Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
R ES E A RC H
W
Total arable land and freshwater withdrawals
ith current diets and production prac- and predominantly Western European producers reconcile to FAO estimates. Emissions from de-
tices, feeding 7.6 billion people is degrad- (12–16) and have not corrected for important meth- forestation and agricultural methane fall within
ing terrestrial and aquatic ecosystems, odological differences between LCAs (12–16). Here, ranges of independent models (17).
depleting water resources, and driving we present a globally reconciled and methodolog-
climate change (1, 2). It is particularly ically harmonized database on the variation in food’s Environmental impacts of the entire
challenging to find solutions that are effective multiple impacts. Our results show the need for food supply chain
across the large and diverse range of producers far-reaching changes in how food’s environmental Today’s food supply chain creates ~13.7 billion
that characterize the agricultural sector. More impacts are managed and communicated. metric tons of carbon dioxide equivalents (CO2eq),
than 570 million farms produce in almost all the 26% of anthropogenic GHG emissions. A further
world’s climates and soils (3), each using vastly Building the multi-indicator 2.8 billion metric tons of CO2eq (5%) are caused
different agronomic methods; average farm sizes global database by nonfood agriculture and other drivers of de-
vary from 0.5 ha in Bangladesh to 3000 ha in We derived data from a comprehensive meta- forestation (17). Food production creates ~32%
Australia (3); average mineral fertilizer use ranges analysis, identifying 1530 studies for potential of global terrestrial acidification and ~78% of
from 1 kg of nitrogen per ha in Uganda to 300 kg inclusion, which were supplemented with addi- eutrophication. These emissions can fundamen-
in China (4); and although four crops provide half tional data received from 139 authors. Studies tally alter the species composition of natural
of the world’s food calories (4), more than 2 million were assessed against 11 criteria designed to ecosystems, reducing biodiversity and ecological
distinct varieties are recorded in seed vaults (5). standardize methodology, resulting in 570 suit- resilience (19). The farm stage dominates, rep-
Further, products range from minimally to heavily able studies with a median reference year of resenting 61% of food’s GHG emissions (81%
processed and packaged, with 17 of every 100 kg of 2010 (17). The data set covers ~38,700 commer- including deforestation), 79% of acidification,
food produced transported internationally, increas- cially viable farms in 119 countries (fig. S2) and and 95% of eutrophication (table S17).
ing to 50 kg for nuts and 56 kg for oils (4). 40 products representing ~90% of global pro- Today’s agricultural system is also incredibly
Previous studies have assessed aspects of this tein and calorie consumption. It covers five im- resource intensive, covering ~43% of the world’s
heterogeneity by using geospatial data sets (6–8), portant environmental impact indicators (18): ice- and desert-free land. Of this land, ~87% is
but global assessments using the inputs, outputs, land use; freshwater withdrawals weighted by for food and 13% is for biofuels and textile crops
and practices of actual producers have been lim- local water scarcity; and GHG, acidifying, and or is allocated to nonfood uses such as wool and
ited by data. The recent rapid expansion of the eutrophying emissions. For crops, yield repre- leather. We estimate that two-thirds of freshwater
life cycle assessment (LCA) literature is providing sents output for a single harvest. Land use in- withdrawals are for irrigation. However, irriga-
this information by surveying producers around cludes multicropping (up to four harvests per tion returns less water to rivers and groundwater
the world. LCA then uses models to translate pro- year), fallow phases (uncultivated periods be- than industrial and municipal uses and pre-
ducer data into environmental impacts with suf- tween crops), and economic allocation to crop dominates in water-scarce areas and times of
ficient accuracy for most decision-making (9–11). coproducts such as straw. This makes it a stron- the year, driving 90 to 95% of global scarcity-
To date, efforts to consolidate these data or build ger indicator of both farm productivity and weighted water use (17).
new large-scale data sets have covered greenhouse food security than yield.
gas (GHG) emissions only (8, 12, 13), agriculture The system we assess begins with inputs (the Highly variable and skewed
only (13–16), small numbers of products (8, 14–16), initial effect of producer choice) and ends at re- environmental impacts
tail (the point of consumer choice) (fig. S1). For We now group products by their primary dietary
each study, we recorded the inventory of out- role and express impacts per unit of primary
1
Department of Zoology, University of Oxford, New Radcliffe puts and inputs (including fertilizer quantity nutritional benefit (Fig. 1 and fig. S3). Immedi-
House, Oxford OX2 6GG, UK. 2School of Geography and the and type, irrigation use, soil, and climatic con- ately apparent in our results is the high variation
Environment, University of Oxford, South Parks Road, Oxford
OX1 3QY, UK. 3Agroscope, Agroecology and Environment Research
ditions). Where data were not reported, for ex- in impact among both products and producers.
