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

Download details:

IP Address: 132.76.61.52

This content was downloaded on 04/12/2016 at 08:21

Please note that terms and conditions apply.

Energy and protein feed-to-food conversion efficiencies in the US and potential food security

gains from dietary changes

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

2016 Environ. Res. Lett. 11 105002

(http://iopscience.iop.org/1748-9326/11/10/105002)

Home Search Collections Journals About Contact us My IOPscience

You may also be interested in:

Diet changea solution to reduce water use?

M Jalava, M Kummu, M Porkka et al.

Local food in Iceland: identifying behavioral barriers to increased production and consumption

rhildur sk Halldrsdttir and Kimberly A Nicholas

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

The imprint of crop choice on global nutrient needs

Esteban G Jobbgy and Osvaldo E Sala

Redefining agricultural yields: from tonnes to people nourished per hectare

Emily S Cassidy, Paul C West, James S Gerber et al.

Historical trade-offs of livestocks environmental impacts

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

The role of diet in phosphorus demand

Genevive S Metson, Elena M Bennett and James J Elser

Embodied crop calories in animal products

Prajal Pradhan, Matthias K B Ldeke, Dominik E Reusser et al.

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.

http://iopscience.iop.org/page/termshttp://iopscience.iop.org/1748-9326/11/10http://iopscience.iop.org/1748-9326http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopsciencehttp://iopscience.iop.org/article/10.1088/1748-9326/9/7/074016http://iopscience.iop.org/article/10.1088/1748-9326/11/11/115004http://iopscience.iop.org/article/10.1088/1748-9326/10/11/114004http://iopscience.iop.org/article/10.1088/1748-9326/10/11/114004http://iopscience.iop.org/article/10.1088/1748-9326/9/8/084014http://iopscience.iop.org/article/10.1088/1748-9326/8/3/034015http://iopscience.iop.org/article/10.1088/1748-9326/10/12/125013http://iopscience.iop.org/article/10.1088/1748-9326/7/4/044043http://iopscience.iop.org/article/10.1088/1748-9326/8/4/044044http://iopscience.iop.org/article/10.1088/1748-9326/10/2/025002http://iopscience.iop.org/article/10.1088/1748-9326/10/2/025002

Environ. Res. Lett. 11 (2016) 105002 doi:10.1088/1748-9326/11/10/105002

LETTER

Energy and protein feed-to-food conversion efficiencies in the USand potential food security gains from dietary changes

AShepon1, G Eshel2, ENoor3 andRMilo1

1 Department of Plant and Environmental Sciences,Weizmann Institute of Science, Rehovot 7610001, Israel2 Radcliffe Institute for Advanced Study,HarvardUniversity, 10Garden Street, Cambridge,MA02138,USA3 Institute ofMolecular Systems Biology, ETHZrich, Auguste-Piccard-Hof 1, CH-8093 Zrich, Switzerland

E-mail: ron.milo@weizmann.ac.il

Keywords: livestock, food security, sustainability

Supplementarymaterial for this article is available online

AbstractFeeding a growing populationwhileminimizing environmental degradation is a global challengerequiring thoroughly rethinking food production and consumption. Dietary choices control foodavailability and natural resource demands. In particular, reducing or avoiding consumption of lowproduction efficiency animal-based products can spare resources that can then yieldmore food. Inquantifying the potential food gains of specific dietary shifts,most earlier research focused on calories,with less attention to other important nutrients, notably protein.Moreover, despite thewell-knownenvironmental burdens of livestock, only a handful of national level feed-to-food conversionefficiency estimates of dairy, beef, poultry, pork, and eggs exist. Yet such high level estimates areessential for reducing diet related environmental impacts and identifying optimal food gain paths.Herewe quantify caloric and protein conversion efficiencies forUS livestock categories.We then usethese efficiencies to calculate the food availability gains expected from replacing beef in theUS dietwith poultry, amore efficientmeat, and a plant-based alternative. Averaged over all categories, caloricand protein efficiencies are 7%8%.At 3% in bothmetrics, beef is by far the least efficient.We findthat reallocating the agricultural land used for beef feed to poultry feed production canmeet thecaloric and protein demands of120 and140million additional people consuming themeanAmerican diet, respectively, roughly 40%of currentUS population.

