Human food vs. animal feed debate. A thorough analysis of environmental footprints

Land Use Policy 67 (2017) 652–659

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Human food vs. animal feed debate. A thorough analysis of environmental footprints

MARK

⁎

Arianna Di Paolaa, , Maria Cristina Rullib, Monia Santinia,c a b c

Euro-Mediterranean Center on Climate Change (CMCC) Foundation, Impacts on Agriculture, Forest and Ecosystem Services (IAFES), Viterbo, Italy Department of Civil and Environmental Engineering, Politecnico di Milano, Milano, Italy Far East Federal University (FEFU), Ajax St., Vladivostok, Russky Island, Russia

A R T I C L E I N F O

A B S T R A C T

Keywords: Human vs. animal food debate Environmental footprints Animal-source proteins Plant-source proteins Feed grains Livestock impacts

Currently, a large portion of grain production is funneled into animal feed despite widespread hunger and undernutrition. In the present work we: (i) estimated the area, water and carbon footprints of animal-source proteins (AP) obtained from intensive farming systems and compared them with those from producing an equivalent amount of plant-source proteins (PP); (ii) postulated a set of straightforward hypotheses to recover environmental resources by cutting down a surplus in the per capita protein intake from three representative regions where intensive animal farming systems account for a great share of animal food production. Our major findings revealed that AP from intensive farming were approximately 2.4 to 33 more expensive in terms of area and water demand and 2.4 to 240 more pollutant in terms of greenhouse gas emissions when compared with PP. Environmental recoveries varied widely according to the hypothesized scenarios, but even the lowest estimates suggested remarkable results. Whether additional proteins supply would be required, crops with large protein content as peas, chickpeas, soybeans, and lupins could help to meet food security, while better compromise between dietary habits and environmental protection could be reached in rich countries by a moderate consumption of meat produced with non-feed grain systems.

1. Introduction Proteins are essential constituents of human nutrition (Day, 2013; World Food Programme (WFP), 2012) and one of the major benefits of consuming animal-source food lies in the protein content they eventually incorporate (Smil, 2014; Sanders, 1999). Currently, an excessive vs. inadequate proteins intake in rich vs. malnourished countries yields to an imbalance in the world's food system (Smil, 2002a), and for the coming decades, emerging countries are expected to increase the demand for animal-source food (Ehui et al., 1998; Kastner et al., 2012; Ranganathan et al., 2016; Tilman and Clark, 2014). The production of animal-source proteins (AP) is highly inefficient (Smil, 2002a; Aiking, 2011): about 75% of global human-driven inputs of reactive nitrogen are used for agriculture and only 30% of these inputs are converted into plant-source proteins (PP) to feed livestock, with the non-recovered fraction mainly lost to the environment, causing degradation and pollution. Then, about 15% of PP from feed crops are estimated to turn into AP for human consumption, while 85% are wasted (Smil, 2002a; Aiking, 2011). From the fifties, industrialized countries intensified the efficiency of ⁎

Corresponding author. E-mail address: [email protected] (A. Di Paola).

http://dx.doi.org/10.1016/j.landusepol.2017.06.017 Received 28 March 2017; Received in revised form 19 June 2017; Accepted 21 June 2017 0264-8377/ © 2017 Elsevier Ltd. All rights reserved.

animal-source food production shifting from extensive, grazing, smallscale, subsistence animal farming systems towards intensive, largescale, geographically-concentrated, specialized production units (Robinson et al., 2011; Delgado et al., 2001), broadly known as Intensive Animal Farming Systems (IAFSs). Even if IAFSs are mostly located in industrialized, large and quickly emerging countries, they represent a global issue accounting for more than 40% of the global animal-source food (FAO et al., 2014), being a relevant source of greenhouse gas (GHG) emissions (Tubiello et al., 2013) and sharing a significant portion of the global crop production to feed livestock (Foley et al., 2011; Smil, 2001; Steinfeld et al., 2006; Manceron et al., 2014) that, paradoxically, could be also devoted to human consumption. Indeed, 40 to 70% of animal diets IAFSs is composed by cereals and legumes, which provide high levels of energy and protein intake for animals (FAO et al., 2014), whilst locally produced roughage represents the major constituent of animal diets extensive small-scale grazing systems. In turns, the Milan Protocol (http://www.milanprotocol.com/), promoted at the 2015 EXPO in Milan, highlighted the great food paradox that “a large portion of crop and food production is funneled into animal feed or biofuels despite the widespread hunger and undernutrition”.

