Science of the Total Environment 637–638 (2018) 221–232

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Organic quinoa (Chenopodium quinoa L.) production in Peru:

Environmental hotspots and food security considerations
using Life Cycle Assessment
Eduardo Cancino-Espinoza 1, Ian Vázquez-Rowe ⁎,1, Isabel Quispe
Peruvian LCA Network, Department of Engineering, Pontificia Universidad Católica del Perú, Avenida Universitaria 1801, San Miguel, Lima 15088, Peru

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Peru is the main producer of quinoa

worldwide with an increased demand
for exports.
• Organic quinoa in Peru was analyzed
using Life Cycle Assessment methodol-
ogy.
• GHG emissions were dominated by on-
field emissions due to fertilization.
• Organic quinoa is a low-carbon protein-
rich product as compared to the litera-
ture.
• Shifting from subsistence to intensive
farming defies environmental
sustainability.

a r t i c l e i n f o a b s t r a c t

Article history: Quinoa is a plant that is cultivated in the Andean highlands across Peru and Bolivia. It is increasingly popular due
Received 21 March 2018 to its high nutritive value and protein content. In particular, the cultivation of organic quinoa has grown substan-
Received in revised form 2 May 2018 tially in recent years since it is the most demanded type of quinoa in the foreign market. Nevertheless, despite the
Accepted 3 May 2018
interest that quinoa has generated in terms of its nutritional properties, little is known regarding the environ-
Available online xxxx
mental profile of its production and processing. Therefore, the main objective of this study was to analyze the en-
Editor: D. Barcelo vironmental impacts that are linked to the production and distribution of organic quinoa to the main export
destinations through the application of the Life Cycle Assessment (LCA) methodology. An attributional LCA per-
Keywords: spective was conducted including data from approximately 55 ha of land used for quinoa production in the re-
Andean crops gions of Huancavelica and Ayacucho, in southern-central Peru. IPCC, 2013 and ReCiPe 2008 were the two
Food policy assessment methods selected to estimate the environmental impact results using the SimaPro 8.3 software. Re-
Food security sults, which were calculated for one 500 g package of organic quinoa, showed that GHG emissions are in the
Industrial ecology upper range of other organic agricultural products. However, when compared to other high protein content
Life Cycle Assessment
food products, especially those from animal origin, considerably low environmental impacts are obtained. For in-
Protein
stance, if 20% of the average annual beef consumption in Peru is substituted by organic quinoa, each Peruvian
would mitigate 31 kg CO2eq/year in their diet. Moreover, when the edible protein energy return on investment
(i.e., ep-EROI) is computed, a ratio of 0.38 is obtained, in the higher range of protein rich food products. However,
future research should delve into the environmental and food policy implications of agricultural land expansion
to produce an increasing amount of quinoa for a growing global demand.
© 2018 Elsevier B.V. All rights reserved.

⁎ Corresponding author.
E-mail address: ian.vazquez@pucp.pe (I. Vázquez-Rowe).
1
Both authors contributed equally to the development of the article.