1. Introduction

The combination of ongoing population rise and the

increasing demand for animal-based products places a

severe strain on world natural resources (Smil 2002,Steinfeld et al 2006, Galloway et al 2007, Wirsenius

et al 2010, Bonhommeau et al 2013). Estimates suggestthat global meat demand would roughly double over

the period 20002050 (Pelletier and Tyedmers 2010,Alexandratos and Bruinsma 2012, Pradhan et al 2013,

Herrero et al 2015). Earlier analyses (Steinfeldet al 2006, Godfray et al 2010, Foley et al 2011, Herrero

et al 2015) of food supply chains identified inefficiencyhotspots that lend themselves to such mitigation

measures as improving yield (through genetics andagricultural practices), increasing energy, nutrient and

water use efficiencies, or eliminating waste. Others

focused on the environmental performance of specific

products, for example animal-derived (deVries and deBoer 2010, Pelletier et al 2010, 2011, 2014, Thoma

et al 2013). A complementary body of work (Pimenteland Pimentel 2003, Eshel and Martin 2006, Eshel

et al 2010, Hedenus et al 2014, Tilman and Clark 2014,

Springmann et al 2016) quantifies the environmentalperformance of food consumption and dietary pat-

terns, highlighting the large environmental impacts

dietary choices can have.Key to estimating expected outcomes of potential

dietary shifts is quantifying the amount of extra foodthat would become available by reallocating resourcescurrently used for feed production to producinghuman food (Godfray et al 2010, Foley et al 2011,

OPEN ACCESS

RECEIVED

8 February 2016

REVISED

8 September 2016

ACCEPTED FOR PUBLICATION

12 September 2016

PUBLISHED

4October 2016

Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

2016 IOPPublishing Ltd

http://dx.doi.org/10.1088/1748-9326/11/10/105002mailto:ron.milo@weizmann.ac.ilhttp://dx.doi.org/10.1088/1748-9326/11/10/105002http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/11/10/105002&domain=pdf&date_stamp=2016-10-04http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/11/10/105002&domain=pdf&date_stamp=2016-10-04http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0

Cassidy et al 2013, Pradhan et al 2013, West et al 2014,Peters et al 2016). One notable effort (Foley et al 2011,Cassidy et al 2013,) suggested that global reallocationto direct human consumption of both feed and biofuelcrops can sustain four billion additional people. Yet,most cultivated feed (corn, hay, silage) is human ined-ible and characterized by yields well above those ofhuman edible crops. Moreover, most previous effortsfocused on calories (Cassidy et al 2013, Pradhanet al 2013), while other key dimensions of human dietsuch as protein adequacy are equally important.

Here we quantify efficiencies of caloric and proteinfluxes in US livestock production. We answer suchquestions as: How much feed must enter the livestockproduction stream to obtain a set amount of edibleend product calories?What is the composition of thesefeed calories in the current US system? Where alongthe production stream do most losses occur? We pro-vide the analysis in terms of both protein and caloriesand use them to explore the food availability impactsof a dietary change within the animal portion (exclud-ing fish) of the American food system using the dietaryshift potential method as described below. While diet-ary changes entail changes in resource allocation andemissions (Hedenus et al 2014, Tilman and Clark2014, Eshel et al 2016, Springmann et al 2016), here wehighlight the food availability gains that can be realizedby substituting the least efficient food item, beef, with

the most efficient nutritionally similar food item,poultry. Because beef and poultry are the least andmost efficient livestock derivedmeats respectively, thissubstitution marks the upper bound on food gainsachievable by any dietary change within the meat por-tion of the mean American diet (MAD). In this studywe focus on substitution of these individual items, andplan to explore the substitution of full diets elsewhere.As a yardstick with which to compare our results, wealso present the potential food availability gains asso-ciated with replacing beef with a fully plant-basedalternative.

2.Methods and data

The parameters used in calculating the caloric andprotein Sankey flow diagrams (figures 1 and 2) arebased on Eshel et al (2015, 2014) and references andsources therein. Feed composition used in figures 1and 2 are derived from NRC data (National ResearchCouncil 1982, 2000). For this work, the MAD is theactual diet of the average American over 20002010(United States Department of Agriculture ERS 2015),with approximate daily loss-adjusted consumption of2500 kcal and 70 g protein per capita (see SI andsupplementary data for additional details).

Figure 1.ASankey flowdiagramof theUS feed-to-food caloricflux from the three feed classes (left) into edible animal products(right). On the right, parenthetical percentages are the food-out/feed-in caloric conversion efficiencies of individual livestockcategories. Caloric values are in Pcal, 1012 kcal. Overall, 1187 Pcal of feed are converted into 83 Pcal edible animal products, reflectingaweightedmean conversion efficiency of approximately 7%.

2

Environ. Res. Lett. 11 (2016) 105002

2.1. Calculating the dietary shift potentialThe dietary shift potential, the number of additionalpeople that can be sustained on a given croplandacreage as part of a dietary shift, is

D =-

- -

( )( )

( )P P l ll l l

1a bUS a b

per capita land gain

MAD a b

updated per capita

MAD land requirement

In equation (1), the left-hand side (Pab) is thenumber of additional people that can be fed onland spared by the replacement of food item awith food item b. PUS300 million denotes the20002010 mean US population; la and lb denotesthe annual per capita land area for producing a setnumber of calories of foods a and b. This definitionreadily generalizes to protein based replacements,and/or to substitution of whole diets rather than spe-cific food items.