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

work by Mekonnen and Hoekstra (2011), which quantified country to global scale green (from rainfall), blue (from irrigation) and grey (virtual polluted volume) water footprints (m3⋅t-1) of 126 global crops over the period 1996–2005 (data available on Water Footprint Network, WFN; www.waterfootprint.org). The total water footprint of selected crops is given by the sum of all three components (i.e. green, blue and gray). Data on average crop yields and water footprints are given in Table S1. Sources of GHG emissions associated with both feed production and livestock management (from FAOSTAT) are deeply presented is Sect. 1.2. Feed Conversion Efficiency (FCE; Table 2) expresses the animals’ ability to convert a generic feeding requirement of entire breeding, having about 15–21% of proteins (Smil, 2002b), into edible animalsource proteins. FCE is generally considered an intrinsic feature of the animal species, and it is related to the animal diet (Herrero et al., 2013). Generally, the composition of feeds varies depending on the livestock type to which it is intended for and the countries where it is produced (Herrero et al., 2013; Archimède et al., 2011). However, qualitatively similar mixtures of cereal grains and legumes (Sanders, 1999) are used in intensive animal farming: maize, barley, wheat or sorghum for energy needs, i.e. giving calories and having specific crop protein content (CPC) less than 15% (Table S1), and soybean or lupins for the protein requirements, having CPC above 35% (Table S1). To be consistent with the definition of FCE, the protein content of a unit of feed mass should be somewhat constant and equal to approximately 20% (Smil, 2002b). Hence, we investigated eight possible feed mixtures (Table S3, see also legend in Fig. 2) composed of one energy and one protein crop (hereinafter C1 and C2, respectively), using a constant C1 to C2 ratio of 7:3. Due to the specific crop protein content (10–13% and 38%, for C1 and C2, respectively), such ratio provides a feed mass having about 20% of PP (i.e. 200 kg of PP per tons of feed mass). The assumption of constant feed protein content (FPC) allowed us to approach the human food vs. animal feed debate from a new perspective. Instead of estimating the footprint for specific foods here we estimated the footprints for food categories under similar conditions (i.e. from qualitatively similar feed mixtures) making them comparable. Again, since the higher the quality of the livestock diet, the higher the efficiency of the feed conversion into AP (Herrero et al., 2013), the feed mixtures here investigated allowed us to focus only on intensive, highly efficient, livestock production (as done in IAFSs). In this first analysis, we assumed that the area occupied by the farms as well as the direct water consumption for livestock activities (e.g. drinking of animals, maintenance of livestock farms etc.) are small enough when compared to the feed water and area demands so that they could be neglected (Mekonnen and Hoekstra, 2012). Environmental footprints were estimated (see Sec. S1.1 and S1.2 in SI) as average values for the period 1996–2005 to maximize comparability among the reference periods of the involved dataset (see Table 1). Area and GHG footprints were also estimated for the period 2006–2012, but this was not possible in case of the water footprint due to the lack of updated data on water consumptions. Environmental footprints of PP and AP were then compared showing their relative costs.

Over a decade ago, Smil (2002b) argued that the replacement of a relatively modest but cumulatively significant share of AP with PP in human diets could have great economic, nutritional and environmental benefits. Indeed, several works paved the promising path of a dietary shift, raising the issue of the sustainability of animal- vs. plant-source foods (Kastner et al., 2012; Pimentel and Pimentel, 2003; Baroni et al., 2007; Wirsenius et al., 2010; de Ruiter et al., 2014; Jalava et al., 2014) and exploring less resource-demanding animal products maintaining the relative contribution of non-feed systems (Davis and D’Odorico, 2015). However, these studies are mostly based on the metric of calories, despite the major source of inefficiency arises in the production of AP (which is expected to increase). Moreover, although a general consensus that IAFSs cause strong impacts on the environment (Aiking, 2011; Steinfeld et al., 2006; Pimentel and Pimentel, 2003; Thornton, 2010; Eshel et al., 2014), the question of feeding livestock with feed grains still remains controversial: on one hand it represents the most efficient feed (Herrero et al., 2013); on the other hand, it acts as the farming system that should not be promoted (Smil, 2014; Aiking, 2011) because of the food paradox and the strong environmental impacts. In this study, we explicitly addressed the global costs of producing AP from intensive, highly efficient, grain-fed livestock in terms of area demand, water consumption and GHG emissions (hereinafter environmental footprints), and compared them the with costs of producing an equivalent amount of PP. Based on the global scale focus and the data availability, we confined the analysis to the agricultural production components. We then moved the spotlight toward the actual per capita protein intakes in three representative regions (Northern America, Western Europe, and Eastern Asia) where IAFSs account for a great share of animal food production (Robinson et al., 2011; Steinfeld et al., 2006). In these regions, we found that the protein intake was higher than those recommended by the World Health Organization (WHO). Hence we postulated a set of straightforward hypotheses to shorten the average protein surplus (AP produced from IAFSs) and assessed the resulting amounts of area demand, water consumption and GHG emissions that could be virtually recovered. Since our study is based on the metric of protein and considers the most efficient animal-source food production (i.e. from feed grains), we believe that our contribution would provide an additional insight into the human food vs. animal feed competition debate. 2. Methods 2.1. Environmental footprints Fig. 1 summarizes the stepwise approach adopted to estimate the environmental footprints of both PP and AP. The whole procedure and the involved dataset are presented in detail in the Supplementary information (Sect. S1.1 and S1.2). Data sources (all available online) and their features are also reported in Table 1. Overall, we hypothesized several feed mixtures and for each of them we estimated the Feed Area Demand (HAF), Feed Water Footprint (WFF) and Feed GHG Emissions (GHGfeeding) for unit mass by tracking back to the average yields, average water consumption and average GHG emissions related to the crops composing the feeds, on global scale (Fig. 1, big red box). Footprints of PP and AP obtainable from feeds were then derived from feed protein content (FPC) and feed conversion efficiencies (FCEs) of livestock, respectively. GHG footprints of AP were considered as the sum of the GHGfeeding (i.e. the indirect sources due to crops composing the feed) plus the direct release from livestock management (GHGfarming, see Sect. S1.2 for details), because of the well assessed high impact of the latter (Robinson et al., 2011). Global average yields were taken from FAOSTAT database (FAO, 2015) which provides values (t⋅ha-1) for around 160 crops and 14 crops’ categories for several decades up to 2014 at country, regional and global level. Water consumptions were taken from the comprehensive