https://doi.org/10.1016/j.scitotenv.2018.05.029
0048-9697/© 2018 Elsevier B.V. All rights reserved.
222 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232
1. Introduction
Quinoa (Chenopodium quinoa L.) is an Andean pseudocereal
flowering plant of the Amaranthaceae family grown as a grain crop for
the commercialization of its seeds (MINAGRI, 2017). Quinoa has
adapted to different geographical areas throughout the Andes, which
has derived in the existence of five different varieties of quinoa based
on their ecotype (León, 2003; Bazile et al., 2015): i) Inter-Andean val-
leys ecotype, which has long growth periods between 2000 and
3000 m above sea level; ii) Highlands ecotype, a smaller variety resis-
tant to frosts; iii) Yungas ecotype, which is common in the Bolivian sub-
tropical forest; iv) Salares ecotype, which has high resistance to saline
soils and has a higher protein content than the other varieties; and,
v) Coastal ecotype, which occurs mainly in the lowlands in northern
Chile. Quinoa is usually cultivated in sandy loam or clay loam soils,
with an approximate pH of 5.5 to 8.5, with good drainage and moderate
slopes. The optimal temperature for quinoa is around 8–15 °C, although
it can withstand up to −4 °C (León, 2003). The vegetative period of qui-
noa ranges from 90 to 240 days.
Quinoa seeds have been an important source of protein for indige-
nous communities in the Andean altiplano in Peru, Bolivia and
Ecuador for centuries, but have only become popular elsewhere in re-
cent decades given their interesting nutritional properties (Jacobsen,
2003). Its protein content, according to León (2003), ranges between
12.5% and 14% dry content, but it is also highly considered due to its
high vitamin levels (e.g., high levels of riboflavin – vitamin B2 – and
α-tocopherol – vitamin E as compared to other cereals) and the fact
that, unlike most cereals, it is gluten free (Ruales and Nair, 1994;
Vega-Gálvez et al., 2010; Ruiz et al., 2014).
In Peru, the production of quinoa has been concentrated in a variety
of agro-ecological and physiographic zones of the altiplano. Despite the
fact that quinoa is highly resistant to pH and droughts, its proliferation
at different heights above 3000 m responds to the need of quinoa
farmers to avoid crops losses (Morlon, 1982; Aguilar and Jacobsen,
2003). In fact, new climate change patterns, that expect longer drought
periods in many areas of the Peruvian Andes (SENAMHI, 2009), consti-
tute an important threat to the production of quinoa, although climatic
uncertainty has not prevented the crop from growing at an annual rate
of 5.8% in terms of cultivated area (FAO, 2016). The most recent avail-
able data for the 2013–2014 campaign indicate that a total of 114,725
metric tons of quinoa were produced in Peru in an area of 68,140 ha
(FAO, 2016). In that same campaign, economic revenue related to qui-
noa exports surpassed 200,000 USD (MINAGRI, 2017). Although world-
wide data on quinoa production is not consistent through different
sources, it coincides in the fact that Peru is the main producer, followed
by Bolivia (59.8% and 38.8%, respectively, according to FAOSTAT).
Ecuador and the north of Chile also have small productions of quinoa, al-
though some farming is starting in France and Spain in an attempt to ex-
tend its production (El Pais, 2016; Quinoa d'Anjou, 2017).
Yield rates vary considerably between varieties of quinoa and re-
gions. For instance, Fairlie-Reinoso (2016) reports that grain yield can
be as high as 3820 kg/ha in Arequipa, whereas in the highlands of
Puno this value tends to range from 800 kg/ha to 2500 kg/ha depending
on the degree of mechanization and technical assistance that the
farmers receive. In addition, it should be noted that organic quinoa pro-
duction rarely surpasses 1600 kg/ha. According to FAOSTAT, the average
yield in Peru in 2014 was 1683 kg/ha (FAO, 2016). However, the cultiva-
tion of organic quinoa has increased tenfold in the period 2008–2016 for
two main reasons (MINAGRI, 2017; Willer and Lernoud, 2014): i) it is
the most demanded type of quinoa in the ever growing foreign market
(Fairlie-Reinoso, 2016); and ii) the economic revenue, if well managed,
is considerably higher (i.e., from 50% to 120% times higher depending on
market fluctuations) than that of conventional quinoa despite the lower
yield rates (Fairlie-Reinoso, 2016). Organic agricultural production in
Peru is regulated by Supreme Decree N°044-2006-AG that fixes a tech-
nical regulation for organic production and Law N°29,196 that
advocates for the promotion of organic products. Regardless of the com-
mon enforcements on organic products, which include the control on
the types of fertilizers and plant growth agents allowed, the clear divi-
sion from conventional cultivation areas, traceability and clear labelling,
Peruvian legislation also states that organic production should aim at
minimizing all forms of pollution and (El Peruano, 2008).
From an environmental perspective, current field management ac-
tivities for organic quinoa lack an irrigation system and present low
mechanization levels. Moreover, organic quinoa presents a minimal
use in the amount of plant protection agents (e.g., Serenade or Xentari)
used, and most of the organic fertilizers are either local guano from cat-
tle or guano de isla from the Peruvian coastal isles. Although evidence in
the scientific literature concerning whether the choice of organic food
products as compared with conventionally-grown products is beneficial
to human health is sparse and inconclusive (Williams, 2002; Smith-
Spangler et al., 2012), organic production is commonly perceived by
consumers as being free of pesticides, better for one's health and better
for the environment (Padel and Foster, 2005; Lee and Hwang, 2016).
This situation explains the high amount of organic quinoa that are
demanded from abroad (Gestión, 2016; MINAGRI, 2017), mainly the
United States, Australia, Canada and the European Union (Fairlie-
Reinoso, 2016), although it is worth noting that consumption of organic
products, including quinoa, has experienced a considerable surge in the
city of Lima (Higuchi, 2015).
However, the environmental profile of quinoa using environmental
assessment tools remains unexplored. Hence, the main objective of
this study is to analyze the environmental impacts that are linked to
the production and distribution of organic quinoa to the main consump-
tion destinations (domestic and abroad) through the application of the
Life Cycle Assessment (LCA) methodology (ISO, 2006a, 2006b). In addi-
tion, the aim of the study is focused on aiding stakeholders in the iden-
tification of environmental hotspots and supporting public policies.
Thereafter, the results obtained were linked to dietary patterns and
food security in Peru in terms of policy support. As far as the researchers
were able to ascertain, this study constitutes the first full LCA on quinoa
production and distribution, providing life cycle inventories and impact
assessment for researchers and other stakeholders.
2. Materials and methods
2.1. Methodological framework, goal and scope
The environmental assessment methodology considered in this
study, LCA, was used following the ISO 14040 standard (ISO, 2006a).
The main goal was to bring forward a thorough environmental analysis
linked to the production and distribution of organic quinoa in Peru to
the main destinations, in order to identify the main environmental
hotspots throughout the production system. The methodological ad-
vance of the study is justified based on the fact that, as far as we were
able to ascertain, this study is the first application of LCA to Andean
grains. Moreover, results are intended to provide a basis to suggest im-
provement actions in a sector that is expected to experience important
growth in the next decade. Finally, results aim to be of interest for the
wider LCA community, since quinoa is starting to be incorporated into
diets worldwide due to its high protein content (Navruz-Varli and
Sanlier, 2016).
Data for the study were collected in late 2016 for the 2015–2016 qui-
noa production season. The function of the system was to deliver or-
ganic quinoa cultivated in several sites to the main export destinations
(mainly abroad, but also the city of Lima) ready to supply wholesalers.
Therefore, the selected functional unit (FU), in other words, the quanti-
fiable reference to which material and energy flows referred (ISO,
2006b), was set as one 500 g packet of quinoa ready for retailing in a su-
permarket in Lima or exported to the United States (US) or the
European Union (EU) by marine freight through the port of Callao
(12° 2′S; 77° 8′W). In addition, as part of the sensitivity analysis (see
 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232
Section 2.6), a nutritional-based FU is presented with the aim of com-
paring quinoa with other food products rich in protein.
2.2. System boundary
The system boundary, which is depicted in Fig. 1, includes the differ-
ent phases of quinoa production from soil tillage up to the distribution
of the final product to a regional distribution center (RDC). All field op-
erations (tillage, sowing, harvesting) were included in the system, as
well as the provision of organic fertilizers and plant protection agents.
Post-harvesting operations included drying, threshing, cleaning and
classification and packaging. Pollutant emissions were also included
throughout the production system, although toxicity-related emissions
were not computed in the study. Given that the retailing and consump-
tion of the final product were excluded from the assessment, the end-
of-life of the final package was also excluded from the system boundary.
2.3. Data acquisition
Primary data were obtained from a cluster of producers located in
the regions of Ayacucho and Huancavelica, in the southern area of the
Peruvian Andes (see Fig. 2). The farmers from Ayacucho were located
in the small village of Huamanguilla (13°00′S; 74° 10′W), whereas the
farmers contacted in Huancavelica were located in the village of
Antaparco (13°04′S; 74°24′W). Both are very small and remote commu-
nities, in which the main economic activity is agriculture. These farmers
are supervised by the Servicios Educativos Promoción y Apoyo Rural
(SEPAR), a non-governmental organization (NGO) from Huancayo
(12°04′S; 75°13′W) that provides support to local institutions for the
sustainable management of rural development (SEPAR, 2017). More
specifically, SEPAR monitors the organic practices that are conducted
by a group of 250 quinoa producers throughout the abovementioned
regions.
Fig. 1. Graphical representation of the system boundary of the organic quinoa production system.
223
A first meeting was programmed with SEPAR at their headquarters
in Huancayo in which their staff provided valuable information regard-
ing the benchmark cultivation conditions (Andrés Vílchez, personal
communication, November 2016). The farmers that belong to the clus-
ter are expected to abide by these recommendations in order to main-
tain the benefits provided by SEPAR. Thereafter, on-site visits to the
farmers were scheduled in which the LCA practitioners were able to
visit 14 farms comprising 63.5 ha of organic quinoa. The production
characteristics of each producer are standardized, although variations
exist in terms of the amount of pesticides and fertilizers used, as these
influence the production target of each producer. Farmers responded
to a set of questions linked to on-site field management activities.
More specifically, farmers provided valuable primary data linked to
yield, machinery use, application of fertilizers and plant protection
agents, provision of seeds and the management of organic residues.
The response rate was 100%, although one farm in Antaparco was finally
excluded since the extremely high amounts of poultry fertilizer re-
ported per hectare rendered the data questionable. A general descrip-
tion of the production areas is shown in Table 1. The specific
production characteristics per producer can be found in Table S1 in
the Supplementary Excel Material (SEM).
All the producers that participated in the study reported using
threshing machines to separate the quinoa seeds from the chaff. How-
ever, only 85% reported using tractors in soil tillage operations. Of
these, all but one reported renting the tractor, whereas one producer
owned a tractor for private use. For all types of machinery used, the pro-
ducers provided the fuel consumption per hour, as well as the number
of hours the machine is needed per hectare. Furthermore, the producers
provided the amount of seeds and the amount of fertilizer and pesticide
used per hectare.
It should be noted that all farmers reported obtaining their seeds
from their own harvest the previous year. In terms of fertilizing agents,
guano de isla and guano de corral were used by farmers to amend their
224 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232

Fig. 2. Geographical location of the quinoa producing regions analyzed in the current study.