To derive the mean per capita land requirement ofthe MAD, l ,MAD we calculate the land needs of each ofthe non-negligible plant and animal based itemsthe MAD comprises. We convert a given per capitaplant item mass to the needed land by dividing theconsumed item mass by its corresponding nationalmean loss adjusted yield. The land needs of thefull MAD is simply the sum of these needs over all

items (see supplementary data). The per capita cropland requirements of the animal based MAD cate-gories (e.g., l ,poultry )lbeef are based on Eshel et al(2014, 2015).

Themodest land needs of poultrymean that repla-cing beef with an amount of poultry that is caloric- orprotein-equivalent spares land that can sustain addi-tional people on a MAD. We denote by citem the kcal(person yr)1 consumption of any MAD item. The setnumber of calories (or protein) consumed in theMAD is different for beef and thus for the calculationof substituting beef with poultry, we multiply the percapita land area of poultry by /c c ,beef poultry the percapita caloric (or protein) beef:poultry consumptionratio in the MAD, which is 1.2 for calories and 0.6 forprotein.

Using equation (1), the caloric dietary shift poten-tial of beef is

D =

-

- -

P

P l lc

c

l l lc

c

.beef poultry

US beef poultrybeef

poultry

MAD beef poultrybeef

poultry

For the beef replacement calculation, the resultantpost-replacement calories (light orange arrows infigure 3(a)) comprise (1) the poultry calories thatreplace the MAD beef calories, plus (2) calories that

Figure 2.TheUS feed-to-food proteinflux from the three feed classes (left) into edible animal products (right). On the right,parenthetical percentages are the food-protein-out/feed-protein-in conversion efficiencies of individual livestock categories. Proteinvalues are inMt (109 kg). Overall, 63Mt of feed protein yield edible animal products containing 4.7Mt protein, an 8%weightedmeanprotein conversion efficiency.

3

Environ. Res. Lett. 11 (2016) 105002

the spared lands can yield if allocated to the productionof MAD-like diet for additional people (national feedland supporting beef minus the land needed toproduce the replacement poultry). The MAD caloriesthat the spared land can sustain is calculated bymultiplying the spared land area by the mean caloricyield of the full MAD with poultry replacing beef,1700Mcal (ac yr)1. The national annual caloriesdue to substituting beef for poultry is

= +

-

- -

( )

C P c c

P l lc

c

l l lc

c

365 365

2

beef poultrynat.

US beef MAD

US beef poultrybeef

poultry

MAD beef poultrybeef

poultry

where c and l are the per capita daily caloric consump-tion and annual land requirements of poultry, beef orthe full MAD, respectively. The first and second termson the right-hand side of equation (2) are terms (1)and (2) of the above explanation, respectively.

To derive the difference between the above repla-cement calories and the replaced beef calories (percen-tages in figure 3), we subtract the original nationalconsumed beef calories P c365 US beef from the aboveequation. The difference between replacement andreplaced caloric fluxes is

-

=

-

- -

( )

C C

c

P l lc

c

l l lc

c

365 3

beef poultrynat.

beefnat.

MAD

US beef poultrybeef

poultry

MAD beef poultrybeef

poultry

As noted above, the quotient on the right-hand sidegives the number of extra people that can be fed,reported in figure 3. An analogous calculation repla-cing calories with protein mass, yields the proteindietary shift potential shown in figure 3(b). Thecurrent calculation of the dietary shift potential alsoenables calculating the food availability gains asso-ciated with any partial replacement. Figure S2 depictsthe relation between the dietary shift potential (addi-tional people that can be fed a full MAD diet) and thepercentage of national beef calories (from MAD)replacedwith poultry.

2.2. The choice of poultry as the consideredsubstituteWe use poultry as the replacement food in our foodavailability calculations for several reasons. First, USpoultry consumption has been rising in recent decadesoften substituting for beef (Daniel et al 2011), suggest-ing it can serve as a plausible replacement. In addition,poultry incur the least environmental burden amongthe major meat categories and thus the calculation of

Figure 3.Dietary shift potential of substituting beef with poultry in themeanAmerican diet (MAD). Percent change in availablecalories due to substituting beef with poultry (panel (a)),+520%. The number of additional people consuming 2500 kcal d1 thatthese calories can sustain (the dietary shift potential) is 116million (upper right parenthetical value). Caloric values are in Pcal, i.e.1012 kcal. The protein gain due to dietary shift frombeef to poultrywill increase by 380% (panel (b)), meeting the protein needs of 142million additional people consuming 70 g protein d1 (as in theMAD). Protein values are inMt (109 kg). The caloric loss followingsubstitution is calculated based on the conversion efficiency for poultry and theMAD. The loss of the plant-based portion ofMAD iscalculated by assessing the loss of each individual plant item throughout the supply chain (see supplementary data); the loss of theanimal-based portion ofMAD is based on the caloric efficiency conversion estimates shown in table 1. A similar calculation isperformed for protein.