Division, LCA Research Group, CH-8046 Zürich, Switzerland. ample, on climate, we used study coordinates Ninetieth-percentile GHG emissions of beef
*Corresponding author. Email: joseph.poore@queens.ox.ac.uk and spatial data sets to fill gaps. We recorded are 105 kg of CO2 eq per 100 g of protein, and
GHG Emissions Land Use Acid. Eutroph. Scty. Water land use (area multiplied by years occupied) is
(kg CO2eq) 10th
(m2year) 10th
(g SO2eq) (g PO43–eq) (kL eq) 370 m2·year. These values are 12 and 50 times
A 100g protein n 0 25 50 75 Pc Mean 0 100 200 300 Pc Mean 0 75 150 0 75 150 0 50 100 greater than 10th-percentile dairy beef impacts
Beef (beef herd) 724 20 50 42 164 (which we report separately given that its pro-
Lamb & Mutton 757 12 20 30 185 duction is tied to milk demand). Tenth-percentile
Beef (dairy herd) 490 9.1 17 7.3 22 GHG emissions and land use of dairy beef are
0 5 10 15 0 5 10 15 0 75 150 0 75 150 0 50 100
then 36 and 6 times greater than those of peas.
Crustaceans (farmed) 1.0k 5.4 18 0.4 2.0
High variation within and between protein-rich
Cheese 1.9k 5.1 11 4.4 41
products is also manifest in acidification, eutro-
Pig Meat 116 4.6 7.6 4.8 11
phication, and water use.
Fish (farmed) 612 2.5 6.0 0.4 3.7
Within the major crops wheat, maize, and rice,
Poultry Meat 326 2.4 5.7 3.8 7.1 90th-percentile impacts are more than three times
Eggs 100 2.6 4.2 4.0 5.7 greater than 10th-percentile impacts on all five
Tofu 354 1.0 2.0 1.1 2.2 indicators. Within major growing areas for these
Groundnuts 100 0.6 1.2 1.8 3.5 crops (the Australian wheat belt, the U.S. corn
Other Pulses 115 10th pctl. 0.5 0.8 4.6 7.3 belt, and the Yangtze river basin), land use be-
ruminant
Peas 438 meat 0.3 0.4 1.2 3.4 comes less variable, but we observe the same
Nuts 199 -2.2 0.3 2.7 7.9 high levels of variation in all other indicators.
Grains 23k 1.0 2.7 1.7 4.6 This variability, even among producers in similar
geographic regions, implies substantial potential
B 1 liter 0 2 4 6 0 3 6 9 0 15 30 0 10 20 0 50 100
to reduce environmental impacts and enhance
1.7 3.2 1.1 8.9
GHG Emissions (kg CO2eq 100g protein–1) GHG Emiss. 52% of U.S. grocery sales and 15% of global sales
(kg CO2eq L–1) (32). This sometimes means that standards
A Land Use Change
12.0
0.5 achieve market transformation (33), where vir-
2.1 2.1
Avg. veg. proteins tually all producers adhere to gain market access.
(excluding nuts) A fourth strategy for producers is setting en-
1.3 vironmental standards. These are particularly
1.0
0.20 important: Although many environmental issues
0.6 0.6
0.3
0.5 0.5 0.5 0.5 can be monitored and mitigated in a flexible
0.2 0.1 0.2
0.21
way, issues such as harmful pesticide usage and
0.02
-0.23
deforestation require strict controls, and issues
-0.2 such as on-farm biodiversity are hard to quantify
-0.76
10th pctl. (28). Procurement, farming organizations, and in-
B Crop Production & Feed Transport veg. proteins Avg. veg.
proteins ternational policy-makers must come together to
2.7 2.7
0.2 implement a safety net for global agriculture—
2.1 2.0 comprehensive standards to manage the worst
1.8 1.8
1.6 1.6 1.7
1.4 1.4 1.4 1.4 and hardest-to-quantify environmental issues,
0.1
1.1
0.8 0.11
extending the successes of existing schemes and
0.8
0.5 0.6
0.77 0.09 enabling a flexible mitigation approach to op-
0.54 erate effectively.
Farms Processors & Retailers Consumers or certified independently. In the United States
these tools have already been integrated with ex-
isting farm software (31); in Africa and South Asia
they are in trials with 2G mobile phones (37); and
in China they have been operated by extension
Validate and Monitor multiple Validate and services with extremely successful results (24).