fields. Guano de isla was assumed to be extracted from the main guano- residue of a previous production process, it was excluded from the sys-
extraction area in southern Peru and transported to the areas under as- tem boundary, while including transport and on-field emissions after
sessment by road. The NPK content of this fertilizer is 12-11-2. The en- spreading. The residual chaff is also used on-field for fertilization pur-
tire life-cycle was considered for this product. In contrast, guano de poses. Emissions related to on-field application of fertilizers were esti-
corral is poultry manure (see Table 2), with low nitrogen (1.77%) and mated using the IPCC, 2006 method (IPCC, 2006), as shown in Table 2.
phosphorus (0.22%) content. Considering that manure production is a A sensitivity analysis for these emissions is further explained in
Section 2.7.
The intensity of pesticide use in the cultivation of quinoa is relatively
Table 1 low. There are two main reasons for this particularity. On the one hand,
General description of the organic quinoa production areas analyzed in the study.
pests and pathogens are rare above 3400 m asl, allowing farmers to
Unit Ayacucho Huancavelica Total Average plant quinoa with minimal or no interventions (Ruiz et al., 2014). On
Area ha 24.5 32 56.5 4.35 the other hand, saponins are abundant in the quinoa plant, namely in
Number of farms # 6 7 13 – the seed coat, which provides the seeds with insecticidal, antibiotic
Yield kg/ha 1192 1719 – 1490 and fungicidal properties (Gómez-Caravaca et al., 2011; Carlson et al.,
Total production kg 29,200 55,000 84,200 –
2012; Ruiz et al., 2014). The farmers that provided data in the current
Poultry fertilizer kg 12,250 10,150 22,400 –
(guano de isla) study reported using Bacillus spp. as plant growth promoting bacteria
Poultry fertilizer kg 23,125 54,250 77,375 – to reduce disease (Testen and Backman, 2013). More specifically, Bacil-
(guano de corral) lus thuingiensis, an insecticide, is applied in cultivation sites below
Tractor use hrs/ha 4.02 5.53 – 4.66 3000 m asl, whereas it is not necessary above this altitude. The applied
Threshing hours hrs/ha 4.96 5.56 – 5.30
dose ranges from 4.5 to 5.4 kg/ha diluted in water. Similarly, Bacillus
 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232 225

Table 2
Estimation methods used to model on-field fertilization emissions from poultry manure.

Compound IPCC (2006) IPCC (1996) Marquina et al. (2013)

Ammonia (NH3)
Dinitrogen monoxide (N2O)
Nitrate (NO3)
Nitrogen oxides (NOX)
Emission factors were calculated based on the model provided – –
by Nemecek et al. (2007) for solid manure.
N2O = (44/28) ∗ (0.01 (Ntot + Ncr) + 0.01 ∗ (14/17) ∗ NO3 emissions are overestimated A constant factor of 0.78% is
NH3 + 0.0075 ∗ (14/62) ∗ NO3) N from biological N fixation being considered as N2O-N
Where Ncr is N in crop residues (kg N ha−1) is considered
Ntot is total N (kg N ha−1)
The emissions were calculated as follows: NO3-N (kg/ha) ¼ N – –
input (kg/ha) ∗ factor for leached fraction. The factor for leached
fraction was assumed to be 0.1, according to Barry (2011).
An emission factor of 2.6% kg NOx-N/kg N applied was used (EEA, 2016). – –

subtilis, a fungicide, is applied in different doses above 3000 m asl (ca. 2.5. Life cycle inventory
1500 mL/ha) than below that altitude (ca. 2200 mL/ha).
Background processes were added using ecoinvent® v3.3 as the ref- The Life Cycle Inventory (LCI) stage constitutes a critical part of any
erence database (Ecoinvent, 2016). Nevertheless, certain datasets were LCA study, in which the inputs and outputs relevant to the production
adapted to Peruvian conditions as shown in Table 3. system under analysis are put together on the basis of the FU selected.
In this study 18 inventories were realized, one for each of the 13 pro-
2.4. Assumptions and limitations ducers, the average for each region, two dividing the sites based on
their altitude (i.e., higher and middle zone, on the one hand, and
A set of assumptions were modelled in the inventory to account for lower zone, on the other) and the overall average. Table 4 presents
epistemic uncertainties (i.e., data gaps) in the data collection process. the average inventory data of the three aggregated inventories men-
Data provided by the producers regarding total production and produc- tioned above.
tivity of the farm do not consider the reduction by loss after the harvest-
ing process. Hence, the loss factor that was considered was 2.7% for the
processing and packaging process and 4% for the distribution process, 2.6. Life cycle impact assessment and assessment methods
following the loss factors provided by Gustavsson et al. (2011) for
Latin America. Secondly, in the case of the threshing machines, no infor- The software used to compute the Life Cycle Impact Assessment
mation was available regarding the hourly fuel consumption. Therefore, phase, which is when material and energy flows aggregated in the LCI
it was assumed that this consumption rate was equivalent to that of the are translated into environmental impact results through the use of
tractors used for quinoa sowing. characterization factors, was SimaPro 8.3.0 (PRè-Product Ecology
A final limitation was the lack of quantified values measuring the Consultants, 2017). Two different assessment methods were selected
precipitation in the areas assessed, since these cultivation sites rely on to compute the results of a total of thirteen midpoint impact categories.
rainfall as the only source of irrigation. The lack of meteorological sta- On the one hand, IPCC, 2013 was the assessment method selected to es-
tions in the vicinity of the cultivation sites with adequate temporal se- timate GHG emissions, since it represents the most updated method
ries of rainfall led to the decision to exclude water scarcity modelling available (IPCC, 2013). On the other hand, ReCiPe 2008 – Midpoint H
from the study. (Goedkoop et al., 2009) was the assessment method selected to esti-
mate the non-GHG emissions. The use of midpoint impact categories
rather than the use of endpoint categories was justified given that end-
Table 3
point characterization factors present higher uncertainty (Vázquez-
List and description of the main dataset modifications that were performed to model back-
ground processes. Rowe et al., 2015). Furthermore, as a first study linked to the environ-
mental impacts of quinoa production, it was deemed more useful to re-
Dataset Database Action taken
port direct emissions to the environment rather than damages to the
Electricity, high voltage, ecoinvent® The electricity grid for Peru was adapted different areas of protection (i.e., human health, ecosystems and re-
production mix (Peru) 3.3 to that of year 2011 based on the mix sources). In fact, the effect-consequence chain to reach these damages
reported by Vázquez-Rowe et al.
(2015).
is yet to be explored in the particular context of Peru. Finally, a series
Diesel, production ecoinvent® Diesel B5 is the main type of diesel used of impact categories that are present in the ReCiPe method selected
3 in Peru. This brand was modelled based were excluded from the computation of the results. Firstly, the climate
on the data obtained from Avadí et al. change impact category was omitted, since this environmental dimen-
(2014).
sion is already being assessed through the more updated IPCC, 2013
Agricultural machinery, EMEP/EEA Adapted from the EMEP/EEA emission
tractor inventory guidebook 2013 (EEA, 2016). method (Hauschild et al., 2013). Secondly, toxicity categories
Freight transport, ecoinvent® Peru is currently using the Euro III (i.e., human toxicity and terrestrial-, freshwater- and marine
operation 3 emissions standards for emissions of ecotoxicity) were not computed due to the lack of on-site data regard-
new vehicles sold in the country. ing important toxic emissions into the environment, such as active in-
Therefore, Euro III emissions were
assumed for transport emissions.
gredients in permitted pesticides for organic sites for which no
Poultry manure (guano de ecoinvent® The production of poultry manure was characterization factors were available, or the emissions of heavy metals
corral), production 3 taken into consideration considering contained in the organic fertilizers. Thirdly, consumptive and degrada-
the average NPK concentration of guano tive water impact categories were not included.
based on the recommendation of the
Finally, the Cumulative Energy Demand (CED) indicator, which esti-
technicians at the cultivation site.
Guano from Peruvian ecoinvent® The production of poultry manure from mates the energy requirements of the selected production systems
islands (guano de isla), 3 the islands considered the average NPK (VDI-Richtlinien, 1997), was used to calculate a dimensionless indicator
production concentrations of guano recommended named edible protein Energy Return on Investment (ep-EROI). ep-EROI
by the Ministry of Agriculture is computed through the calculation of the coefficient between the pro-
(MINAGRI, 2016).
tein energy output of a selected food product and the energy inputs
226 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232