2.2. Opportunities for land recovery, water saving, and GHG emission reduction Despite IAFSs represent a global issue, so that the impacts' calculation asks for the worldwide view, they are mostly located in industrialized, large and quickly emerging countries where they provide a great share of animal-source food (Robinson et al., 2011; Steinfeld et al., 2006). Thus, we analysed the actual daily per capita protein intakes (from vegetal products total and animal products total items in the Food Sheet Balance of FAOSTAT) for three representative regions defined by the FAO as: Northern America (United States of America; 653

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

Fig. 1. Schematic overview of the methodological approach. White boxes represent mass store, valves represent proportional relationships between compartments. Red color acts for AP, green for PP. C1 = energetic crop, C2 = protein crop. Feed are supposed to be composed of a constant ratio C1:C2 of 7:3 (see Methods). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Bermuda; Canada; Greenland; Saint Pierre and Miquelon), Western Europe (Austria, Belgium, France, Germany; Luxembourg, Liechtenstein; Monaco; Netherlands; Switzerland) and Eastern Asia (China, mainland; China Hong Kong SAR; China, Macao SAR; China, Province of Taiwan; Democratic People's Republic of Korea; Japan; Mongolia; Republic of Korea). Overall, actual protein intakes were compared with protein requirement suggested by the World Health Organization (WHO), that correspond to an average daily protein intake of 0.83 g per Kg of body weight (WHO, 2007). Considering an average body weight of 75 Kg, the recommended amount of 62.25 g day−1 of proteins was considered. Since we found a positive balance between actual vs. suggested average daily per capita protein intakes in all these regions (both for 1996–2005 and 2006–2012 periods), we considered a set of straightforward scenarios to cut down the average surplus (Balance of proteins, Bp), either completely or partially, by reducing the AP from IAFSs. Related virtual opportunities of land recovery, water saving, and GHG

Table 2 Animal-based food Protein Content (APC, 0–1) and Feed to animal protein Conversion Efficiency (FCE, Kg(AP)t−1 (feed)) of the edible weight for major animal-based foods. Conversion factors are expressed as kg of edible animal protein (AP) obtained from one ton of feed mass (see SI for further details).

APC FCE

Milk

Fish

Eggs

Chicken

Swine

Beef

0.035 50.0

0.180 78.3

0.130 31.0

0.200 44.4

0.140 14.9

0.150 6.0

emissions' reduction were then explored. Scenario 1) abating an amount of daily per capita proteins supply from each animal food item proportionally to their actual consumption; Scenario 2) Abating 1 g (or the maximum possible if less than 1 g) of daily per capita protein supply from all animal food items without necessarily offsetting the Bp;

Table 1 Data sources, derived variables and units. Dataset

Variable

Symbol

Unit

Period

Reference

FAOSTAT Production Food Balance Sheet Food Balance Sheet Population Emission–Agriculture; Harmonizes World Soil Database

Yield animal food productions protein per capita supply Population GHG emissions

Y – – – –

t ha−1 t g d−1per capita in habitant Mt (CO2-eq)

1996–2012 1996–2012 1996–2011 1996–2011 1996–2012

FAO FAO FAO FAO FAO

Crop Water Footprint

WF

m3 t−1

1996–2005

Mekonnen and Hoekstra (2012)

Feed Conversion Efficiency Animal Protein Content Crop Protein Content Healthy protein requirement

FCE APC CPC

Kg(AP) t−1 (feed) Dimensionless Dimensionless g d−1 per capita

(2015) (2015) (2015) (2015) (2015), FAO (2006), FAO (2012)

Water Footprint Network Literature

654

–

Smil (2002a,b) Smil (2002a) Day (2013) WHO (2007)

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

Fig. 2. Area(HAAP), water (WFAP) and GHG (GHGAP) footprints of AP for the period 1996–2005 (black) and 2006–2012 (red). AP footprints were estimated for each food item considering eight different feed mixtures listed in legend. Food items are: M = milk; F = fish; E = eggs; C = chicken; S = swine; B = beef. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