4

Environ. Res. Lett. 11 (2016) 105002

the dietary shift potential presented here serves as anupper bound on possible food gains achievable by anysubstitutionwithin themeat portion of theMAD.

Plant-based diets can also serve as a viable replace-ment for animal products, and confer larger meanenvironmental (Eshel et al 2014, 2016) and food avail-ability gains (Godfray et al 2010). Recognizing that themajority of the populationwill not easily become exclu-sive plant eaters, herewe choose to present the less radi-cal and perhapsmore practical scenario of replacing theenvironmentally most costly beef with the moreresource efficient poultry. We also augment this calcul-ationwith a plant-based alternative diet as a substitute.

Finally, poultry stands out in its high kcal g1 and gprotein g1 values and its desirable nutritional profile.Per calorie, it can deliver more protein than beef whiledelivering as much or more of the other essentialmicronutrients (figure S1).While it is tricky to comparethe protein quality of beef and poultry, we can use thebiological value (modified essential amino acid indexand chemical score index Ihekoronye 1988) and theprotein digestible corrected amino acid score, the pro-tein indicator of choice of the FAO. Within inevitablevariability, the protein quality of poultry is similar tothat of beef using both metrics (Sarwar 1987, Ihekor-onye 1988, Lpez et al 2006, Barrn-Hoyos et al 2013).While the FAOhas recently introduced anupdated pro-tein quality score (DIAASdigestible indispensableamino acid score) (FAO Food and Nutrition paper No.92 2011), to our knowledge no reliable DIAAS datacomparing beef andpoultry exists.

3. Results

The efficiency and performance of the animal portionof the American food system is presented in table 1(see detailed calculations in supplementary files),highlighting a dichotomy between beef and the otheranimal categories, consistent with earlier environmen-tal burden estimates (Eshel et al 2014).

The calories flow within the US from feed to live-stock to human food is presented in figure 1. From leftto right are primary inputs (concentrated feed, pro-cessed roughage and pasture) feeding the five second-ary producer livestock categories, transformed intohuman consumed calories. We report energy fluxes inPcal=1012 kcal, roughly the annual caloric needs of amillion persons. Annually, 1200 Pcals of feed fromall sources (or 800 Pcals when pasture and bypro-ducts are excluded) become 83 Pcals of loss adjustedanimal based human food. This is about 7% overallcaloric conversion efficiency. The overall efficiencyvalue arises from weighting the widely varied categoryspecific efficiencies, from 3% for beef to 17% for eggsand dairy, by the average US consumption (rightmostpart of figure 1). Concentrate feed consumption, suchas maize, is distributed among pork, poultry, beef anddairy, while processed roughage and pasture (50% oftotal calories) feed almost exclusively beef. The con-centrated feed category depicted in figure 1 alsoincludes byproducts. We note that because detailedinformation on the distribution of byproducts as feedfor the different animal categories is lacking, we can-not remove them from the feed to food efficiency calc-ulation. Yet, our analysis shows that for the years20002010 the contribution of byproducts to the totalfeed calories (and protein) was less than 10% (see SIspreadsheet) and so their effect on the values is quanti-tatively small. The results reported in all figures arecorrected for import-export imbalances, such that thepresented values refer to the feed used to produce theanimal-derived food domestically consumed in theUS(i.e., excluding feed used for livestock to be exported,and including imported feed, albeit quite minor in theUS context).

While calories are widely used to quantify foodsystem performance, proteinwhich is often invokedas the key nutritional asset of meatoffers an impor-tant complementary dimension (Tessari et al 2016).The flow of protein in the American livestock produc-tion system, which supplies 45 g protein person1

Table 1.Key parameters (std. dev.)used in evaluatingUS feed allocation and conversion among animal categories (Eshelet al 2015) and energy (caloric) and protein efficiency.

Parameter Units Beef Poultry Pork Dairy Eggs

Feed intake per LW kg/kg LW 144 1.90.4 3.11.3 N/A N/AFeed intake per EW kg/kg EW 3613 4.20.8 62.5 N/A N/AFeed intake perCW kg/kgCW 499 5.41.4 94 2.60.6 2.41.2Feed caloric content kcal g1 2.30.6 3.41.4 3.62 2.80.9 3.42.4Food caloric content kcal g1 3.20.3 2.30.1 2.80.2 1.20.1 1.40.1Caloric conversion efficiency % 2.90.7 134 94 174 179Feed protein content % 123 177 1711 155 1712Food protein content % 152 202 141.4 60.6 131.3Protein conversion efficiency % 2.50.6 217 94.5 144 3116

Note: LW=live weight (USDA reported slaughter live weight); EW=edible weight (USDA reported retail boneless edibleweight); CW=consumed weight (USDA reported loss-adjusted weight). N/A, denotes not applicable as the parameter isrelevant only for CW. Feed caloric content refers to metabolizable energy and feed protein content refers to crude protein. For

further information on all data sources and calculations see SI and supplementary data.