Table 4 test, named the Mann-Whitney-Wilcoxon, was conducted based on

Life Cycle Inventory for the production of organic quinoa. Data reported per 500 g package the assumption that all the observations from the groups are indepen-
of organic quinoa.
dent of each other (Lorenzo-Toja et al., 2018).
Unit Average Average Average In terms of uncertainty, it is important to note that the data collected
farmer farmer farmer and modelled in this study are prone to different sources of uncertainty.
(Ayacucho) (Huancavelica)
Epistemic uncertainty has been partially identified in Section 2.4 when
Inputs from nature describing some of the data gaps in the foreground system. Similarly,
Water, for pesticide dilution mL 180.5 202.1 169.1
the background system used through the ecoinvent® 3.3 dataset is
Occupation, arable land m2*a 3.36 4.20 2.91
prone to uncertainty, since the unit processes that nourish the fore-
Inputs from technosphere ground system are not specific to the conditions modelled in the
Fuels and machinery
study. Despite the fact that LCA studies tend to have a deterministic ap-
Diesel (Tractor) g 5.14 5.55 5.30
Diesel (Threshing) g 1.77 2.09 1.62 proach to environmental impacts, considering the purpose of
Tractor (use) g 0.391 0.421 0.402 supporting policy-making, it was decided to conduct a stochastic ap-
Thresher (use) g 1.78 2.09 1.62 proach to understand the distribution of results for quinoa production
Fertilizers (Qin and Suh, 2017). For this case study, Monte Carlo Simulation
Guano de Isla mg 133.0 210.6 91.82
(MCS) was applied to the average inventory of the 13 site sample, con-
Guano de corral mg 459.5 396.0 493.2
Pesticides sidering real distribution of data for the foreground system (when ap-
Bacillus subtilis strain QST 713a mL 0.58 0.80 0.45 plicable) and the use of the geometric standard deviations through the
Bacillus thuingiensisb mL 1.03 1.23 0.14 use of the Pedigree Matrix for the remaining data points (Ciroth et al.,
Transport of raw materials
2016). According to the modelling done in SimaPro, approximately
Truck load (16–32 t) – kgkm 105.2 165.9 103.0
fertilization 76% of total unit processes included in the quinoa production dataset
Marine freight – pesticides kgkm 10.39 14.63 7.44 had a predefined standard deviation.
Truck load (16–32 t) – gkm 272.4 383.6 264.8
pesticides 3. Results
Truck load (16–32 t) – plastic gkm 26.91 26.91 36.46
bags
Materials 3.1. Global Warming Potential (GWP) results for the average producer of
Low density polyethylene g 2.14 2.14 2.14 organic quinoa
(LDPE) - production
Transport of final product to
The GHG emissions of the average producer added up to 441 g CO2eq
Lima
Truck load (16–32 t) – organic tkm 115.8 110.1 110.2
per FU (i.e., one 500 g packet of quinoa ready for retailing). Considering
quinoa a total of 13 producers with 56.5 ha, on average, the main carrier of en-
vironmental impact in terms of CC was the emissions of fertilizers
Outputs
500 g package production p 1 1 1
(58.5%), followed by the production of fertilizers (12.6%), machinery
Quinoa in package g 500 500 500 use on-field and for threshing (12.1%), and, to a lesser extent, transport
Emissions from fertilization emissions linked to the delivery of the quinoa to Lima (10.7%) and the
NOx (guano de corral) g 0.453 0.39 0.486 intermediate transport of raw materials (4.9%), as depicted in Fig. 3. Ab-
NO3 (guano de corral) g 3.60 3.10 3.87
solute values per FU can be found in Table S2 in the SEM.
NH3 (guano de corral) g 3.09 2.73 3.28
N2O (guano de corral) g 0.382 0.358 0.394
NOx (guano de isla) mg 0.889 1.4 0.617 3.2. Geographical variation of the global warming potential (GWP) results
NH3 (guano de isla) g 3.98 6.22 2.79
N2O (guano de isla) g 0.525 0.71 0.427
Climate change results, when subdivided per geographical area,
NO3 (guano de isla) g 7.07 11.14 4.9
showed a higher emission of GHGs per FU in the region of Ayacucho,
a
Bacillus subtilis strain QST 713is sold under the commercial name Serenade ASO. which groups 6 of the farms assessed. In this sense, the average emis-
b
Bacillus thuingiensis is sold under the commercial name Xentari WDG.
sion in this area added up to 569 g CO2eq per FU, whereas in the region
of Huancavelica the average value was 30% lower (i.e., 397 g CO2eq per
linked to the production of the product selected – i.e., CED impact cate- FU). Despite this difference, results obtained when the Mann-Whitney-
gory (Tyedmers, 2000; Hall et al., 2009). ep-EROI allows broad compar- Wilcoxon test was conducted suggest that there is no evidence for a sig-
isons of protein sources by calculating a ratio of the edible protein nificant difference between the GHG emissions of the two samples of
energy content of a food product relative to the total industrial energy quinoa producers in the regions of Huancavelica and Ayacucho. Simi-
expended in its production/acquisition (Vázquez-Rowe et al., 2014). larly, differences in environmental impact were identified when com-
paring sites based on their altitudinal location. Cultivation sites located
2.7. Sensitivity, uncertainty and statistical analysis in lower zones, as shown in Fig. 4, presented higher GHG emissions
per FU (565 g CO2eq) as compared to farms located in higher areas
A sensitivity analysis was conducted in order to assess the variation (406 g CO2eq).
of results using different calculation methods for on field N2O emissions.
Besides the use of the IPCC, 2006 method (IPCC, 2006) in the baseline 3.3. Sensitivity and uncertainty analysis
scenario, two additional methods were conducted to estimate these
emissions. On the one hand, the IPCC, 1996 method, which is the previ- The baseline scenario results, using the IPCC, 2006 method to esti-
ous method suggested by IPCC to monitor this type of emissions, was mate on-field N2O fertilization emissions, show an intermediate result
applied (IPCC, 1996). On the other hand, a study performed by in terms of total GHG emissions as compared to the other two methods
Marquina et al. (2013) was used, which monitored on-field emissions computed. The use of the IPCC, 1996 standard increased total emissions
in tropical conditions in South America (see Table 2 for further details). to 570 g CO2eq per FU, whereas the assumption of a 0.78% N2O-N emis-
Additionally, for the statistical analysis, a non-parametric test was sion factor monitored by Marquina et al. (2013) lowered the total GHG
carried out to ascertain if there are any differences in terms of environ- emissions to 267 g CO2eq per FU (see Fig. 5).
mental impact between the two clusters of producers in the regions of The baseline scenario assumed the arrival of quinoa in the city of
Ayacucho and Huancavelica. In this sense, the selected non-parametric Lima for distribution within the city. If the main international export
 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232 227