kg(AP)−1, respectively), with sorghum-based feed being the most expensive in terms of footprints. Swine meat (S) follows with high scores (ranges of 0.018–0.052 ha kg(AP)−1, 95–186 m3·H2O kg(AP)−1and 22–28 kg−1 respectively). The footprints of eggs (E), fish (F) and (CO2-eq) kg(AP), chicken (C) show scores approximately reduced by half and lower variability than swine meat (ranges 0.003–0.025 ha kg(AP)−1, 18–90 m3·H2O kg(AP)−1 and 2–17 kg(CO2-eq) kg(AP)−1, respectively), with fish being the less expensive animal-source food in terms of area, water, and GHG emissions. Milk production seems to require less land and water than beef, swine, eggs, and chicken, but it is second after beef in terms of GHG emissions. Environmental footprints were consistent between the two time periods considered in this study. Focusing on the feed compositions, maize-soybean is the mixture that demands less land (see Tables S3 and S4) and, unsurprisingly, also the most common feed mix (Sanders, 1999). Because of its relatively low yields, sorghum is the energy crop demanding most land and water, with other conditions being equal. The comparison between environmental footprints (Fig. 3, Tables S6 and S12) are shown by means of ratios: HAAP to HAPP and WFAP to WFPP for land and water consumption (hereinafter EAP) and GHGAP to GHGPP for GHG emissions (hereinafter EAP_GHG). AP result approximately from 2.4 (for fish) to 33 (for beef) more expensive in terms of land and water footprints, with milk, fish, eggs, and chicken scoring relatively lower values (EAP < 7), swine and beef the highest (from 12 for swine up to 33 for beef). As shown in Sect. S1.1 (SI), EAP only depends on the FCE, thus its scores are independent of time frames (left panel, Fig. 3). As far as the GHG emissions are concerned (Fig. 3, right panel), the cost of producing AP vs. PP is noticeably higher: EAP_GHG ranged from ca. 2.4 (for fish) to approximately 240 (for beef), with fish, eggs, and

Scenario 3) Abating the whole Bp from each animal food items, where possible, in order to explore their potential contribution. Details and data source used to explore the opportunities for land recovery, water saving and GHG emission reductions together with scenarios for offsetting Bp are deeply presented in Sect. S1.3 of SI. Data source are also listed in Table 1. For a more intuitive understanding, potential recoveries of area, water, and GHG emissions were compared with the global estimates of arable plus permanent croplands (FAO, 2015), water use in agriculture (http://www.globalagriculture.org/report-topics/water.html) and global emissions from the agriculture sector (excluding energy), reported in Table 3. Since global average of water use in agriculture is given by the components of rainfall and irrigations (i.e. it does not consider the virtual polluted water), opportunities of water saving were explored (and then compared with global average consumption) using AP water footprints recalculated considering only the green (from rainfall) plus blue (from irrigation) components.

eq)

3. Results Area, water and GHG footprints of AP (HAAP, WFAP, GHGAP, respectively) are shown in Fig. 2 and are also reported in Tables S4, S5, and S11. Area, water and GHG footprint of PP (HAPP, WFPP, GHGPP, respectively) were reported in Tables S3 and S8, but not shown in Fig. 2 due to their relatively low scores. Overall, beef (B, Fig. 2) appears the most demanding food commodity in terms of land and water use and releases the larger amount of GHG. The footprints of beef also show large variability with feed composition (ranges of 0.044–0.13 ha kg(AP)−1, 236–463 m3·H2O kg(AP)−1 and 221–240 kg(CO2655

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

farming component dominates the GHG emissions. The balance between daily per capita protein supply and requirement, as suggested by the WHO, results to be always positive for all the regions considered (Fig. 4 and Table 3), yielding an average surplus (Bp) of +27.9 and +33.6 g d−1 per capita, respectively, for the first and second period. Due to the large difference between AP and PP footprints, opportunities of land and water saving, while reducing GHG emissions are reported only assuming the offset of the AP (Fig. 5). Due to their high AP footprints, the reduction of beef and swine show the highest potential contribution in terms of environmental benefit. Opportunities for area recovery, water saving and GHG emissions’ reductions from all the scenarios are always less than 8% for the food items of fish, eggs, milk, and chicken(8% obtained from chicken, Scenario 3, in terms of area recovery). By contrast, achieving the reduction target Bp, as proposed by Scenario 1 from swine and beef, would allow: (i) a global land recovery ranging from 6(7)% to 34(38)% for the first(second) time frame; (ii) saving an amount of freshwater ranging from 5to 34%; (iii) and reducing GHG emissions from 2(2)% to 20(21)% for the first(second) time frame. Similar paths can be achieved from the Scenario 2: Global land recovery from 1% to 6% in both time frames; (ii) water saving ranging from 1 to 4% for the reference period 1996–2005; (iii) and reducing GHGs up to 4(3)% for the first(second) time frame. Lastly, according to Scenario3, the potential land and water saving afforded by a reduction of AP from swine meat or beef would be around: (i) 9(9) to 47(46)% for land with respect to the first(second) time frame; (ii) 7 to 33% for freshwater consumption with respect to the first time frame; (iii) 3(3) to 29(25)% for GHG emissions with respect to the first (second) time frame.

