Home Search Collections Journals About Contact us My IOPscience

Comparative analysis of environmental impacts of agricultural production systems, agricultural

input efficiency, and food choice

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2017 Environ. Res. Lett. 12 064016

(http://iopscience.iop.org/1748-9326/12/6/064016)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 191.101.88.53
This content was downloaded on 17/06/2017 at 02:17

Please note that terms and conditions apply.

You may also be interested in:

Potential of extensification of European agriculture for a more sustainable food system, focusing
on nitrogen
Hans J M van Grinsven, Jan Willem Erisman, Wim de Vries et al.

Comparison of production-phase environmental impact metrics derived at the farm- and national-scale
for United States agricultural commodities
Christine Costello, Xiaobo Xue and Robert W Howarth

Impacts of European livestock production: nitrogen, sulphur, phosphorus and greenhouse gas
emissions, land-use, water eutrophication and biodiversity
Adrian Leip, Gilles Billen, Josette Garnier et al.

Historical trade-offs of livestock’s environmental impacts

Kyle Frankel Davis, Kailiang Yu, Mario Herrero et al.

Energy and protein feed-to-food conversion efficiencies in the US and potential food security gains
from dietary changes
A Shepon, G Eshel, E Noor et al.

Nitrogen use in the global food system: past trends and future trajectories of agronomic
performance, pollution, trade, and dietary demand
Luis Lassaletta, Gilles Billen, Josette Garnier et al.

Whole farm quantification of GHG emissions within smallholder farms in developing countries
Matthias Seebauer

Global versus local environmental impacts of grazing and confined beef production systems
P Modernel, L Astigarraga and V Picasso
 Environ. Res. Lett. 12 (2017) 064016 https://doi.org/10.1088/1748-9326/aa6cd5

LETTER

Comparative analysis of environmental impacts of agricultural

OPEN ACCESS
production systems, agricultural input efficiency, and food
RECEIVED
28 December 2016 choice
REVISED
5 April 2017
Michael Clark1,4 and David Tilman2,3
ACCEPTED FOR PUBLICATION 1
Natural Resources Science and Management, University of Minnesota, St Paul, Minnesota 55108, United States of America
12 April 2017 2
Department of Ecology, Evolution and Behavior, University of Minnesota, St Paul, Minnesota 55108, United States of America
PUBLISHED 3
Bren School of Environmental Science and Management, University of California Santa Barbara, California 93106, United States of
16 June 2017
America
4
Author to whom any correspondence should be addressed.
Original content from E-mail: [email protected]
this work may be used
under the terms of the
Keywords: life cycle assessment, environmental sustainability, food, organic, agricultural efficiency
Creative Commons
Attribution 3.0 licence. Supplementary material for this article is available online
Any further distribution
of this work must
maintain attribution to
the author(s) and the
title of the work, journal
Abstract
citation and DOI. Global agricultural feeds over 7 billion people, but is also a leading cause of environmental
degradation. Understanding how alternative agricultural production systems, agricultural input
efficiency, and food choice drive environmental degradation is necessary for reducing agriculture’s
environmental impacts. A meta-analysis of life cycle assessments that includes 742 agricultural
systems and over 90 unique foods produced primarily in high-input systems shows that, per unit
of food, organic systems require more land, cause more eutrophication, use less energy, but emit
similar greenhouse gas emissions (GHGs) as conventional systems; that grass-fed beef requires
more land and emits similar GHG emissions as grain-feed beef; and that low-input aquaculture
and non-trawling fisheries have much lower GHG emissions than trawling fisheries. In addition,
our analyses show that increasing agricultural input efficiency (the amount of food produced per
input of fertilizer or feed) would have environmental benefits for both crop and livestock
systems. Further, for all environmental indicators and nutritional units examined, plant-based
foods have the lowest environmental impacts; eggs, dairy, pork, poultry, non-trawling fisheries,
and non-recirculating aquaculture have intermediate impacts; and ruminant meat has impacts
∼100 times those of plant-based foods. Our analyses show that dietary shifts towards low-impact
foods and increases in agricultural input use efficiency would offer larger environmental benefits
than would switches from conventional agricultural systems to alternatives such as organic
agriculture or grass-fed beef.

Introduction These impacts are likely to increase globally over the

next several decades because of increases in population
Global agriculture feeds over 7 billion people, but is growth and income-dependent dietary shifts towards
also a major cause of multiple types of environmental more meat-based diets (Tilman et al 2011, Bajzelj et al
degradation. Agricultural activities emit 25%–33% of 2014, Tilman and Clark 2014, Springmann et al 2016).
greenhouse gases (Steinfeld et al 2006, Edenhofer et al We need to understand the linkages between diets,
2014, Tubiello et al 2014); occupy 40% of Earth’s land agricultural production practices, and environmental
surface (FAO 2016a); account for >70% of freshwater degradation if we are to reduce agriculture’s environ-
withdrawals (Molden 2007), drive deforestation and mental impacts while providing a secure food supply
habitat fragmentation (Ramankutty and Foley 1999) for a growing global population. To quantify these
and resultant biodiversity loss (IUCN 2016); and processes and linkages, we review and synthesize
eutrophy and acidify natural aquatic and terrestrial published information from 742 food production
ecosystems with agrochemicals (Vitousek et al 1997). systems of over 90 foods from 164 published life cycle