Monitor multiple
communicate impacts using communicate Second, policy-makers would set targets on
impacts
impacts supply-chain data impacts
environmental indicators and incentivize them
Require
by providing producers with credit or tax breaks
Meet targets by sustainability Meet targets or by reallocating agricultural subsidies that now
choosing from standards exceed half a trillion dollars a year worldwide
multiple practice Policy
changes (38). Third, the assessment tools would provide
Set and incentivize mitigation targets Incentivize multiple mitigation and productivity enhancement
sustainable
Define and regulate sustainability standards consumption options to producers. Ideally these tools would be-
Researchers
come platforms that consolidate the vast amounts
Provide multiple mitigation options of research conducted by scientists around the
world, while also sharing producer best practices.
Fig. 4. Graphical representation of the mitigation framework. In particular, practice sharing offers a very effec-
tive way to engage producers (24). Maximum
flexibility also ensures least-cost mitigation (39)
animal and vegetable proteins will hold into the tons of CO2 from the atmosphere each year over and supports producer-led innovation (24).
21. E. Beza, J. V. Silva, L. Kooistra, P. Reidsma, . Eur. J. Agron. 82, 35. M. Springmann et al., Nat. Clim. Change 7, 69–74 (2016). on the manuscript and Tyana for illustrations. Funding: This work
206–222 (2017). 36. K. Denef, K. Paustian, S. Archibeque, S. Biggar, D. Pape, was unfunded. Author contributions: J.P. conducted the analysis
22. Z. Cui et al., Biogeosciences 11, 2287–2294 (2014). “Report of greenhouse gas accounting tools for agriculture and and wrote the manuscript. J.P. and T.N. contributed to the study
23. J. K. Ladha et al., Global Change Biol. 22, 1054–1074 forestry sectors” (Interim report to U.S. Department of design and data interpretation and reviewed the manuscript.
(2016). Agriculture under contract no. GS23F8182H, ICF International, Competing interests: The authors declare no competing interests.
24. Z. Cui et al., Nature 555, 363–366 (2018). 2012). Data and materials availability: A Microsoft Excel file allowing full
25. P. Smith et al., in Climate Change 2014: Mitigation of Climate 37. GSM Association (GSMA), “Creating scalable, engaging mobile replication of this analysis, containing all original and recalculated
Change (Cambridge Univ. Press, 2015), pp. 811–922. solutions for agriculture” (GSMA, 2017). data, has been deposited in the Oxford University Research Archive
26. E. Song, B. L. Nelson, J. Staum, SIAM/ASA J. Uncertain. 38. Organisation for Economic Co-operation and Development (doi.org/10.5287/bodleian:0z9MYbMyZ).
Quantif. 4, 1060–1083 (2016). (OECD), “Agriculture policy monitoring and evaluation 2017”
27. Q. Yu et al., Agric. Syst. 153, 212–220 (2017). (OECD, 2017).
28. R. N. German, C. E. Thompson, T. G. Benton, Biol. Rev. SUPPLEMENTARY MATERIALS
39. K. Segerson, Annu. Rev. Resour. Economics 5, 161–180
Cambridge Philos. Soc. 92, 716–738 (2017). (2013). www.sciencemag.org/content/360/6392/987/suppl/DC1
29. R. Lal, Global Chang. Biol. 10.1111/gcb.14054 (2018). 40. European Parliament and Council, “Establishing a common Materials and Methods
30. P. Smith et al., Agric. Ecosyst. Environ. 118, 6–28 Supplementary Text
organization of the markets in agricultural products”
(2007). Figs. S1 to S14
[Regulation (EU) 1308/2013, European Union, 2013].
31. K. B. Waldman, J. M. Kerr, Annu. Rev. Resour. Econ. 6, 429–449 Tables S1 to S17
(2014). References (41–151)
ACKN OWLED GMEN TS
32. Euromonitor; www.euromonitor.com Data S1 and S2
33. D. C. Nepstad, W. Boyd, C. M. Stickler, T. Bezerra, We thank the many researchers who provided us with additional
A. A. Azevedo, Philos. Trans. R. Soc. London Ser. B 368, data, acknowledged in data S1. We are grateful to R. Grenyer, 5 October 2017; resubmitted 8 December 2017
20120167 (2013). P. Smith, E. J. Milner-Gulland, C. Godfray, G. Gaillard, L. de Baan, Accepted 17 April 2018
34. A. Mottet et al., Global Food Sec. 14, 1–8 (2017). Y. Malhi, D. Thomas, K. Javanaud, and K. Afemikhe for comments 10.1126/science.aaq0216
SUPPLEMENTARY http://science.sciencemag.org/content/suppl/2018/05/30/360.6392.987.DC1
MATERIALS
RELATED file:/content
CONTENT
http://science.sciencemag.org/content/sci/363/6429/eaaw9908.full
REFERENCES This article cites 85 articles, 11 of which you can access for free
http://science.sciencemag.org/content/360/6392/987#BIBL
PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science is a registered trademark of AAAS.