Fig. 3. Relative GHG emission contributions for the average producer of organic quinoa. Results computed by using the IPCC method (IPCC, 2013).

routes for quinoa are considered, an increase of 6.1% in GHG emissions is
observed if the system boundary is extended to the port of Miami in the
US (i.e., 468 g CO2eq per FU), and 14.7% (506 g CO2eq per FU) when ma-
rine freighted to Amsterdam. For the Monte Carlo analysis, a simulation
was carried out with 1000 iterations for the average producer. Results
show a mean value of 445.2 g CO2eq per FU, with a standard deviation
of ±28.8 g CO2eq considering a 95% confidence interval.
3.4. Non-GHG environmental impacts for the average producer of organic
quinoa
For the remaining impact categories computed in the study, produc-
tion and emissions of fertilizers were the predominant source of envi-
ronmental impact. This was the case, particularly in terms of
terrestrial acidification (TA) and particulate matter formation (PMF),
in which on-site fertilizer emissions accounted for 93% and 80% of the
overall burdens respectively (see Fig. 6). In terms of specific substances,
in the case of TA, ammonia emissions were responsible for 93% of the
total burdens, whereas in PMF, ammonia accounted for 80% of the im-
pact and NOx for 12%. Regardless of PMF, in another air quality category,
photochemical oxidant formation, NOx dominated the environmental
burdens. Fuel production and combustion activities and, to a lesser ex-
tent, on-site fertilizer emissions were mostly responsible for these im-
pacts. ep-EROI results show a relatively high value (37.4%).
4. Discussion and conclusions
4.1. Environmental hotspots in the production of quinoa
Production of organic quinoa in the regions evaluated has proved to
be a relatively traditional process, in which the use of machinery, while
existent, is limited. In this sense, unlike in other agricultural systems
evaluated using LCA, GHG emissions are clearly dominated by the fertil-
izing emissions, with fuels and extraction of raw materials having sub-
stantially lower relative contributions (Vázquez-Rowe et al., 2016).
Cultivation of organic quinoa in this area of Peru is essentially of sub-
sistence nature for farmers. However, despite some years in which the
price of quinoa has not been very competitive (MINAGRI, 2017), in gen-
eral terms the crop has boosted the revenue of many farmers. Moreover,
the fact that these farmers have become integrated in a producer orga-
nization with the support of SEPAR has permitted a reduction in costs
through what is known as a “sharing economy” (Knickel et al., 2018).
An example of this sharing economy is the use of a common platform
to obtain training on agronomic and resource management. Moreover,
current mechanization of the cultivation sites consists of renting out
machinery for the length of time required, optimizing resources. How-
ever, additional revenues through time may imply a will by farmers to
invest in technology to increase efficiency in an attempt to further aug-
ment profits. Regardless of the fact that these actions would increase the

Fig. 4. Total GHG emissions for the geographical and altitudinal division per functional unit (FU; one 500 g package of organic quinoa). Results computed by using the IPCC method (IPCC,
2013).
228 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232

Fig. 5. Total GHG emissions per functional unit (FU; one 500 g package of organic quinoa) based on the selection of on-field fertilization emission methods. Results computed by using the
IPCC method (IPCC, 2013).

reliance on fossil fuels in the production system, presumably increasing
environmental impacts, technological investment may also create debt
for farmers and their families in a rural context in which the use of
banks and familiarization with financial instruments is low (Knickel
et al., 2018). Consequently, any improvement actions suggested must
be integrated in a wider scope of actions in which resilience and the
adaptive capacity of these quinoa-producing communities must be
evaluated with care, integrating social and economic variables.
Despite the fact that the farmers that participated in the study have
established some type of cooperation under the supervision and sup-
port of SEPAR, rural activities in Peru, especially considering that these
farmers are smallholders, are still eminently of subsistence nature. The
results in this study suggest that for organic quinoa production the
level of mechanization is low, which could constitute an opportunity
to maintain a low-carbon agricultural scheme regardless of certain tech-
nological advancements. This would be in line with the strategy that is
currently being adopted by the Common Agricultural Policy (CAP) of
the European Union, through the implementation of “more territorially
and environmentally balanced, climate-friendly, resilient, competitive
and innovative” agricultural practices (EC, 2018). The advantages of
this policy would imply that Peru would align itself with low carbon
and climate resilient agricultural economy standards, as well as with
sustainable consumption and production standards (e.g., product envi-
ronmental footprints) (Finkbeiner, 2014).
Given the low-carbon nature of the production system under analy-
sis, on-field fertilizer emissions dominated impacts linked to GHG emis-
sions. In fact, the emissions of N2O linked to the spreading of organic
fertilizers accounted for 30–70% of the total impact depending on the
method used in its calculation, as shown in Fig. 5. This sensitivity analy-
sis is important given the lack of robust data on on-field fertilizer emis-
sions in Latin America (Marquina et al., 2013). Nevertheless, the
sensitivity analysis conducted demonstrates that, despite the relatively
low GHG emissions in the cultivation of organic quinoa, this stage of
the supply chain is still the most relevant. Moreover, considering the
elongated shelf time of quinoa, it is not plausible to assume that this
product will be air freighted when exported abroad. Therefore, in this
context, we argue that future research should delve into the market dy-
namics of the expected expansion of quinoa production and on the ef-
fects that this expansion could have in terms of land use changes
(LUCs) in Andean regions.

Fig. 6. Relative non-GHG environmental impacts for the average producer. Results computed by using the IPCC method (IPCC, 2013). OD = ozone depletion; TA = terrestrial acidification;
POF = photochemical oxidant formation; PMF = particulate matter formation; MD = metal depletion; FD = fossil depletion.
 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232 229

Table 5
Comparative analysis in terms of GHG emissions between organic quinoa and key protein-rich food products in the Peruvian diet. Data source: ENAPREF (2012) and Vázquez-Rowe et al.
(2017a).

Product Protein per kg of product GHG emissions per kg of product GHG emissions per g of protein Mean consumption per capita
(g) (kg CO2eq) (g CO2eq) (kg/year)

Quinoa (organic) 136 0.88 6.47 0.9a

Beef 213 28.73 134.88 5.6
Fresh cow cheese 175 8.86 50.63 2.4
Fresh cow's milk 31 1.39 44.84 3.0
Pork 144 5.85 40.63 1.0
Rice 78 2.66 34.10 46.5
Tuna 229 5.9 25.76 0.9
Chicken meat 214 4.12 19.25 23.3
Fish 234 4.41 18.85 7.0
a
This value includes the consumption of quinoa, quilete and cañihua.