© 2017 IOP Publishing Ltd

Environ. Res. Lett. 12 (2017) 064016

assessments (LCAs). LCAs are an internationally The analyses and results presented here expand on
recognized way to account the inputs, outputs, and current knowledge of how food production system,
environmental impacts of a food production system. agricultural input efficiency, and food choice affect
Using our meta-analysis of LCAs, we examine the agriculture’s environmental impacts. The results can be
comparative environmental impacts of different food used to create a more sustainable agricultural future.
production systems, different agricultural input
efficiencies, and different foods.
Food production systems such as organic agricul- Methods
ture and grass-fed beef have been proposed as
potential ways to reduce agriculture’s environmental Publication selection and issues covered
impacts (e.g. Ponisio et al 2014). Organic agriculture, We searched Web of Knowledge, PubMed, AGRICOLA,
for example, is often promoted as having lower and Google Scholar for food LCAs published before July
environmental impacts relative to high-input conven- 2015. We excluded several publications because a lack of
tional systems because it replaces agrochemical inputs defined system boundaries made direct comparisons
with natural inputs such as manure or with ecosystem with other LCAs impossible. In addition, some LCAs
services such as pest control (Azadi et al 2011). Recent conducted by for-profit companies were excluded
analyses examining the comparative impacts of because of potential biases. In total, we used 164
organic and conventional systems have, of necessity, publications that analyzed 742 unique food production
been limited to a few environmental indicators or in systems a (supplementary table 1 available at stacks.iop.
statistical strength of their inferences because of small org/ERL/12/064016/mmedia). We used five different
sample size (Mondelaers et al 2009, Seufert et al 2012, environmental indicators in our analyses. These
Tuomisto et al 2012, Ponisio et al 2014). Recent indicators are greenhouse gas emissions, land use,
increases in the number of published LCAs enables energy use, acidification potential (a measure of nutrient
more complete analysis of the comparative impacts of loading), and eutrophication potential (a measure of
organic and conventional systems across a range of nutrient runoff) to give a broad overview of the
environmental indicators and foods. In addition, we environmental impacts of food production. The data for
combine de novo analyses to determine the compara- other environmental indicators, such as biodiversity
tive environmental impacts of three other sets of impacts, were not present in adequate amounts to
production systems: grass-fed and grain-fed beef; include in our analyses.
trawling and non-trawling fisheries; and greenhouse Our analyses include all relevant pre-farm and on-
grown and open-field produce. farm activities (fertilizer production and application,
Increases in agricultural input efficiency, or the seed production, farm energy use, feed and fodder
amount of food produced per unit of fertilizer or feed production, manure production (when used for
input, may also reduce agriculture’s environmental fertilizer), manure management, infrastructure con-
impact (e.g. Robertson and Swinton 2005). Agricul- struction, etc) and their associated environmental
tural systems depend on fertilizer and feed inputs to impacts up until a food leaves the farm. Our analyses
obtain and/or maintain high productivity. However, are thus of ‘cradle-to-farm gate’ activities; a paucity of
excessive application of these inputs increases data on post-farm gate impacts limited our ability to
agriculture’s environmental impact without increas- analyze them in a balanced manner, although a
ing yields or farmer profits (Vitousek et al 2009). Our previous analysis showed that the vast majority of a
analyses examine the extent to which increases in food’s greenhouse gas emissions stem from ‘cradle-to-
agricultural input efficiency could reduce the farm gate’ activities (Weber and Matthews 2008).
environmental impact of producing a given type of In-depth examples of the activities included in ‘cradle-
food. to-farm gate’ system boundary can be found in
Previous analyses have shown that foods can differ Pelletier (2008), Hokazono and Hayashi (2012), and
greatly in their environmental impact (e.g. Clune et al Torrellas et al (2012).
2017). However, these have been limited to animal- The majority of LCA publications included in
based foods (de Vries and de Boer 2010, Nijdam et al these analyses are from agricultural systems in Europe,
2012) or to a single environmental indicator (e.g. North America, and Australia and New Zealand (86%
Mekonnen and Hoekstra 2010, Clune et al 2017). It is of systems are from these regions). Systems from
thus currently unclear how foods differ in their China (2%), Japan (2%), the rest of Asia (5%), South
impacts across a range of environmental indicators, America (4%), and Africa (.4%) are much less
and whether foods with low impacts for one common. The results presented here are therefore
environmental indicator have similarly low impacts indicative of highly industrialized systems and should
for other environmental indicators. Our meta-analysis be interpreted with this in mind. However, because the
enables us to make these comparisons for five majority of systems analyzed here are highly industri-
environmental indicators: greenhouse gas emissions alized systems, comparisons across publications will be
(GHGs), land use, fossil fuel energy use, eutrophica- more indicative of environmental differences between
tion potential, and acidification potential. foods than if production systems were highly variable.

2
Environ. Res. Lett. 12 (2017) 064016

We found sufficient data to compare the environ- alternative production systems by food item within
mental impacts of four sets of alternative production publication. We first calculated the ratio of impacts of
systems: organic versus conventional systems; grass- different production systems by food item within each
fed versus grain-fed beef; trawling versus non-trawling publication, and then calculated the response ratio by
fisheries; and greenhouse-grown versus open-field taking the natural log of the ratio of impacts (Hedges
produce. We were also able to examine how et al 1999). We then aggregated foods into groups of
agricultural input efficiency, or the amount of food similar food types (cereals; fruits; vegetables; pulses,
produced per unit of agricultural input, affects a food’s nuts and oil crops; dairy and eggs; and meats) to
environmental impact, as well as how foods differ in improve the power of statistical tests. We tested for
their environmental impacts across the five environ- significant differences between alternative production
mental indicators examined. systems using t-tests on the response ratio.