Table 3 Daily per capita intake of AP and PP in the selected regions and corresponding balance with respect to the daily per capita protein supply of 62.25 g (Rp) recommended by WHO (2007) obtained for the periods of 1996–2005 and 2006–2011. *Average across regions are weighted by the region’s population, also in the table. Global arable plus permanent crop area, green and blue water use and GHG emissions in agriculture are also reported. Unit

1996–2005

2006–2011

Northern America

71.6

71.2

Western Europe

65.9

65.9

Eastern Asia

29.8

36.6

39.7

44.8

Northern America

41.3

40.5

Western Europe

38.0

39.8

Eastern Asia

53.9

54.6

50.5

51.0

Average daily per capita AP supply

−1

gd per capita

Weighted Average* Average daily per capita PP supply

g d−1 per capita

weighted Average* −1

Average daily per capita protein supply (grant total)

gd per capita

90.1

95.8

Balance with respect to Rp

g d−1 per capita

+27.9

+33.6

Population 0.31·109

0.34·109

Western Europe

0.18·10

9

0.19·109

Eastern Asia

1.50·109

1.56·109

9

1.54 109

Northern America

World Arable land plus permanent crop area World Green plus Blue water used in agriculture World GHG emissions in the agricultural sector

Ha

1.52 10

Km3

7735

Mt(CO2-eq)

4707

5196

4. Discussion Motivated by the world nutrition paradox “a large portion of crop and food production is funnelled into animal feed or biofuels despite widespread hunger and undernutrition”, expressed in the Milan Protocol (http:// www.milanprotocol.com/) at the 2015 EXPO in Milan, we performed an explicit global comparison between the footprints of animal-source proteins obtained from intensive, highly efficient, livestock production and plant-source proteins. Our major findings revealed that producing AP from intensive

chicken scoring relatively low values (EAP_GHG < 10), followed by milk and swine (EAP_GHG between 30–41). Since protein content of the feed mix has been maintained constant at ∼20%, EAP and EAP_GHG will remarkably change among food items but slightly among feed compositions. Greatest differences are found between AP and PP in terms of GHG footprints because the animal

Fig. 3. The relative (environmental) costs of producing AP with respect to PP. Left panel: area and water demand (EAP, defined as the ratios of HAAP:HAPP and WFAP:WFPP). Box plots represent the median (horizontal bars in bold), the interquartile range (boxes) and the extension to maximum and minimum values (whiskers). EAP only depends on FCE, thus its values resulted equal between land and water and between time frames (see Sect. S1.1 in S.I.). Right panel: as the left panel but for GHG emissions (EAP_GHG, defined as the ratio GHGAP:GHGPP). Symbols for food items as in Fig. 2.

656

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

Fig. 4. Average daily per capita protein supply by regions for the two time frames (left bars: 1996–2005; right bars 2006–2012).

hypothesized scenario to bring down the surplus. However, we can speculate on the most trivial (and perhaps the easiest to be pursued) Scenario 2 which assumes a reduction of just 1 g of AP per capita per day. Integrated into real actions, Scenario 2 suggests that if people from Northern America, Western Europe, and Eastern Asia reduced the consumption of ca. 30 g of AP per month from beef (the equivalent of the proteins stored approximately in one beef-burger of 200 g), the planet would recover at least 2% up to 6% of the actual world arable land, 2–4% of the world water used in agriculture and 3–4% of the global carbon emissions originating from the agriculture sector. These percentages are apparently small but converted into the absolute values of land surface, water volumes and emissions they appear highly significant. For instance, let we consider that clear-cutting activities in tropical rain forest proceed approximately at 0.5% per year with respect to the global arable land (Achard et al., 2014). Besides the meaningful comparison of +2% potential land recovery from a minimal animal-source food reduction (one beef-burger per month in the previous example) vs. −0.5% from stopping clear cutting activities, we deal with a great paradox again: soybean cultivation for animal feed has been a major driver of deforestation in the Brazilian Amazonian region (Smil, 2014; Steinfeld et al., 2006; Garnett, 2008). Similar considerations should be made also for the phenomenon of foreign large-scale acquisition of agricultural land (D’Odorico and Rulli, 2013; Rulli and D’Odorico, 2013). Obviously, these calculations are just the best approximation of relevant global totals aimed at revealing realistic opportunities. Another minor consideration on our findings regards the area and GHG footprints (water footprints lack updated data) of both PP and AP, which did not show appreciable change over time, likely reflecting the slight yield increase that has taken places over the last decade. However, our results also suppose several approximations that need to be further investigated. First of all, we excluded from our analysis the environmental impacts related to the whole food chain. We did not consider, for example, the GHG emissions due to energy use in the cultivation practices, food processing, and transportation, which can be demonstrably relevant (Steinfeld et al., 2006; Garnett, 2008), albeit they do not represent the biggest components of the sector’s impact on climate change (IPCC, 2007; Gephart et al., 2014a). Hence, our estimates were conservative and the actual emission footprints could be reasonably higher. Second, we treated only quantitative aspects yet livestock production causes strong impacts, as water and soil pollutions (Steinfeld et al.,