4.2. Comparison with other food production systems
When compared with other food products, namely those of cereal
nature or with high protein content, quinoa ranks as one of the food
products with the lowest GHG emissions per unit of protein, as shown
in Table 5. In this sense, 6.47 g CO2eq are generated for the production
of 1 g of protein contained in quinoa, whereas in the case of rice the
value rises to 34.1 g CO2eq. Similarly, when compared to animal-based
products, chicken meat presents the most attractive low-carbon option
(i.e., 18.9 g CO2eq), but still substantially higher than that of quinoa. If a
different indicator is considered such as ep-EROI, similar results are
identified. For instance, the calculated ep-EROI for quinoa in this study
(i.e., 37.4%) is higher than that for chicken (25%), swine (3–7%) or
tuna – 12% (Pimentel and Pimentel, 2003;Parker et al., 2015).
A recent study by Vázquez-Rowe et al. (2017a) analyzed the GHG
emissions linked to the average Peruvian diet. The study suggests that
annual per capita emissions on average account for 1.08 t CO2eq, rang-
ing from 966 kg CO2eq to 1.77 t CO2eq depending on the city assessed.
Moreover, in household consumption, red meat and poultry accounted,
on average, for ca. 48% of total GHG emissions. In this sense, Table 6 pre-
sents the reductions in GHG emissions that would be attained if 20% of
protein intake from selected protein-rich food products were
substituted by the consumption of the same amount of protein from
quinoa. Results show that, for instance, in the case of beef, 30.9 kg
CO2eq per capita would be reduced annually. On a national level this
would imply a reduction of 973 t CO2eq per year. These results should
be used cautiously, since, as mentioned above, higher quinoa consump-
tion would imply significant LUCs. Nevertheless, it is feasible to presume
that cattle ranching activities (in Peru or abroad) linked to an increase in
demand for red meat in the emerging middle class will translate into
higher GHG emissions than LUCs triggered by the expansion of the agri-
cultural frontier for quinoa production (Bustamante et al., 2012; Cohn
et al., 2014; Vázquez-Rowe et al., 2017a). Having said this, when consid-
ering the amount of agricultural land destined to quinoa cultivation in
the regions of Ayacucho and Huancavelica, it is observed that quinoa
only represents 2.4% and 0.1%, respectively, whereas fallow and set-
aside land account for 14% and 8% in these regions, suggesting that
quinoa expansion could still be attained without compromising LUCs
(Graesser et al., 2015).
4.3. Quinoa and food security
Quinoa is commonly identified with cereals due to its texture and
plant-based origin. However, its high protein content, as well as the
fact that it contains all protinogenic amino acids, makes it an attractive
product to compete with animal protein (Vega-Gálvez et al., 2010;
Ruiz et al., 2014). However, despite the fact that quinoa originates
from the Peruvian highlands, its consumption in the average Peruvian
diet, considering the latest available data from the Peruvian Statistics In-
stitute (INEI), is low (ENAPREF, 2012). According to the survey con-
ducted in 2008 and 2009 in over 36,000 households across the nation,
rice is the most consumed cereal in Peru (47.4 kg per person and
year), whereas quinoa is included in the group “other cereals”, together
with kiwicha or quilete (Amaranthus caudatus) and cañihua
(Chenopodium pallidicaule). This group only adds up to 0.6 kg/year per
capita on a national average, with slightly higher numbers in rural
areas (1.1 kg/y) and in Andean cities (1.2 kg/y). There are reasons to hy-
pothesize that in future surveys the amount of quinoa per capita will
trend upwards for a number of reasons, including higher national pro-
duction, although most is exported, and an increasing perception, also
domestically, of the nutritional benefits of consuming this product.
Regardless of the nutritional and dietary benefits of consuming qui-
noa on a regular basis, an additional issue regarding food security is
worth highlighting. In this sense, when the average Peruvian diet is an-
alyzed, it is evident that Peruvian households rely on very few food
products, in a diet that is animal-based, consisting mainly of poultry,
and with a high consumption of staples and cereals (ENAPREF, 2012;
Vázquez-Rowe et al., 2017a). From a health perspective this diet lacks
sufficient amounts of fruits and vegetables, so one may argue that qui-
noa is not a critical product to introduce in the diet, unless its benefits
justify substituting some other products from a social, environmental
or economic point of view.
From a food security perspective, however, it is important to note
that Peru is a country that is extremely vulnerable to natural disasters,

Table 6
GHG emission mitigation potential of substituting 20% of protein intake from selected protein-rich food products when substituted by the consumption of the same amount of protein from
quinoa.

Product Mean annual consumption per capita GHG emissions per capita Total protein content Per capita variation
considering a 20% of protein
replacement by quinoa