Description of environmental indicators Agricultural input efficiency

Five environmental indicators were used in this In determining how agricultural input efficiency, or
analysis: greenhouse gas emissions, land use, energy the amount of food produced per unit of agricultural
use, acidification potential, and eutrophication po- input, affects a food’s environmental impact, we
tential. The analyses were limited to these indicators performed regressions between a food’s environmen-
because a very limited number of publications tal impact and its nutrient use efficiency in crop
reported data for other indicators such as human systems or its feed use efficiency in livestock systems.
health, ecotoxicity, or biodiversity. An explanation of We limited analyses to non-rice cereal crops and non-
the indicators included in the analyses is below. ruminant livestock because flooding in rice paddies
Greenhouse gas emissions (GHGs) are reported in and digestive processes in ruminants do not make
carbon dioxide equivalents, and include the green- them directly comparable with other crop and
house gas emissions from carbon dioxide, methane, livestock systems. There is not adequate data to
and nitrous oxide. GHGs from activities in the results perform similar analyses limited to ruminant systems:
presented include, but are not limited to, fertilizer comparisons would be severely limited for beef (n ¼ 7
production and application, manure management, for GHGs and n < 5 for all other indicators), and only
enteric fermentation. three studies provide feed use efficiency in dairy
Energy use is reported in kilojoules and includes the systems. For the analysis on nutrient use efficiency, we
energy used during pre-farm and on-farm activities excluded crop systems that applied manure because
including, but not limited to, fertilizer production, the variable nitrogen content of manure made it
infrastructure construction and machinery use. impossible to calculate nitrogen inputs in these
Land use is a measurement of how much land is systems. In total, we examined the agricultural input
occupied during food production. It accounts for land efficiency of 49 non-rice cereal production systems
used to grow crops and/or livestock feed, to house and 53 non-ruminant livestock production systems.
animals, and to pasture ruminants.
Acidification potential is reported in SO2 equiv- Different foods
alents and includes acidification potential from sulfur LCAs commonly report a food’s environmental impact
dioxide, nitrogen oxides, nitrous oxide, and ammonia, on a per mass basis (e.g. impacts per kg of food).
among others. Acidification potential is a measure- However, because the nutritional values of foods come
ment of the potential increase in acidity of an from their calories, protein, and/or micronutrients, and
ecosystem. Excess acidification makes it more difficult not from mass per se, we also calculated a food’s
for plants to assimilate nutrients, and thus results in environmental impacts per kilocalorie, gram protein,
decreased plant growth. Activities such as fertilizer and USDA serving (2016). To compare differences
application, fuel combustion, and manure manage- between broad types of foods, we aggregated foods into
ment are included in the results presented here. 13 food groups composed of similar foods (supplemen-
Eutrophication potential (a measure of nutrifica- tary table 2).
tion) is reported in PO4 equivalents and includes
eutrophication potential from phosphate, nitrogen
oxides, ammonia, and ammonium, among others. Results and discussions
Eutrophication is a measurement of the increase in
nutrients entering an ecosystem. Eutrophication has Environmental impacts of alternative food
substantial environmental impacts including, but not production systems
limited to, algal blooms and aquatic dead zones. Organic versus conventional agriculture
Organic agriculture is a fast-growing sector in many
Alternative production systems western nations, perhaps because it is perceived as
To control for environmental and agronomic differ- being more sustainable or healthier than conventional
ences between publications, as well as differences in agricultural systems (Rigby and Cáceres 2001). Our
nutrient contents between foods, we compared analyses based on 46-paired organic—conventional

3
Environ. Res. Lett. 12 (2017) 064016

Figure 1. Response ratio of the environmental impacts of organic and conventional food production systems. Comparisons were
made within publication to control for agronomic and environmental differences between publications. Plotted on a log base 2 scale,
where a ratio greater than one indicates organic systems have higher impacts; a ratio less than one indicates organic systems have lower
impacts. Bars are means and standard errors.

systems examine the comparative environmental on manure increase the portion of reactive nitrogen in
impacts of these agricultural systems across five organic systems that turns into nitrous oxide, a potent
environmental indicators and a broad range of foods. greenhouse gas (Myhre et al 2013), causing organic
We found that organic systems require 25%–110% and conventional systems to have similar GHG
more land use (p < 0.001; n ¼ 37), use 15% less emissions. Because we limited comparisons to within
energy (p ¼ .0452; n ¼ 33), and have 37% higher publication, the results presented here are therefore
eutrophication potential (p ¼ .0383; n ¼ 20) than indicative of comparative environmental differences of
conventional systems per unit of food. In addition, organic and conventional systems at a local scale. It is,
organic and conventional systems did not significantly however, possible that the comparative environmental
differ in their greenhouse gas emissions (p ¼ .5923; impacts of organic and conventional systems might
n ¼ 44) or acidification potential (p ¼ .299; n ¼ 26), differ at a regional, national, or global scale (e.g.
although these were 4% lower and 13% higher in Bengtsson et al 2005 and Phalan et al 2011).
organic systems, respectively (figure 1). Previous analyses have shown that increasing
The differences in environmental impacts between nutrient application and adopting techniques such as
organic and conventional systems are primarily driven rotational farming, cover cropping, multi-cropping,
by differences in nutrient management techniques. and polyculture in organic systems can halve the land
Organic agriculture is largely dependent on manure as use difference between organic and conventional
a nitrogen input in contrast to conventional agricul- systems (Seufert et al 2012, Ponisio et al 2014).
ture’s use of synthetic fertilizers. Application of Additionally, while the overall pattern is for higher
manure, which releases nutrients in response to land use in organic systems, organic systems have
environmental conditions and not crop nutrient similar land use for legumes and perennial crops while
demand (Seufert et al 2012), often results in temporal the land use difference between organic and conven-
mismatches between nutrient availability and nutrient tional systems is smaller in rain-fed systems and in
demand and thereby increases the proportion of systems with weakly-acidic to weakly-alkaline soils
nutrients that are not assimilated by plants (Cassman (Pimentel et al 2005, Seufert et al 2012).
et al 2002). These temporal mismatches in organic Organic systems might offer health and environ-
systems result in reduced crop growth and yields and mental benefits we could not investigate with our data
thus in increased land use. In addition, nutrient set. Organic foods have higher micronutrient con-
applications not incorporated into plant growth cause centrations (Hunter et al 2011, Palupi et al 2012) and
eutrophication and acidification, thereby driving the lower pesticide residues (Baker et al 2002) than
higher eutrophication potential and tendency for conventional foods, although these differences may
higher acidification potential in organic systems. In not translate into improved human health outcomes
contrast, energy use is lower in organic systems (Dangour and Lock 2010, Hunter et al 2011). On-farm
because of organic’s reduced reliance on energy- and near-farm biodiversity (Mäder et al 2002,
intensive synthetic fertilizer and pesticide inputs. Bengtsson et al 2005, Hole et al 2005) tends to be
GHG emissions are similar in organic and conven- higher in organic agricultural systems, probably
tional systems because of the trade-off between because of its lower fertilizer, herbicide and pesticide
application of synthetic fertilizer in conventional inputs. In addition, soil organic carbon is higher in
systems and use of manure in organic systems. Indeed, organic systems (Gattinger et al 2012) because manure
while production of conventional fertilizer is energy- application promotes carbon storage in agricultural
and GHG-intensive, mismatches between nutrient soils. However, organic agriculture would likely have a
availability and demand in organic systems dependent net negative impact on biodiversity and soil organic