5

Environ. Res. Lett. 11 (2016) 105002

d1 to theMAD, is shown in figure 2. Overall, 63Mt (1Mt=109 kg) feed protein per year are converted byUS livestock into 4.7Mt of loss-adjusted edible animalbased protein. This represents an overall weighted-mean feed-to-food protein conversion efficiency of8% for the livestock sector. Protein conversion effi-ciencies by individual livestock categories span an11-fold range, more than twice the correspondingrange for calories, from 31% for eggs to 3% for beef(see SI formore details).

By isolating visually and numerically the contribu-tions from pasture, which are derived from land that isunfit for production of most other foods, figures 1 and2 quantify expected impacts of dietary shifts. Of those,we choose to focus on substituting beef with poultry.Because these are themost and least resource intensivemeats, this substitution constitutes an upper boundestimate on food gains achievable by any meat-to-meat shift. Lending further support to the beef-to-poultry substitution choice, poultry is relatively nutri-tionally desirable (see the methods section and figureS1), andjudging by its ubiquity in the MADpala-table tomanyAmericans.

We quantify the dietary shift potential (a term wefavor over the earlier diet gap Foley et al 2011), thenumber of additional people a given cropland acreagecan sustain if differently reallocated as part of a dietaryshift.While here we estimate the dietary shift potentialof the beef-to-poultry substitution, the methodologygeneralizes to any substitution (see methods sectionfor further information and equations). The beef-to-poultry dietary shift potentials are premised on reallo-cating the cropland acreage currently used for produ-cing feed for US beef (excluding pastureland) toproducing feed for additional poultry production.Subtracting from beefs high quality land require-ments those of poultry gives the spared land thatbecomes available for feeding additional people.Dividing this spared acreage by the per capita landrequirements of the MAD diet (modifying the latterfor the considered substitution) yields the numberof additional people sustained by the dietarysubstitution.

We calculate the dietary shift potential for beef (asdefined above and in the methods section) by quanti-fying the land needed for producing calorie- and pro-tein-equivalent poultry substitution, and theirdifferences from the land beef currently requires. Wederive the number of additional people this land cansustain by dividing the areal difference thus found bythe per capita land demands of the whole modifiedMAD,0.5 acres (2103 m2) per year.

Evaluating this substitution, and taking note of fullsupply chain losses, we obtain the overall dietary shiftpotential of beef to poultry on a caloric basis to be120 million people (40% of current US popula-tion; figure 3, panel (a)). That is, if the (non-pasture)land that yields the feed US beef currently consumewas used for producing feed for poultry instead, and

the added poultry production was chosen so as to yieldexactly the number of calories the replaced beef cur-rently delivers, a certain acreage would be spared,because of poultrys lower land requirements. If, inaddition, that spared land was used for growing a vari-ety of products with the same relative abundance as inthe full MAD (but with poultry replacing beef), theresultant human edible calories would have risen to sixtimes the replaced beef calories (figure 3, panel (a)).For protein-conserving dietary shift (figure 3, panel(b)), the dietary shift potential is estimated at 140million additional people (consuming 70 g proteinperson1 d1 as in the full MAD). As the protein qual-ity of poultry and beef are similar (see the methodssection and references therein), this substitutionentails no protein quality sacrifices.

As a benchmark with which to compare the beef topoultry results, we next consider the substitution ofbeef with a plant based alternative based on themetho-dology developed in Eshel et al (2016). In that study,we derive plant based calorie- and protein-conservingbeef-replacements. We consider combinations of 65leading plant items consumed by the average Amer-ican thatminimize land requirements with themass ofeach plant item set to 15 g d1 to ensure dietarydiversity.We find that these legume-dominated plant-based diets substitute beef with a dietary shift potentialof190million individuals.