farming is approximately 2.4 to 33 more expensive in terms of land and water demand and 2.4 to 240 more pollutant in terms of GHG emissions than those of producing PP, calling for several considerations on the general habits of animal-food consumption and, in particular, the share of proteins provided by IAFSs. The world’s richest population is far from being vegan or vegetarian (Smil, 2014), being confident that fish, meat, eggs and dairy products provide in addition to proteins a range of essential nutrients as iron, calcium, vitamin B12 and fat (Sanders, 1999). However, in developed countries, an increasing number of people are eschewing meat while opting for a vegetarian diet (Sanders, 1999), which in turns has been proven to be nutritionally adequate when based on a wider variety of foods such as salads, fruits, nuts, and pulses on a regular basis (Day, 2013; Sanders, 1999; Ranganathan et al., 2016). With special regards to protein contents, crops such as lupins, soybean, chickpeas and dry peas incorporate more proteins per weight than meat (see Table S1) and impact less on the environment. Likewise, these crops could also help to meet food security in poor countries where diets are mainly plantbased, but the choice of foods is more restricted and thus nutritional deficiencies are more likely. Surprisingly, FAOSTAT does not report the actual consumption of PP from foods such as chickpeas and lupins (i.e. they result as missing data), suggesting that today the consumption of these foods is still relatively low, despite their nutritional worth. Furthermore, these crops are mostly legumes, responsible for nitrogen fixation in the soil through their residues and leaving nitrogen oxides from the residues after harvest and, thus, have lower net emission footprint (Lüscher et al., 2014). Again, an increase in vegetable protein intake and a decrease in animal protein intake reduce the risk of osteoporosis, heart disease and cancers developments (McEvoy et al., 2012; Sellmeyer et al., 2001; Key et al., 1999; Appleby et al., 1999; Freeland-Graves and Nitzke, 2013). Although these evaluations are beyond the scope of our work, information on safe diets should complement our findings on environmental sustainability. When we looked at the actual per capita protein supply (in Northern America, Western Europe, and Eastern Asia) with respect to the health requirement suggested by the WHO. We found a surplus in both time periods (27.9 and 33.6 g day−1 per capita, respectively) that motivated us to determine to what extent land consumption, water use, and GHG emissions could be reduced assuming the offset of the surplus from AP. Best opportunities resulted from the reduction of swine and beef productions (Fig. 5). Estimations of environmental benefits vary widely according to the 657

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

Fig. 5. Potential global recovery of arable area, water consumption, and GHG emissions. Numbers in brackets refer to the hypothesized scenarios for offsetting the average surplus of daily per capita protein intakes resulted in the addressed regions. Symbols for food items and box plots as in Figs. 2 and 3, respectively.

diet to produce livestock, and the FCEs we used (from Ref. (Smil, 2002a)) fall in the lower-end range reported by the comprehensive work of Herrero et al. (2013), being consistent with the fact that they are among the most efficient conversion factors and related to livestock diets of high quality. On the other hand, feed grains require monocultures (and land and water) for crops that are also suitable for human nutrition. In this regard, our findings lead us to agree with Smil (2014), suggesting that due to their large impacts on land, water, and GHGs emissions, feed grains should be limited and better compromises can be reached by a moderate consumption of meat produced with non-feed grains systems (stover, crop residues, and sustainable grazing).

2006), that should also be taken into account. For instance, in our analyses, fish appears as the animal food with lower footprints and, in this sense, Gephart et al. (2014b) and Troell et al. (2014) agreed on how the marine proteins rather than terrestrial proteins are a fundamental and sustainable source for a low water footprint diet, especially in water-stressed regions. However, we must be careful in not translating these results as the free pass to further intensify fish catching and aquaculture fish production and end up with serious consequences of water quality and sea overexploitation (Lam et al., 1994; Wu, 1995; Kalantzi et al., 2013). While aquaculture becomes more important for feeding the growing world population, it also incorporates more terrestrially-based feeds, and this should be accounted for in fishery management policies, in the promotion of sustainable aquaculture (Gephart et al., 2014c), and in the selection of feed mixtures (Pahlow et al., 2015). Third, we analyzed the animal food produced only by IAFSs (i.e. livestock mostly fed with worth grains) because of its controversial aspect. On the one hand, feed grains are proven to be the most efficient

5. Conclusion Overall, we pointed out remarkable results that could bring only strong benefits in return: plant-source proteins demand less natural (land and water) resources, release less GHGs, help the natural regeneration of the soil and, from a medical point of view, appear to be 658