4
Environ. Res. Lett. 12 (2017) 064016

of grass, silage, and fodder are often lower, possibly

because the land on which they are grown is often less
fertile than that used to produce feed (e.g. maize, soy,
etc) used in grain-fed systems. The combination of
higher feed inputs and lower nutritional crop yields for
feeds drive the higher land use observed in grass-fed
systems. Additionally, because grass-fed cattle grow
slower and are slaughtered 6–12 months older than
grain-fed cattle, lifetime methane emissions, and thus
GHGs per unit of food, tend to be higher for grass-fed
beef. The source of GHGs in grass-fed and grain-fed
Figure 2. Response ratio of the environmental impact of
grass-fed and grain-fed beef. Comparisons were made within
systems further supports this explanation. Indeed,
publication to control for agronomic and environmental 30% and 52% of GHGs in grain-fed systems result
differences between study locations. A ratio greater than one from feed production and enteric fermentation,
indicates grass-fed beef has higher impacts; a ratio less than
one indicates grass-fed beef has lower impacts. Bars are means respectively. In contrast, feed production and enteric
and standard errors. are responsible for 20% and 61% of GHGs,
respectively, in grass-fed systems.
Grass-fed beef may have environmental and
carbon at larger spatial scales because of the greater human health benefits we could not analyze with
land clearing required under organic agriculture and our data. For example, grass-fed systems promote soil
because biodiversity (Balmford et al 2005, Phalan et al carbon sequestration (Derner and Schuman 2007) and
2011) and carbon stocks (Gilroy et al 2014) decrease within-pasture nutrient cycling while simultaneously
dramatically with conversion from natural habitats. decreasing eutrophication (Smith et al 2013). Addi-
Although organic systems have higher land use tionally, grass-fed beef has higher micronutrient
and eutrophication potential and tend to have higher concentrations and a fatty acid profile that might
acidification potential, this should not be taken as an lead to improved human health outcomes relative to
indication that conventional systems are more consumption of grain-fed beef (Daley et al 2010).
sustainable than organic systems. Conventional Furthermore, grass-fed beef may promote food
practices require more energy use and are reliant on security in cropland-scarce regions because it can be
high nutrient, herbicide, and pesticide inputs that can grown on land not suitable for crop production (Smith
have negative impacts on human health (Townsend et al 2013).
2003, Schwarzenbach et al 2010, Mostafalou and
Abdollahi 2013) and the environment (Vitousek et al Trawling versus non-trawling fisheries versus
2009, Foley et al 2011). Developing production aquaculture
systems that integrate the benefits of conventional, We classified commercial fisheries into trawling
organic, and other agricultural systems is necessary for fisheries—where nets are physically dragged across
creating a more sustainable agricultural future. a seabed—and non-trawling fisheries (midwater
trawl, short and long-line fishing, and seine nets).
Grass-fed versus grain-fed beef Our analyses of 10 paired systems of trawling and
We quantitatively analyzed the environmental differ- non-trawling fisheries show that trawling fisheries
ences between grass-fed and grain-fed beef using 7 emit 2.8 times more GHGs than non-trawling
paired grass- and grain-fed systems. We define grass-fed fisheries (p ¼ .004; n ¼ 10) (figure 3) because of
systems as those where beef is raised solely on pasture or the high fuel requirements of dragging a net across a
seasonally on pasture and supplemented diets of grass, seabed. Response ratios differ greatly between fish,
silage, and fodder while overwintering. We found that with non-schooling fish (flat fish) having compara-
grass-fed beef had higher land use requirements than tively higher impacts under trawl fisheries than do
grain-fed beef (p ¼ .0381, n ¼ 4). Grass-fed and grain- fish that form schools (mackerel, cod). Previous
fed beef had similar impacts per unit food for the other analyses have also shown that trawl fisheries
environmental impacts examined (p > .05 for all other negatively impact non-targeted species through high
indicators), although grass-fed beef had, on average, bycatch rates relative to other fish capture methods
19% higher GHGs (p ¼ .2218; n ¼ 7) per unit food than and through ecosystem degradation from dragging a
grain-fed beef (figure 2). net across a seabed (Dayton et al 1995). Shifting from
The higher land use and tendency for higher GHG trawling to non-trawling fisheries would thus
emissions in grass-fed beef stem from the lower simultaneously decrease GHGs, bycatch rates, and
macronutrient densities and digestibility of feeds used ecosystem degradation.
in grass-fed systems (Feedipedia 2016) because they Aquaculture, which accounts for ∼45% of global
cause grass-fed beef to require higher feed inputs per fish production, could be a sustainable alternative to
unit of beef produced than grain-fed systems. wild-caught fisheries (FAO 2016b). Our examination
Furthermore, the nutritional yields (e.g. kcal ha1) of 142 fishery and aquaculture systems indicates that,

5
Environ. Res. Lett. 12 (2017) 064016

Figure 3. Response ratio of the greenhouse gas emissions of Figure 5. Response ratio of environmental impacts of
trawling and non-trawling fisheries (e.g. line, purse and seine greenhouse grown and open field produce. Comparisons
net). A ratio greater than one indicates trawling fisheries have were made within publication to control for agronomic and
higher greenhouse gas emissions than non-trawling fisheries. environmental differences between study locations. Bars are
Bars are means and standard errors. means and standard errors. A ratio greater than one indicates
greenhouse systems have higher impacts; a ratio less than one
indicates greenhouse systems have lower impacts.

rice-catfish agriculture-aquaculture systems have

comparatively low impacts (Folke and Kautsky
1992, Páez-Osuna 2001).