4.Discussion

In this study we quantify the caloric and proteincascade through the US livestock system from feed toconsumed human food. Overall,

sustained individuals, respectively. This potentialproduction increase can serve as food collateral in faceof uncertain food supply (e.g. climate change), orexported to where food supply is limited. In the case ofenvisioning various scenarios resulting in only partialsubstitution to poultry consumption, the currentcalculation also enables to deduce the food gainsassociated with substituting only a certain percentageof national beef calories with poultry (see figure S2).Our purpose here is not to endorse poultry consump-tion, nor can our results be construed as such. Rather,the results simply illustrate the significant food avail-ability gains associated with the rather modest andtractable dietary shift of substituting beef with lessinefficient animal based alternatives. Substitution ofother food items with other nutritionally similaranimal food items is also plausible (e.g., pork for beef),yet the food gains expected from such replacementsare considerably lower (see supplementary data).Substitution of beef with non-meat animal basedproducts (dairy and eggs) is possible on a caloric orprotein basis (see supplementary data), yet given theirdissimilar nutritional profile, amore elaborate metho-dology is required to construct and analyze such a shift(Tessari et al 2016). The dietary shift potential ofreplacing beef with a plant based alternative (domi-nated by legumes) (Eshel et al 2016) amounts to190million additional people. Thus while plant basedalternatives offer the largest food availability gains,poultry is not far behind.

We note that the substitution of beef for eitherpoultry or plants also entails vast reductions indemand for pastureland. The effects of dietary shiftson demand for agricultural inputs (such as fertilizer orwater) for the production of food on the land sparedfrom growing feed for beef requires furtherinvestigation.

This paper offers a system wide view of feed tofood production in the US, and introduces the dietaryshift potential as a method for quantifying possiblefood availability gains various dietary shifts confer.Building on this work, future work can quantify thedietary shift potential of full diets (e.g. Peterset al 2016), enhance the realism of various considereddietary shifts, and better integrate nutritional con-siderations, micronutrients in particular, in the assess-ment of expected outcomes.

Acknowledgments

We thank David Canty, David St-Jules, Avi Flamholz,Avi Levy, Tamar Makov, and Lisa Sasson for theirimportant help with this manuscript. This researchwas funded by the European Research Council (Pro-ject NOVCARBFIX 646827). RM is the Charles andLouiseGartner professional chair.

References

AlexandratosN andBruinsma J 2012World agriculture towards2030/2050: the 2012 revision ESAWorking paper 12-03FAO, Rome

Barrn-Hoyos JM,Archuleta AR, Falcn-VillaM, del R,Canett-Romero R,Cinco-Moroyoqui F J,Romero-Barancini A L andRueda-Puente EO2013 Proteinquality evaluation of animal food proteins by in-vitromethodologies FoodNutr. Sci. 04 37684

Bonhommeau S,Dubroca L, Le PapeO, Barde J, KaplanDM,Chassot E andNieblas A-E 2013 Eating up theworlds foodweb and the human trophic level Proc. Natl Acad. Sci. USA110 2061720

Cassidy E S,West PC,Gerber J S and Foley J A 2013Redefiningagricultural yields: from tonnes to people nourished perhectareEnviron. Res. Lett. 8 034015

Daniel CR,Cross A J, KoebnickC and Sinha R 2011Trends inmeatconsumption in theUSAPublic HealthNutr. 14 57583

deVriesM and de Boer I 2010Comparing environmental impactsfor livestock products: a review of life cycle assessments Livest.Sci. 128 111

Eshel G andMartin PA 2006Diet, energy, and global warming EarthInteract. 10 117

Eshel G,Martin PA andBowenEE 2010 Land use and reactivenitrogen discharge: effects of dietary choices Earth Interact. 14115

Eshel G, SheponA,MakovT andMilo R 2014 Land, irrigationwater, greenhouse gas and reactive nitrogen burdens ofmeat,eggs &dairy production in theUnited States Proc. Natl Acad.Sci. USA 111 119962001

Eshel G, SheponA,MakovT andMilo R 2015 PartitioningUnitedStates feed consumption among livestock categories forimproved environmental cost assessments J. Agric. Sci. 15343245

Eshel G, SheponA,Noor E andMilo R 2016 Environmentallyoptimal, nutritionally aware beef replacement plant baseddiets. In pressEnviron. Sci. Technol. 50 81648

FAOFood andNutrition paperNo. 92 2011Dietary protein qualityevaluation in human nutritionReport of an FAOExpertConsultation (Auckland,NewZealand) (http://fao.org/ag/humannutrition/35978-02317b979a686a57aa4593304ffc17f06.pdf)

Foley J A et al 2011 Solutions for a cultivated planetNature 47833742

Galloway JN et al 2007 International trade inmeat: the tip of thepork chopAmbio 36 6229

GodfrayHC J, Beddington J R, Crute I R,Haddad L, LawrenceD,Muir J F, Pretty J, Robinson S, Thomas SMandToulminC2010 Food security: the challenge of feeding 9 billion peopleScience 327 8128

Hedenus F,Wirsenius S and JohanssonD JA 2014The importanceof reducedmeat and dairy consumption formeetingstringent climate change targetsClim. Change 124 7991

HerreroM,Wirsenius S,Henderson B, Rigolot C, Thornton P,Havlk P, de Boer I andGerber P 2015 Livestock and theenvironment:What havewe learned in the past decade?Annu. Rev. Environ. Resour. 40 177202