(kg/year) (kg CO2eq/year) (g/year) kg CO2eq/year %

Beef 5.6 160.89 1192.8 30.63 19.0

Fresh cow cheese 2.4 21.26 420 3.71 17.4
Pork 1.0 5.85 144 0.98 16.8
Chicken meat 23.3 3627 4986 12.75 13.3
Rice 46.5 123.69 3627 20.04 16.0
Fish 7.0 30.87 1638 4.05 13.1
230 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232
Fig. 7. Variation of the wholesale monthly market price of selected food basket products in Lima between July 2016 and June 2017. a) Price per kilogram of product; b) Price per gram of
protein.
including the El Niño-Southern Oscillation (ENSO) phenomenon, as
well as very intense earthquakes (Mitchell, 1999). For instance, the
city of Lima, with a population close to 11 million people, experienced
a week without a centralized potable water supply in March 2017 due
to the mudslides occurring along the Peruvian coast in that period
(Vázquez-Rowe et al., 2017b). This event tested the resilience of
Lima's infrastructure, showing that the redundancy of its waterworks
and other centralized public services (e.g., electricity or gas) is low.
Fig. 7a shows the evolution of the wholesale monthly price of three
selected common food products (i.e., chicken meat, lemon and beef) in
the Peruvian food basket and quinoa. Results show that the mudslides
occurring in March 2017 did not affect meat or quinoa, but created a
strong effect on the price of lemon in the market, mainly attributable
to the fact that lemon, produced in the northern coast, was not able to
reach Lima due to agricultural land and infrastructure damage
(Vázquez-Rowe et al., 2017b). Fig. 7b, in contrast, shows the cost in
Peruvian soles (PEN) of 1 g of protein obtained from quinoa as com-
pared to beef and chicken meat, showing that the market price is com-
petitive to that of chicken meat and substantially cheaper than beef.
Consequently, it is hypothesized that in the case of a major seismic
event, the city of Lima could be badly hit in terms of its infrastructure,
including extended periods without electricity (PREDES, 2009), which
could lead to food scarcity. Given this scenario, the consumption of per-
ishable food products (e.g., beef, fruit, poultry…) would become chal-
lenging. This usually leads to the distribution of cereal- and staple-
based products, as well as canned products (e.g., tuna) in humanitarian
crises (Maxwell et al., 2010). However, we argue that quinoa could play
a key role in terms of humanitarian response in Peru considering its
high protein content, the relatively low cost if compared to other high
protein products and the fact that it is a long-term, non-perishable
foodstuff.
5. Conclusions
As far as the researchers were able to ascertain, this study constitutes
the first full LCA on organic quinoa production, packaging and distribu-
tion. It is expected that the results obtained will constitute an important
benchmark for the scientific community working in the field of environ-
mental sustainability of agriculture, food diets or nutrition, considering
the current increase in global demand for this Andean food product. The
results show that organic quinoa presents environmental impacts that
are in a similar range to other agricultural products, such as wheat or
rice, although in general terms quinoa impacts appear on the upper
side of the range. Nevertheless, if the environmental impact results are
compared to other food products rich in protein, it is observed that qui-
noa presents an outstanding environmental profile when an attribu-
tional LCA perspective is applied. In fact, considerable GHG emission
reductions could be attained in the average Peruvian diet if quinoa
were to partially substitute other protein rich products, such as red
meat or seafood products. However, it must be considered that the pro-
duction of quinoa worldwide is concentrated in small areas of Peru and
Bolivia, with a limited production volume. If [organic] quinoa were to
expand throughout other agricultural areas, further research should
focus on how that would affect the market, as well as the environmental
consequences of this expansion in terms of LUCs. It is feasible to hypoth-
esize that quinoa production expansion at high altitudes would not
have major effects in terms of GHG emissions due to the limited carbon
stock in Andean areas, but this modelling would allow understanding
the trade-offs linked to indirect LUCs due to the displacement of other
crops (e.g., barley, wheat…) elsewhere.
The farmers that provided data for this study are smallholder subsis-
tence farmers with a low degree of mechanization in their agronomic
operations. Moreover, they practice a “sharing economy” in terms of
hiring machinery and obtaining training on agronomic and resource
management. Hence, future efforts should focus on transitioning from
subsistence to commercial agriculture without abandoning low carbon
practices. In fact, this strategy, which could be implemented in a greater
part of the Peruvian Andes, would be aligned with the current climate
resilient agricultural economy standards of the European Union, gaining
competitiveness in that market.
Acknowledgements
The authors wish to express their gratitude to Jorge Perea Valencia
and Andrés Vílchez Melo, Executive Director and Director of Develop-
ment and Regional Competitiveness, respectively, of Servicios Educativos
de Promoción y Apoyo Rural (SEPAR) for facilitating data acquisition in
the production areas. The authors also thank the Instituto de Ciencias
de la Naturaleza, Territorio y Energías Renovables (INTE) at the Pontificia
Universidad Católica del Perú for financial support. Dr. Pedro
Villanueva-Rey and BSc Gustavo Larrea-Gallegos are acknowledged for
valuable scientific exchange.
Appendix A. Supplementary Excel Material
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2018.05.029.
References
Aguilar, P.C., Jacobsen, S.E., 2003. Cultivation of quinoa on the Peruvian Altiplano. Food
Rev. Int. 19 (1–2), 31–41.
Avadí, Á., Vázquez-Rowe, I., Fréon, P., 2014. Eco-efficiency assessment of the Peruvian an-
choveta steel and wooden fleets using the LCA+ DEA framework. J. Clean. Prod. 70,
118–131.
 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232
Barry, M.T., 2011. Life Cycle Assessment and the New Zealand Wine Industry: A Tool to
Support Continuous Environmental Improvement. M. Sc. Dissertation. Massey Uni-
versity, Wellington, NZ.
Bazile, D., Martinez, E.A., Fuentes, F., Chia, E., Namdar-Irani, M., Olguin, P., Saa, C., Thomet,
M., Vidal, A., 2015. In: Bazile, D., Bertero, H.D., Nieto, C. (Eds.), State of the Art Report
on Quinoa Around the World in 2013. FAO, Rome, pp. 401–421.
Bustamante, M.M., Nobre, C.A., Smeraldi, R., Aguiar, A.P., Barioni, L.G., Ferreira, L.G., Longo,
K., May, P., Pinto, A.S., Ometto, J.P., 2012. Estimating greenhouse gas emissions from
cattle raising in Brazil. Clim. Chang. 115 (3–4), 559–577.
Carlson, D., Fernández, J.A., Poulsen, H.D., Nielsen, B., Jacobsen, S.E., 2012. Effects of quinoa
hull meal on piglet performance and intestinal epithelial physiology. J. Anim. Physiol.
Anim. Nutr. 96, 198–205.
Ciroth, A., Muller, S., Weidema, B., Lesage, P., 2016. Empirically based uncertainty factors
for the pedrigree matrix in ecoinvent. Int. J. Life Cycle Assess. 21, 1338–1348.
Cohn, A.S., Mosnier, A., Havlík, P., Valin, H., Herrero, M., Schmid, E., O'Hare, M.,
Obersteiner, M., 2014. Cattle ranching intensification in Brazil can reduce global
greenhouse gas emissions by sparing land from deforestation. Proc. Natl. Acad. Sci.
111 (20), 7236–7241.
EC, 2018. Rural development 2014–2020: EU legislation. European Commission. Re-
trieved from:. https://ec.europa.eu/agriculture/rural-development-2014-2020/legis-
lation_en (Latest access: January 19th 2018).
Ecoinvent, 2016. Ecoinvent database. Available at. http://www.ecoinvent.org/.
EEA, 2016. EMEP/EEA Emission Inventory Guidebook 2016. European Environment
Agency Retrieved from:. https://www.eea.europa.eu/themes/air/emep-eea-air-
pollutantemission-inventory-guidebook/emep (Latest access: November 21st 2017).
El Pais, 2016. La quinua se aleja de los Andes. Retrieved from:. http://elpais.com/elpais/
2016/03/10/estilo/1457648149_520065.html (Published March 10th 2016, in
Spanish).
El Peruano, 2008. Ley N°29196. Ley de promoción de la producción orgánica y ecológica.
Retrieved from:. https://www.senasa.gob.pe/senasa/normas-sobre-produccion-