Greenhouse grown versus open-field produce

Greenhouse systems allow crops to be grown in
climates and regions not suitable for crop production.
Our analysis of five paired greenhouse–open-field
systems shows that greenhouse production systems
tend to emit almost three times more GHGs (figure 5;
p ¼ .089) because of the energy required to maintain
Figure 4. Greenhouse gas emissions from non-trawling (e.g. greenhouses at ideal growing conditions. While our
line, purse and seine net) and trawling fisheries, and from
non-recirculating (e.g. pond, bag, flow-through) and analyses show that, on average, greenhouse produc-
recirculating aquaculture systems per gram protein. tion systems tend to have higher energy use than open-
Significant differences are denoted by letters and were field systems, it is important to note that energy
calculated using a Tukey post-hoc test.
requirements and thus greenhouse gas emissions can
differ greatly between greenhouses. For example,
greenhouses that are both heated and lighted will
on average, non-recirculating aquaculture (e.g. aqua- require substantially more energy to maintain than
culture in ponds, fjords, rivers, etc) and non-trawling will greenhouses that are neither heated nor lighted.
fisheries emitted similar GHGs per unit of food and Land use in greenhouse systems was, on average, one
had emissions similar to pork, poultry, and dairy quarter of that in open fields, but this difference was
(figures 4 and S1). In contrast, trawling fisheries and not significant (p ¼ .166; n ¼ 3). Crop yields are
recirculating aquaculture (in tanks and other systems higher, and thus land use lower, in greenhouse systems
in which pumps and filters are used) emitted several because they are maintained at ideal conditions for
times more GHGs than non-trawling fisheries and plant growth. The limited sample size of these analyses
non-recirculating aquaculture because of their high prevents concrete conclusions from being drawn.
energy requirements (figure 4). Aquaculture-raised Future analyses examining the environmental differ-
fish from non-recirculating systems could thus be a ences between greenhouse and open-field production
lower-emission alternative to trawling fisheries, an systems are needed to fully elucidate their comparative
equal-emission alternative to non-trawling fisheries, environmental impacts.
and could alleviate pressure on over-harvested
fisheries (Costello et al 2012).
There can be marked differences in environmental Environmental impacts of agricultural input
impacts even among the lower-impact non-recirculat- efficiency
ing aquaculture systems. For instance, aquaculture at
high fish densities can eutrophy closed bodies of water We found large differences among studies in the
and cause gene exchange between farmed and wild fish environmental impacts of producing the same food
varieties (FAO 2016b). In addition, shrimp aquacul- (supplemental figure 1). To examine why foods may
ture systems that require deforestation of mangroves vary in their environmental impacts, we analyzed
have high environmental impacts while integrated agricultural input efficiency, or the amount of food

6
Environ. Res. Lett. 12 (2017) 064016

Figure 6. Correlations between nitrogen use efficiency, or calories produced per g of nitrogen input, and the environmental impacts of
non-rice cereal crops. Regression lines are reciprocal fits between nitrogen use efficiency and a food’s environmental impact. All
relationships are significant at p < .05 except for acidification potential.

Figure 7. Correlations between feed use efficiency, or kcal of food produced per kcal of feed input, and environmental impacts in non-
ruminant livestock systems. Regression lines are reciprocal fits between feed use efficiency and a food’s environmental impacts. All
relationships are significant at p < .05.

produced per unit of fertilizer or feed input, in 49 non- environmental indicator are almost always downward
rice cereal production systems and 53 non-ruminant sloping and significant. Increasing agricultural input
livestock systems. We found that higher agricultural efficiency reduces a food’s environmental impact
input efficiency is consistently associated with lower because of the environmental impacts of producing
environmental impacts for both non-rice cereal agricultural inputs such as fertilizer, pesticides, and
systems (figure 6) and non-ruminant livestock systems livestock feeds. However, the environmental benefits
(figure 7). While the fits shown in figures 6 and 7 are of increasing agricultural input efficiency would not be
across all food items, fits for individual food by equal across all systems, with improvements in

7
Environ. Res. Lett. 12 (2017) 064016

Figure 8. Environmental impacts of broad groups of foods per kilocalorie. The environmental indicators examined are greenhouse
gas emissions, land use, energy use, acidification potential (Acid. Pot.) and eutrophication potential (Eut. Pot.). Bars show means and
standard errors. Plant-based foods are in green; dairy and eggs are in grey; meats are in red; and seafood is in blue. Data from foods
grown in greenhouses are not included when plotting this figure. Trawl Fishery ¼ bottom-trawling fisheries; NT Fishery ¼ all other
fisheries (e.g. line, purse net, seine net, etc); Recirc Aqua ¼ recirculating aquaculture; NR Aqua ¼ non-recirculating aquaculture (e.g.
pond, net pen, flow-through, etc).

agricultural input efficiency having the largest climatic and soil conditions often affect agricultural
environmental benefit in the least efficient systems. input efficiency. Indeed, spatially locating food
Further, improving efficiency in more efficient systems production in areas with the most suitable climatic
may only be possible at an economic cost. Emphasis and soil conditions for a crop can increase agricultural
should therefore be placed on improving efficiency in input efficiency and decrease environmental impacts
less efficient systems, although efficiency improve- (Polasky et al 2008, Johnson et al 2014, Chaplin-
ments in more efficient systems would still have Kramer et al 2015). For example, preferentially
environmental benefits. locating agricultural land to maximize single ecosys-
Several technologies and management techniques tem services would increase carbon stores by
can increase agricultural input efficiency. Precision ∼6 billion metric tonnes (worth ∼$1 trillion 2012
farming, where nutrient and pesticide inputs are USD) (Johnson et al 2014) and substantially decrease
temporally and spatially applied to match crop projected rates of agriculturally-driven biodiversity
requirements, has increased fertilizer input efficiency loss (Chaplin-Kramer et al 2015). Globally leveraging
and farmer profits without decreasing crop yields for a environmental and soil conditions to increase
variety of crops in geographically diverse areas agricultural input efficiency could thus provide
(Robertson and Vitousek 2009). Conservation tillage substantial environmental benefits.
and cover cropping, particularly with nitrogen fixing
crops because they simultaneously reduce required
nitrogen inputs, also increase fertilizer input efficiency Environmental Impacts of different foods
by reducing nutrient loss from agricultural systems
(Robertson and Vitousek 2009, Ponisio et al 2014). Many analyses have shown that dietary choice can
Feed input efficiency in livestock systems can also be greatly influence the environmental impacts of the
increased. For example, pork from pigs fed diets agricultural food system (de Vries and de Boer 2010,
supplemented with amino acids required less feed and Nijdam et al 2012, Tilman and Clark 2014, Clune et al
emitted 5% fewer GHGs and had 28% lower 2017), although these analyses were limited to animal-
eutrophication potential than pork from pigs fed based foods or a single environmental indicator. Our
unsupplemented diets (Ogino et al 2013). Similar analyses expand on these earlier studies and show that
benefits have also been found in poultry, beef, and foods with low impact for one environmental
dairy systems (Robertson and Vitousek 2009). In indicator tend to have low impacts for all environ-
addition, using agricultural wastes and byproducts as mental indicators examined (figure 8). Indeed, for all
animal feeds could reduce the environmental impacts indicators examined, ruminant meat (beef, goat and
of livestock production by 20% without reducing food lamb/mutton) had impacts 20–100 times those of
quality or farmer profits (zu Ermgassen et al 2016). plants while milk, eggs, pork, poultry, and seafood had
The location of food production can also influence impacts 2–25 times higher than plants per kilocalorie
its environmental impact because differences in of food produced. This clear trend of ruminant meat