IhekoronyeA I 1988 Estimation of the biological value offood proteins by amodified equation of the essentialamino acid index and the chemical score Food/Nahrung 327838

LpezMMS,Kizlansky A and Lpez L B 2006 Evaluacin de lacalidad de las protenas en los alimentos calculando el escorede aminocidos corregido por digestibilidadNutr. Hosp. 214751

National ResearchCouncil 1982United States-Canadian Tables ofFeedComposition: Nutritional Data forUnited States andCanadian Feeds, Third Revision (WashingtonDC: TheNational Academies)

National ResearchCouncil 2000Nutrient Requirements of BeefCattle: Seventh Revised Edition: Update 2000 (Washington,DC: TheNational Academies)

7

Environ. Res. Lett. 11 (2016) 105002

http://dx.doi.org/10.4236/fns.2013.44048http://dx.doi.org/10.4236/fns.2013.44048http://dx.doi.org/10.4236/fns.2013.44048http://dx.doi.org/10.1073/pnas.1305827110http://dx.doi.org/10.1073/pnas.1305827110http://dx.doi.org/10.1073/pnas.1305827110http://dx.doi.org/10.1088/1748-9326/8/3/034015http://dx.doi.org/10.1017/S1368980010002077http://dx.doi.org/10.1017/S1368980010002077http://dx.doi.org/10.1017/S1368980010002077http://dx.doi.org/10.1016/j.livsci.2009.11.007http://dx.doi.org/10.1016/j.livsci.2009.11.007http://dx.doi.org/10.1016/j.livsci.2009.11.007http://dx.doi.org/10.1175/EI167.1http://dx.doi.org/10.1175/EI167.1http://dx.doi.org/10.1175/EI167.1http://dx.doi.org/10.1175/2010EI321.1http://dx.doi.org/10.1175/2010EI321.1http://dx.doi.org/10.1175/2010EI321.1http://dx.doi.org/10.1175/2010EI321.1http://dx.doi.org/10.1073/pnas.1402183111http://dx.doi.org/10.1073/pnas.1402183111http://dx.doi.org/10.1073/pnas.1402183111http://dx.doi.org/10.1017/S0021859614000690http://dx.doi.org/10.1017/S0021859614000690http://dx.doi.org/10.1017/S0021859614000690http://dx.doi.org/10.1017/S0021859614000690http://dx.doi.org/10.1021/acs.est.6b01006http://dx.doi.org/10.1021/acs.est.6b01006http://dx.doi.org/10.1021/acs.est.6b01006http://fao.org/ag/humannutrition/35978-02317b979a686a57aa4593304ffc17f06.pdfhttp://fao.org/ag/humannutrition/35978-02317b979a686a57aa4593304ffc17f06.pdfhttp://fao.org/ag/humannutrition/35978-02317b979a686a57aa4593304ffc17f06.pdfhttp://dx.doi.org/10.1038/nature10452http://dx.doi.org/10.1038/nature10452http://dx.doi.org/10.1038/nature10452http://dx.doi.org/10.1038/nature10452http://dx.doi.org/10.1579/0044-7447(2007)36[622:ITIMTT]2.0.CO;2http://dx.doi.org/10.1579/0044-7447(2007)36[622:ITIMTT]2.0.CO;2http://dx.doi.org/10.1579/0044-7447(2007)36[622:ITIMTT]2.0.CO;2http://dx.doi.org/10.1126/science.1185383http://dx.doi.org/10.1126/science.1185383http://dx.doi.org/10.1126/science.1185383http://dx.doi.org/10.1007/s10584-014-1104-5http://dx.doi.org/10.1007/s10584-014-1104-5http://dx.doi.org/10.1007/s10584-014-1104-5http://dx.doi.org/10.1146/annurev-environ-031113-093503http://dx.doi.org/10.1146/annurev-environ-031113-093503http://dx.doi.org/10.1146/annurev-environ-031113-093503http://dx.doi.org/10.1002/food.19880320818http://dx.doi.org/10.1002/food.19880320818http://dx.doi.org/10.1002/food.19880320818http://dx.doi.org/10.1002/food.19880320818

PelletierN, IbarburuMandXinH2014Comparison of theenvironmental footprint of the egg industry in theUnitedStates in 1960 and 2010Poult. Sci. 93 24155

PelletierN, Lammers P, StenderD and Pirog R 2011 Life cycleassessment of high- and low-profitability commodity anddeep-bedded niche swine production systems in theUpperMidwesternUnited StatesAgric. Syst. 103 599608

PelletierN, Pirog R andRasmussenR 2010Comparative life cycleenvironmental impacts of three beef production strategies intheUpperMidwesternUnited StatesAgric. Syst. 103 3809

PelletierN andTyedmers P 2010 Forecasting potential globalenvironmental costs of livestock production 20002050Proc.Natl Acad. Sci. USA 107 183714