Land Use Policy 67 (2017) 652–659

A. Di Paola et al.

IPCC, 2007. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In: Parry, M.L., Canziani, O.F., Palutikof, J.P., van der Linden, P.J., Hanson, C.E. (Eds.), Cambridge University Press, Cambridge, UK 976pp. Jalava, M., Kummu, M., Porkka, M., Siebert, S., Varis, O., 2014. Diet change – a solution to reduce water use? Environ. Res. Lett. 9 (7), 074016. Kalantzi, I., Shimmield, T.M., Pergantis, S.A., Papageorgiou, N., Black, K.D., Karakassis, I., 2013. Heavy metals, trace elements and sediment geochemistry at four Mediterranean fish farms. Sci. Total Environ. 444, 128–137. Kastner, T., Ibarrola Rivas, M.J., Koch, W., Nonhebel, S., 2012. Global changes in diets and the consequences for land requirements for food. Proc. Natl. Acad. Sci. U. S. A. 109 (18), 128–137. Key, T.J., Davey, G.K., Appleby, P.N., 1999. Health benefits of a vegetarian diet. Proc. Nutr. Soc. 58, 271–275. Lüscher, A., Mueller-Harvey, I., Soussana, J.F., Rees, R.M., Peyraud, J.L., 2014. Potential of legume-based grassland-livestock systems in Europe: a review. Grass Forage Sci. 69 (2), 206–228. Lam, K.S., MacKay, D.W., Lau, T.C., Yam, V., 1994. Impact of marine fish farming on water quality and bottom sediment: a case study in the sub-tropical environment. Mar. Environ. Res. 38 (2), 115–145. Manceron, S., Ben-Ari, T., Dumas, P., 2014. Feeding proteins to livestock: global land use and food vs. feed competition. OCL 21 (4), D408. http://dx.doi.org/10.1051/ocl/ 2014020. McEvoy, C.T., Temple, N., Woodside, J.V., 2012. Vegetarian diets, low-meat diets and health: a review. Public Health Nutr. 15 (12), 2287–2294. Mekonnen, M.M., Hoekstra, A.Y., 2011. The green, blue and grey water footprint of crops and derived crop products. Hydrol. Earth Syst. Sci. 15 (5), 1577–1600. Mekonnen, M.M., Hoekstra, A.Y., 2012. A global assessment of the water footprint of farm animal products. Ecosystems 15 (3), 401–415. Pahlow, M., van Oel, P.R., Mekonnen, M.M., Hoekstra, A.Y., 2015. Increasing pressure on freshwater resources due to terrestrial feed ingredients for aquaculture production. Sci. Total Environ. 536, 847–857. Pimentel, D., Pimentel, M., 2003. Sustainability of meat-based and plant-based diets and the environment. Am. J. Clin. Nutr. 78 (3), 660S–663S. Ranganathan, J., Vennard, D., Waite, R., Dumas, P., Lipinski, B., Searchinger, T., GLOBAGRI-WRR model authors, 2016. Shifting Diets for a Sustainable Food Future. Working Paper, Installment 11 of Creating a Sustainable Food Future. World Resource Institute, Washington D.C Accessible at: http://www.worldresourcesreport. org. Robinson, T.P., Thornton, P.K., Franceschini, G., Kruska, R.L., Chiozza, F., Notenbaert, A., Cecchi, G., Herrero, M., Epprecht, M., Fritz, S., You, L., Conchedda, G., See, L., 2011. Global Livestock Production Systems. Food and Agriculture Organization of the United Nations (FAO) and International Livestock Research Institute (ILRI), Rome. Rulli, M.C., D’Odorico, P., 2013. The water footprint of land grabbing. Geophys. Res. Lett. 23 (40), 6130–6135. http://dx.doi.org/10.1002/2013GL058281. Sanders, T.A.B., 1999. The nutritional adequacy of plant-based diets. Proc. Nutr. Soc. 58 (02), 265–269. Sellmeyer, D.E., Stone, K.L., Sebastian, A., Cummings, S.R., 2001. A high ratio of dietary animal to vegetable protein increases the rate of bone loss and the risk of fracture in postmenopausal women. Am. J. Clin. Nutr. 73 (1), 118–122. Smil, V., 2001. Enriching the Earth: Fritz Haber, Carl Bosch, and the Transformation of Food Production. MIT Press, Cambridge (MA) ISBN 9780262194495. Smil, V., 2002a. Nitrogen and food production: proteins for human diets. Ambio 31 (2), 126–131. Smil, V., 2002b. Worldwide transformation of diets, burdens of meat production and opportunities for novel food proteins. Enzyme Microb. Techol. 30 (3), 305–311. Smil, V., 2014. Eating meat: evolution patterns, and consequences. Global Food Secur. 3, 67–71. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M., De Haan, C., 2006. Livestock's Long Shadow. FAO, Rome p. 392. Thornton, P.K., 2010. Livestock production: recent trends future prospects. Philos. Trans. R. Soc. B 365, 2853–2867. Tilman, D., Clark, M., 2014. Global diets link environmental sustainability and human health. Nature 515 (7528), 518–522. Troell, M., Metian, M., Beveridge, M., Verdegem, M., Deutsch, L., 2014. Comment on ‘Water footprint of marine protein consumption – aquaculture’s link to agriculture. Environ. Res. Lett. 9, 109001. Tubiello, F.N., Salvatore, M., Rossi, S., Ferrara, A., Fitton, N., Smith, P., 2013. The FAOSTAT database of greenhouse gas emissions from agriculture. Environ. Res. Lett. 8, 015009. WHO, 2007. Protein and Amino Acid Requirements in Human Nutrition. WHO Technical Reports Series No. 935. Wirsenius, S., Azar, C., Berndes, G., 2010. How much land is needed for global food production under scenarios of dietary changes and livestock productivity increases in 2030? Agric. Syst. 103 (9), 621–638. World Food Programme (WFP), 2012. Nutrition at the World Food Programme. Programming for Nutrition-Specific Interventions. World Food Programme, Rome, Italy Available at: http://documents.wfp.org/stellent/groups/public/documents/ communications/wfp258650.pdf. Wu, R.S.S., 1995. The environmental impact of marine fish culture: towards a sustainable future. Mar. Pollut. Bull. 31 (4–12), 159–166. de Ruiter, H., Kastner, T., Nonhebel, S., 2014. European dietary patterns and their associated land use: variation between and within countries. Food Policy 44, 158–166.