8
Environ. Res. Lett. 12 (2017) 064016

having high impacts and other animal-based foods be placed on improving efficiency in the least efficient
having intermediate impacts also holds when foods are agricultural systems.
examined per gram protein, USDA serving, or unit The difference in environmental impacts between
mass (supplemental figure 1). Isocaloric shifts from foods is large compared to the difference between
high-impact to lower-impact but nutritionally similar production systems and systems with different
foods, such as shifts from ruminant meats to fish, agricultural input efficiencies producing the same
pork, poultry, or legumes, would have large diet- food. Ruminant meats, for example, have impacts that
related environmental benefits while also improving are 3–10 times those of other animal-based foods and
human health outcomes (e.g. Tilman and Clark 2014). 20–100 times those of plant-based foods for all
These dietary shifts, however, would likely decrease the indicators examined. Because the majority of produc-
total cost of the diet; it is possible that increased tion systems included in these analyses are from
consumption of other material goods could offset the Europe and North America, the results presented here
environmental benefits of consuming lower-impact are indicative of trends in highly industrialized and
foods. high-input agricultural systems. Analyses of the
Most of the 742 LCA food analyses used were environmental impacts of low-input agricultural
based on high-input systems in Europe and North systems are necessary to elucidate the extent to which
America; the results presented here are thus indicative the trends observed here also apply to lower-input
of the impacts of high-input systems in developed agricultural systems.
nations. In contrast, the impacts of low-input systems The analyses presented here greatly expand current
common in developing nations are not yet well knowledge of the environmental impacts of food
studied, although a recent analysis indicates that production. However, there are still large knowledge
GHGs may be higher in these systems because of lower gaps which, if addressed, would further our under-
agricultural input efficiency (Herrero et al 2013). LCA standing of agriculture’s environmental impacts. For
analyses on less-studied but nutritionally and cultur- example, analyses on the environmental impacts of
ally important foods such as quinoa, cassava, and agricultural systems in low-income countries, on
millet, as well as analyses from additional regions and staple crops not common in Westernized diets
management regimes would provide further insight (quinoa, yams, sorghum, millet, etc), on fish produced
and a clearer understanding of the environmental via aquaculture, and on agricultural input efficiency in
impacts of different foods and food systems globally. non-cereal crops and in ruminant systems are limited.
In addition, agricultural production has a multitude of
environmental impacts beyond the five environmental
Conclusions indicators analyzed here; few LCAs analyses have
examined agriculture’s other environmental impacts
Our analyses show that the comparative environmen- such as water use, pesticide use, or impact on
tal impacts of agricultural production systems differ biodiversity. Analyses into these, and other, under-
depending on the systems, food, and environmental studied aspects of agriculture’s environmental impacts
indicator examined. Per unit of food produced, are needed to more fully elucidate agriculture’s entire
organic systems had higher land use and eutrophica- environmental impact.
tion potential, tended to have higher acidification Despite current knowledge gaps, it is clear that
potential, did not offer benefits in GHGs, but had current agricultural trajectories would substantially
lower energy use; trawling fisheries emitted almost increase agriculture’s environmental impacts by
3 times more GHGs than non-trawling fisheries; grass- midcentury (Tilman et al 2001,Tilman et al 2011,
fed beef required more land and tended to emit more Bajzelj et al 2014, Tilman and Clark 2014). Many
GHGs than grain-fed beef; and high agricultural interventions would, however, greatly reduce agricul-
efficiency was consistently correlated with lower ture’s future environmental impacts. Adoption of low-
environmental impacts. Combining the benefits of meat and no-meat diets in nations with excess meat
different production systems, for example organic’s consumption (Springmann et al 2016), sustainable
reduced reliance on chemical inputs with the high increases in crop yields (Foley et al 2011, Mueller et al
yields of conventional systems would result in a more 2012), and adoption of low-impact and otherwise
sustainable agricultural system. more efficient agricultural systems (Robertson and
Agricultural input efficiency, or the amount of Vitousek 2009), would offer large environmental
food produced per unit of input, is inversely correlated benefits. In addition, over 30% of food production is
with a food’s environmental impact in non-rice cereal wasted; reducing food waste would offer environmen-
systems and non-ruminant livestock systems. Increas- tal benefits without requiring shifts in production
ing agricultural input use efficiency would have practices or diets (Foley et al 2011). Implementing
environmental benefits without necessitating dietary policy and education initiatives designed to increase
change. However, because the marginal environmental adoption of lower-impact foods, of lower-impact
benefits of increasing agricultural input efficiency is production systems, and of systems with high
larger in less efficient systems, special emphasis should agricultural input efficiency is necessary before