Peters C J, Picardy J, Darrouzet-Nardi A F,Wilkins J L,GriffinT S and FickGW2016Carrying capacity of USagricultural land: ten diet scenarios Elem. Sci. Anthr. 4 000116

Pimentel D andPimentelM2003 Sustainability ofmeat-based andplant-based diets and the environmentAm. J. Clin. Nutr. 78660S3S

PradhanP,LdekeMKB,ReusserDE andKropp J P 2013Embodiedcrop calories in animal productsEnviron. Res. Lett.8 44044

SarwarG 1987Digestibility of protein and bioavailability of aminoacids in foods, effects on protein quality assessmentWorldRev. Nutr. Diet. 54 2670

Smil V 2002Nitrogen and food production: proteins for humandietsAmbio 31 12631

SpringmannM,GodfrayHC J, RaynerMand Scarborough P 2016Analysis and valuation of the health and climate changecobenefits of dietary change Proc. Natl Acad. Sci. 113414651

SteinfeldH,Gerber P,Wassenaar T, Castel V, RosalesM andDeHaanC 2006 Livestocks Long Shadow: EnvironmentalIssues andOptions (Rome: Food andAgricultureOrganization of theUnitedNations)

Tessari P, Lante A andMoscaG2016 Essential amino acids:masterregulators of nutrition and environmental footprint? Sci. Rep.6 26074

ThomaG et al 2013Greenhouse gas emissions frommilkproduction in theUnited States: a cradle-to grave life cycleassessment circa 2008 Int. Dairy J. 31 S314

TilmanDandClarkM2014Global diets link environmentalsustainability and human healthNature 515 51822

United States Department of Agriculture ERS 2015 Loss-AdjustedPer Capita FoodAvailabilityData (http://ers.usda.gov/data-products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-documentation.aspx)

West PC et al 2014 Leverage points for improving global foodsecurity and the environment Science 345 3258

Wirsenius S, Azar C andBerndes G 2010Howmuch land is neededfor global food production under scenarios of dietary changesand livestock productivity increases in 2030?Agric. Syst. 10362138

8

Environ. Res. Lett. 11 (2016) 105002

http://dx.doi.org/10.3382/ps.2013-03390http://dx.doi.org/10.3382/ps.2013-03390http://dx.doi.org/10.3382/ps.2013-03390http://dx.doi.org/10.1016/j.agsy.2010.07.001http://dx.doi.org/10.1016/j.agsy.2010.07.001http://dx.doi.org/10.1016/j.agsy.2010.07.001http://dx.doi.org/10.1016/j.agsy.2010.03.009http://dx.doi.org/10.1016/j.agsy.2010.03.009http://dx.doi.org/10.1016/j.agsy.2010.03.009http://dx.doi.org/10.1073/pnas.1004659107http://dx.doi.org/10.1073/pnas.1004659107http://dx.doi.org/10.1073/pnas.1004659107http://dx.doi.org/10.12952/journal.elementa.000116http://dx.doi.org/10.1088/1748-9326/8/4/044044http://dx.doi.org/10.1159/000415302http://dx.doi.org/10.1159/000415302http://dx.doi.org/10.1159/000415302http://dx.doi.org/10.1579/0044-7447-31.2.126http://dx.doi.org/10.1579/0044-7447-31.2.126http://dx.doi.org/10.1579/0044-7447-31.2.126http://dx.doi.org/10.1073/pnas.1523119113http://dx.doi.org/10.1073/pnas.1523119113http://dx.doi.org/10.1073/pnas.1523119113http://dx.doi.org/10.1073/pnas.1523119113http://dx.doi.org/10.1038/srep26074http://dx.doi.org/10.1016/j.idairyj.2012.08.013http://dx.doi.org/10.1016/j.idairyj.2012.08.013http://dx.doi.org/10.1016/j.idairyj.2012.08.013http://dx.doi.org/10.1038/nature13959http://dx.doi.org/10.1038/nature13959http://dx.doi.org/10.1038/nature13959http://ers.usda.gov/data-products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-documentation.aspxhttp://ers.usda.gov/data-products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-documentation.aspxhttp://ers.usda.gov/data-products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-documentation.aspxhttp://dx.doi.org/10.1126/science.1246067http://dx.doi.org/10.1126/science.1246067http://dx.doi.org/10.1126/science.1246067http://dx.doi.org/10.1016/j.agsy.2010.07.005http://dx.doi.org/10.1016/j.agsy.2010.07.005http://dx.doi.org/10.1016/j.agsy.2010.07.005http://dx.doi.org/10.1016/j.agsy.2010.07.005

1. Introduction2. Methods and data2.1. Calculating the dietary shift potential2.2. The choice of poultry as the considered substitute

3. Results4. DiscussionAcknowledgmentsReferences