healthier (FAO, 2006). While the transition of our energy/production systems may require enormous investments and take a long time, people, in turn, can start changing their food choices today and with little investment. In the light of these results and the current know-how, we hope our study will contribute to increase the awareness and interest in strategies to reduce feed grains livestock production while supporting the benefit of a moderate yet sustainable animal food demand, and increase food security and nutritionally adequate intakes in poor countries. Acknowledgements This work was supported by the Italian Ministry of Education, University and Research and the Italian Ministry of Environment, Land and Sea under the project GEMINA (n. 232/2011). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.landusepol.2017.06. 017. References Achard, F., Beuchle, R., Mayaux, P., Stibig, H.J., Bodart, C., Brink, A., Lupi, A., et al., 2014. Determination of tropical deforestation rates and related carbon losses from 1990 to 2010. Global Change biology 20 (8), 2540–2554. Aiking, H., 2011. Future protein supply. Trends Food Sci. Technol. 22, 112e120. Appleby, P.N., Thorogood, M., Mann, J.I., Key, T.J., 1999. The Oxford vegetarian study: an overview. Am. J. Clin. Nutr. 70 (3), 525s–531s. Archimède, H., Eugène, M., Marie Magdeleine, C., Boval, M., Martin, C., Morgavi, D.P., Lecompte, P., Doreau, M., 2011. Comparison of methane production between C3 and C4 grasses and legumes. Anim. Feed Sci. Technol. 166, 59–64. Baroni, L., Cenci, L., Tettamanti, M., Berati, M., 2007. Evaluating the environmental impact of various dietary patterns combined with different food production systems. Eur. J. Clin. Nutr. 61 (2), 279–286. D’Odorico, P., Rulli, M.C., 2013. The fourth food revolution. Nat. Geosci. 6 (6), 417–418. http://dx.doi.org/10.1038/ngeo1842. Davis, K.F., D’Odorico, P., 2015. Livestock intensification and the influence of dietary change: a calorie-based assessment of competition for crop production. Sci. Total Environ. 538, 817–823. Day, L., 2013. Proteins from land plants–potential resources for human nutrition and food security. Trends Food Sci. Technol. 32 (1), 25–42. Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S., Courbois, C., 2001. Livestock to 2020: the next food revolution. Outlook Agric. 30 (1), 27–29. Ehui, S., Li-Pun, H., Mares, V., Shapiro, B., 1998. The role of livestock in food security and environmental protection. Outlook Agric. 27, 81–88. Eshel, G., Shepon, A., Makov, T., Milo, R., 2014. Land, irrigation water, greenhouse gas, and reactive nitrogen burdens of meat, eggs, and dairy production in the United States. Proc. Natl. Acad. Sci. U. S. A. 111 (33), 11996–12001. FAO, IDF, IFCN, 2014. World Mapping of Animal Feeding Systems in the Dairy Sector. FAO, Rome, Italy. FAO, 2006. Fertilizers Used by Crops. FAO Fertilizer And Plant Nutrition Bulletin. N. 17. FAO/IIASA/ISRIC/ISSCAS/JRC, 2012. Harmonized World Soil Database (version 1.2). FAO, Rome, Italy and IIASA, Laxenburg, Austria. FAO, 2015. Food and Agriculture Organization (FAO). Faostat Database. Available at: http://faostat.fao.org/ (last Accessed: 13 December 2015). Foley, J.A., Ramankutty, N., Brauman, K.A., Cassidy, E.S., Gerber, J.S., Johnston, M., Mueller, N.D., O’Connell, C., Ray, D.K., West, P.C., Balzer, C., Bennett, E.M., Carpenter, S.R., Hill, J., Monfreda, C., Polasky, S., Rockström, J., Sheehan, J., Siebert, S., Tilman, D., Zaks, D.P.M., 2011. Solutions for a cultivated planet. Nature 478 (7369), 337–342. Freeland-Graves, J.H., Nitzke, S., 2013. Position of the academy of nutrition and dietetics: total diet approach to healthy eating. J. Acad. Nutr. Dietetics 113 (2), 307–317. Garnett, T., 2008. Cooking up a Storm: Food, Greenhouse Gas Emissions and Our Changing Climate. Food Climate Research Network, Centre for Environmental Strategy, University of Surrey. Gephart, J.A., Pace, M.L., D’Odorico, P., 2014a. Freshwater savings from marine protein consumption. Environ. Res. Lett. 9. http://dx.doi.org/10.1088/1748-9326/9/1/ 014005. Gephart, J.A., Pace, M.L., D’Odorico, P., 2014b. Reply to comment on ‘Water footprint of marine protein consumption – aquaculture's link to freshwater’. Environ. Res. Lett. 9. http://dx.doi.org/10.1088/1748-9326/9/10/109002. Herrero, M., Havlík, P., Valin, H., Notenbaert, A., Rufino, M.C., Thornton, P.K., Blümmel, M., Weiss, F., Grace, D., Obersteiner, M., 2013. Biomass use, production, feed efficiencies, and greenhouse gas emissions from global livestock systems. Proc. Natl. Acad. Sci. U. S. A. 110 (52), 20888–22089.

659