9
Environ. Res. Lett. 12 (2017) 064016

agriculture causes substantial, and potentially irre- Gattinger A et al 2012 Enhanced top soil carbon stocks under
organic farming Proc. Natl Acad. Sci. USA 109 18226–31
versible, environmental damage. Gilroy J J, Woodcock P, Edwards F A, Wheeler C, Baptiste B L,
Medina Uribe C A, Haugaasen T and Edwards D P 2014
Cheap carbon and biodiversity co-benefits from forest
Acknowledgments regeneration in a hotspot of endemism Nat. Clim. Change
4 503–7
We would like to thank Kaitlin Kimmel, George Furey, Hedges L V, Gurevitch J and Curtis P S 1999 The meta-analysis
Adam Clark, David Williams, and James Gerber for of response ratios in experimental ecology Ecology 80
1150–6
comments on the manuscript. We would also like to Herrero M, Havlík P, Valin H, Notenbaert A, Rufino M C,
thank the Balzan Foundation, the McKnight Presi- Thornton P K, Blümmel M, Weiss F, Grace D and
dential Chair (DT), and the University of California, Obersteiner M 2013 Biomass use, production, feed
Santa Barbara for support. efficiencies, and greenhouse gas emissions from global
livestock systems Proc. Natl Acad. Sci. USA 110
20888–93
References Hokazono S and Hayashi K 2012 Variability in environmental
impacts during conversion from conventional to organic
Azadi H, Schoonbeek S, Mahmoudi H, Derudder B, De Maeyer farming: a comparison among three rice production
P and Witlox F 2011 Organic agriculture and sustainable systems in Japan J. Clean. Prod. 28 101–12
food production system: main potentials Agric. Ecosyst. Hole D G, Perkins A J, Wilson J D, Alexander I H, Grice P V
Environ. 144 92–4 and Evans A D 2005 Does organic farming benefit
Bajzelj B, Richards K S, Allwood J M, Smith P, Dennis J S, biodiversity? Biol. Conserv. 122 113–30
Curmi E and Gilligan C A 2014 Importance of food- Hunter D, Foster M, McArthur J O, Ojha R, Petocz P and
demand management for climate mitigation Nat. Clim. Samman S 2011 Evaluation of the micronutrient
Change 4 924–9 composition of plant foods produced by organic and
Baker B P, Benbrook C M, Groth E and Lutz Benbrook K 2002 conventional agricultural methods Crit. Rev. Food Sci.
Pesticide residues in conventional, integrated pest Nutr. 51 571–82
management (IPM)-grown and organic foods: insights IUCN 2016 IUCN Red List Threat. Species Version 2016-5
from three US data sets Food Addit. Contam. 19 427–46 Johnson J A, Runge C F, Senauer B, Foley J and Polasky S 2014
Balmford A, Green R E and Scharlemann J P W 2005 Sparing Global agriculture and carbon trade-offs Proc. Natl Acad.
land for nature: exploring the potential impact of changes Sci. USA 2014 1–6
in agricultural yield on the area needed for crop Mäder P, Fließbach A, Dubois D, Gunst L, Fried P and Niggli U
production Glob. Change Biol. 11 1594–605 2002 Soil fertility and biodiversity in organic farming
Bengtsson J, Ahnström J and Weibull A C 2005 The effects of Science 296 1694–7
organic agriculture on biodiversity and abundance: a meta- Mekonnen M M and Hoekstra A Y 2010 The Green, Blue and
analysis J. Appl. Ecol. 42 261–9 Grey Water Footprint of Farm Animals and Animal
Cassman K G, Dobermann A and Walters D T 2002 Products vol 1 (Delft: UNESCO-IHE Institute for Water
Agroecosystems, nitrogen-use efficiency, and nitrogen Education)
management A J. Hum. Environ. 31 132–40 Molden D 2007 Comprehensive assessment of water
Chaplin-Kramer R et al 2015 Spatial patterns of agricultural management in agriculture Water for Food, Water for Life:
expansion determine impacts on biodiversity and carbon A Comprehensive Assessment of Water Management in
storage Proc. Natl Acad. Sci. 112 7402–7 Agriculture (London: Earthscan International Water
Clune S, Crossin E and Verghese K 2017 Systematic review of Management Institute)
greenhouse gas emissions for different fresh food categories Mondelaers K, Aertsens J and Van Huylenbroeck G 2009 A
J. Clean. Prod. 140 766–83 meta-analysis of the differences in environmental impacts
Costello C, Ovando D, Hilborn R, Gaines S D, Deschenes O and between organic and conventional farming Br. Food J. 111
Lester S E 2012 Status and solutions for the world’s 1098–119
unassessed fisheries Science 338 517–20 Mostafalou S and Abdollahi M 2013 Pesticides and human
Daley C A, Abbott A, Doyle P S, Nader G A and Larson S 2010 chronic diseases: evidences, mechanisms, and perspectives
A review of fatty acid profiles and antioxidant content in Toxicol. Appl. Pharmacol. 268 157–77
grass-fed and grain-fed beef Nutr. J. 9 10 Mueller N D, Gerber J S, Johnston M, Ray D K, Ramankutty N
Dangour A and Lock K 2010 Nutrition-related health effects of and Foley J A 2012 Closing yield gaps through nutrient
organic foods: a systematic review Am. J. Clin. Nut. 92 and water management Nature 490 254–7
203–10 Myhre G et al 2013 Anthropogenic and natural radiative forcing
Dayton P K, Thrush S F, Agardy M T and Hofman R J 1995 Climate Change 2013: The Physical Science Basis.
Environmental effects of marine fishing Aquat. Conserv. Contribution of Working Group I to the Fifth Assessment
Mar. Freshw. Ecosyst. 5 205–32 Report of the Intergovernmental Panel on Climate Change
Derner J D and Schuman G E 2007 Carbon sequestration and ed T F Stocker, D Qin, G-K Plattner, M Tignor, S K Allen,
rangelands: a synthesis of land management and J Boschung, A Nauels, Y Xia, V Bex and P Midgley
precipitation effects J. Soil Water Conserv. 62 77–85 (Cambridge: Cambridge University Press) pp 659–740
Edenhofer O et al 2014 Climate Change 2014: Mitigation of Nijdam D, Rood T and Westhoek H 2012 The price of protein:
Climate Change (Cambridge: Cambridge University Press) review of land use and carbon footprints from life cycle
FAO 2016a www.faostat.fao.org assessments of animal food products and their substitutes
FAO 2016b The State of World Fisheries and Aquaculture: Food Policy 37 760–70
Contributing to Food Security and Nutrition for All (Rome: Ogino A, Osada T, Takada R, Takagi T, Tsujimoto S, Tonoue T,
FAO) Matsui D, Katsumata M, Yamashita T and Tanaka Y 2013
Feedipedia 2016 Animal Feed Resources Information System Life cycle assessment of Japanese pig farming using low-
(INRACIRADAFZ and FAO) www.feedipedia.org protein diet supplemented with amino acids Soil Sci. Plant
Foley J A et al 2011 Solutions for a cultivated planet Nature 478 Nutr. 59 107–18
337–42 Páez-Osuna F 2001 The environmental impact of shrimp
Folke C and Kautsky N 1992 Aquaculture with its environment: aquaculture: causes, effects, and mitigating alternatives
prospects for sustainability Ocean Coast. Manage. 17 5–24 Environ. Manage. 28 131–40

10
Environ. Res. Lett. 12 (2017) 064016

Palupi E, Jayanegara A, Ploeger A and Kahl J 2012 Comparison of Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M and de
nutritional quality between conventional and organic dairy Haan C 2006 Livestock’s Long Shadow: Environmental
products: a meta-analysis J. Sci. Food Agric. 92 2774–81 Issues and Options (Rome: Food and Agriculture
Pelletier N 2008 Environmental performance in the US broiler Organization of the United Nations)
poultry sector: life cycle energy use and greenhouse gas, Tilman D, Balzer C, Hill J and Befort B L 2011 Global food
ozone depleting, acidifying and eutrophying emissions demand and the sustainable intensification of agriculture
Agric. Syst. 98 67–73 Proc. Natl Acad. Sci. USA 108 20260–4
Phalan B, Onial M, Balmford A and Green R E 2011 Reconciling Tilman D and Clark M 2014 Global diets link environmental
food production and biodiversity conservation: land sustainability and human health Nature 515 518–22
sharing and land sparing compared Science 333 1289–91 Tilman D, Fargione J, Wolff B, D’Antonio C, Dobson A,
Pimentel D, Hepperly P, Hanson J, Douds D and Seidel R 2005 Howarth R, Schindler D, Schlesinger W H, Simberloff D
Environmental, energetic, and economic comparisons of and Swackhamer D 2001 Forecasting agriculturally driven
organic and conventional farming systems Bioscience 55 573 global environmental change Science 292 281–4
Polasky S et al 2008 Where to put things? Spatial land Torrellas M, Antón A, López J C, Baeza E J, Parra J P, Muñoz P
management to sustain biodiversity and economic returns and Montero J I 2012 Torrellas M, Antón A, López J C,
Biol. Conserv. 141 1505–24 Baeza E J, Parra J P, Muñoz P and Montero J I 2012 LCA
Ponisio L C, M’Gonigle L K, Mace K C, Palomino J, de Valpine of a tomato crop in a multi-tunnel greenhouse in Almeria
P and Kremen C 2014 Diversification practices reduce Int. J. Life Cycle Assess. 17 863–75
organic to conventional yield gap Proc. R. Soc. B 282 1799 Townsend A R 2003 Human health effects of a changing global
Ramankutty N and Foley J A 1999 Estimating historical changes nitrogen cycle Front. Ecol. Environ. 1 240–6
in global land cover: croplands from 1700 to 1992 Glob. Tubiello F N, Salvatore M, Cóndor Golec R D Ferrara A, Rossi
Biogeochem. Cycles 13 997–1027 S, Biancalani R, Federici S, Jacobs H and Flammini A 2014
Rigby D and Cáceres D 2001 Organic farming and the Agriculture, Forestry and Other Land Use Emissions by
sustainability of agricultural systems Agric. Syst. 68 Sources and Removals by Sinks vol 2 (Rome: Food and
21–40 Agriculture Organization of the United Nations) pp 4–89
Robertson G P and Swinton S M 2005 Reconciling agricultural Tuomisto H L, Hodge I D, Riordan P and Macdonald D W 2012
productivity and environmental integrity: a grand Does organic farming reduce environmental impacts? a
challenge for agriculture Front. Ecol. Environ. 3 38–46 meta-analysis of European research J. Environ. Manage.
Robertson G P and Vitousek P M 2009 Nitrogen in agriculture: 112 309–20
balancing the cost of an essential resource Annu. Rev. United States Deparment of Agriculture 2016 ChooseMyPlate.gov
Environ. Resour. 34 97–125 Vitousek P M, Aber J D, Howarth R W, Likens G E A P,
Schwarzenbach R P, Egli T, Hofstetter T B, von Gunten U and Schindler D W, Schlesinger W H and Tilman D 1997
Wehrli B 2010 Global water pollution and human health Human alteration of the global nitrogen cycle: sources and
Annu. Rev. Environ. Resour. 35 109–36 consequences Ecol. Appl. 7 737
Seufert V, Ramankutty N and Foley J A 2012 Comparing the Vitousek P M et al 2009 Nutrient imbalances in agricultural
yields of organic and conventional agriculture Nature 485 development Science 324 1519–20
229–32 de Vries M and de Boer I J M 2010 Comparing environmental
Smith J, Sones K, Grace D, MacMillan S, Tarawali S and Herrero impacts for livestock products: a review of life cycle
M and International 2013 Beyond milk, meat, and eggs: assessments Livest. Sci. 128 1–11
role of livestock in food and nutrition security Anim. Weber C L and Matthews H S 2008 Food-miles and the relative
Front. 3 6–13 climate impacts of food choices in the United States
Springmann M, Godfray H C J, Rayner M and Scarborough P Environ. Sci. Technol. 42 3508–13
2016 Analysis and valuation of the health and climate zu Ermgassen E K H J Phalan B, Green B and Balmford A 2016
change cobenefits of dietary change Proc. Natl Acad. Sci. Reducing the land use of EU pork production: where
USA 113 4146–51 there’s swill, there’s a way Food Policy 58 35–48

11