organica/ (Last accessed: March 18th 2018).
ENAPREF, 2012. Perú: Consumo per cápita de los principales alimentos 2008–2009.
Encuesta Nacional de Presupuestos Familiares (ENAPREF). Dirección Técnica de
Demografía e Indicadores Sociales. Instituto Nacional de Estadística e Informática
(INEI). May, 2012 (in Spanish).
Fairlie-Reinoso, A., 2016. La quinua en el Perú: cadena exportadora y políticas de gestión
ambiental. First edition. INTE-PUCP, Lima (86pp).
FAO, 2016. FAOSTAT. Statistics Division. Food and Agriculture Organization of the United
Nations Retrieved from:. http://www.fao.org/faostat/en/#home (Last accessed: Janu-
ary 9th 2017).
Finkbeiner, M., 2014. Product environmental footprint—breakthrough or breakdown for
policy implementation of life cycle assessment? Int. J. Life Cycle Assess. 19, 266–271.
Gestión, 2016. Perú se consolidó como primer exportador de quinua a nivel mundial. Re-
trieved from:. http://gestion.pe/economia/peru-se-consolido-como-primer-
exportador-quinua-nivel-mundial-2156129 (Last accessed: April 20th 2018).
Goedkoop, M., Heijungs, R., Huijbregts, M., de Schryver, A., Struijs, J., van Zelm, R., 2009.
ReCiPe 2008. A Life Cycle Impact Assessment Method Which Comprises Harmonised
Category Indicators at the Midpoint and the Endpoint Level. Report I: Characterisa-
tion. Ministry of Housing, Spatial Planning and Environment (VROM) Retrieved
from:. www.lcia-recipe.info (last accessed: 19 May 2015).
Gómez-Caravaca, A.M., Segura-Carretero, A., Fernández-Gutiérrez, A., Caboni, A.F., 2011.
Simultaneous determination of phenolic compounds and saponins in quinoa
(Chenopodium quinoa Willd.) by a liquid chromatography–diode array detection–
electrospray–ionization-time-of-flight mass spectrometry methodology. J. Agric.
Food Chem. 59, 10815–10825.
Graesser, J., Aide, T.M., Grau, H.R., Ramankutty, N., 2015. Cropland/pastureland dynamics
and the slowdown of deforestation in Latin America. Environ. Res. Lett. 10 (3),
034017.
Gustavsson, J., Cederberg, C., Sonesson, U., Van Otterdijk, R., Meybeck, A., 2011. Global
Food Losses and Food Waste. Food and Agriculture Organization of the United Na-
tions, Rome.
Hall, C.A.S., Balogh, S., Murphy, D.J.R., 2009. What is the minimum EROI that a sustainable
society must have? Energies 2, 25–47.
Hauschild, M.Z., Goedkoop, M., Guinée, J., Heijungs, R., Huijbregts, M., Jolliet, O., Margni,
M., De Schryver, A., Humbert, S., Laurent, A., Sala, S., Pant, R., 2013. Identifying best
existing practice for characterization modeling in life cycle impact assessment. Int.
J. Life Cycle Assess. 18 (3), 683–697.
Higuchi, A., 2015. Characteristics of consumers of organic products and the increase in the
supply of these products in metropolitan Lima, Peru. Apuntes 42 (77), 57.
IPCC, 1996. Reference manual. IPCC Guidelines for National Greenhouse Gas Inventories.
Vol. 3. UK Meteorological Office, Bracknell (Revised 1996).
IPCC, 2006. Chapter 11: N2O Emissions from Managed Soils, and CO2 Emissions from
Lime and Urea Application. Agriculture, Forestry and Other Land Use. 2006 IPCC
Guidelines for National Greenhouse Gas Inventories Vol. 4 Retrieved from:. http://
www.ipccnggip.iges.or.jp/public/2006gl/pdf/4_Volume4/V4_11_Ch11_N2O&CO2.pdf
(Latest access: May 18th 2016).
IPCC, 2013. Climate Change 2013. The Physical Science Basis. Working Group I Contribu-
tion to the 5th Assessment Report of the IPCC. Intergovernamental Panel on Climate
Change Retrieved from:. http://www.climatechange2013.org (Latest access: June
30th 2016).
ISO, 2006a. ISO 14040. Environmental Management e Life Cycle Assessment - Principles
and Framework. International Organization for Standardization.
ISO, 2006b. ISO 14044. Environmental Management e Life Cycle Assessment - Require-
ments and Guidelines. International Organization for Standardization.
Jacobsen, S.E., 2003. The worldwide potential for quinoa (Chenopodium quinoa Willd.).
Food Rev. Int. 19, 167–177.
231
Knickel, K., Redman, M., Darnhofer, I., Ashkenazy, A., Calvão Chebach, T., Sūmane, S.,
Tisenkopfs, T., Zemeckis, R., Atkociuniene, V., Rivera, M., Strauss, A., Kristensen, L.S.,
Schiller, S., Koopmans, M.E., Rogge, E., 2018. Between aspirations and reality: making
farming, food systems and rural areas more resilient, sustainable and equitable.
J. Rural. Stud. https://doi.org/10.1016/j.jrurstud.2017.04.012.
Lee, H.J., Hwang, J., 2016. The driving role of consumers' perceived credence attributes in
organic food purchase decisions: a comparison of two groups of consumers. Food
Qual. Prefer. 54, 141–151.
León, H., 2003. Cultivo de la quinua en Puno–Perú. Descripción, manejo y producción.
Universidad Nacional del Altillano-Facultad de Ciencias Agrarias (17, in Spanish).
Lorenzo-Toja, Y., Vázquez-Rowe, I., Marín-Navarro, D., Crujeiras, R.M., Moreira, M.T.,
Feijoo, G., 2018. Dynamic environmental efficiency assessment for wastewater treat-
ment plants. Int. J. Life Cycle Assess. 23, 357–367.
Marquina, S., Donoso, L., Perez, T., Gil, J., Sanhueza, E., 2013. Losses of NO and N2O emis-
sions from Venezuelan and other worldwide tropical N-fertilized soils. J. Geophys.
Res. Biogeosci. 118, 1094e1104.
Maxwell, D., Webb, P., Coates, J., Wirth, J., 2010. Fit for purpose? Rethinking food security
responses in protracted humanitarian crises. Food Policy 35 (2), 91–97.
MINAGRI, 2016. El guano de las islas. Propiedades y Usos. Retrieved from:. http://minagri.
gob.pe/portal/download/pdf/novedades/presentacion-agrorural.pdf (Last accessed:
July 1st 2016, in Spanish).
MINAGRI, 2017. La Quinua: Producción y Comercio del Perú. Dirección General de
Políticas Agrarias. Dirección de Estudios Económicos e Información Agraria.
Ministerio de Agricultura y Riego. Boletín. Perfil Técnico N°2 (in Spanish).
Mitchell, J.K., 1999. Megacities and natural disasters: a comparative analysis. GeoJournal
42 (2), 137–142.
Morlon, P., 1982. Valorización de La diversidad ecológica. In Tecnologías Agrícolas
Tradicionales de los Andes Centrales Perspectivas Para El Desarrollo Lima, Perú:
COFIDE, PNUD y. UNESCO (in Spanish).
Navruz-Varli, S., Sanlier, N., 2016. Nutritional and health benefits of quinoa (Chenopodium
quinoa Willd.). J. Cereal Sci. 69, 371–376.
Nemecek, T., Kägi, T., Blaser, S., 2007. Life cycle inventories of agricultural production sys-
tems. Final Report Ecoinvent. Vol. 2 (0 No, 15).
Padel, S., Foster, C., 2005. Exploring the gap between attitudes and behaviour: under-
standing why consumers buy or do not buy organic food. Br. Food J. 107 (8),
606–625.
Parker, R.W., Vázquez-Rowe, I., Tyedmers, P.H., 2015. Fuel performance and carbon foot-
print of the global purse seine tuna fleet. J. Clean. Prod. 103, 517–524.
Pimentel, D., Pimentel, M., 2003. Sustainability of meat-based and plant-based diets and
the environment. Am. J. Clin. Nutr. 78, 660–663.
PREDES, 2009. Diseño de escenario sobre el impacto de un sismo de gran magnitud en
Lima Metropolitana y Callao. Centro de Estudios y Prevención de Desastres – PREDES.
Retrieved from:. http://www.predes.org.pe/predes/images/dis_esc_lima.pdf (Latest
access: January 18th 2018).
PRè-Product Ecology Consultants, 2017. SimaPro 8.3. PRè Consultants, The Netherlands.
Qin, Y., Suh, S., 2017. What distribution function do life cycle inventories follow? Int. J. Life
Cycle Assess. 22, 1138–1145.
Quinoa d'Anjou, 2017. . Retrieved from:. http://www.quinoadanjou.fr (Last accessed: Jan-
uary 15th 2017).
Ruales, J., Nair, B.M., 1994. Properties of starch and dietary fibre in raw and proc-
essed quinoa (Chenopodium quinoa, Willd.) seeds. Plant Foods Hum. Nutr. 45,
223–246.
Ruiz, K.B., Biondi, S., Oses, R., Acuña-Rodríguez, I.S., Antognoni, F., Martinez-Mosqueira,
E.A., et al., 2014. Quinoa biodiversity and sustainability for food security under cli-
mate change. A review. Agron. Sustain. Dev. 34 (2), 349–359.
SENAMHI, 2009. Escenarios Climáticos en el Perú para el año 2030. Segunda
comunicación nacional de cambio climático. Resumen Técnico. Ministry of the Envi-
ronment, Peru Retrieved from:. http://redpeia.minam.gob.pe/admin/files/item/
4d77e7ad5bb27_Resumen_Escenarios_climaticos_del_Peru.pdf (in Spanish).
SEPAR, 2017. Servicios Educativos Promoción y Apoyo Rural. Retrieved from:. http://
separ.org.pe (Latest access: September 8th 2017).
Smith-Spangler, C., Brandeau, M.L., Hunter, G.E., Bavinger, J.C., Pearson, M., Eschbach, P.J.,
Sundaram, V., Liu, H., Schirmer, P., Stave, C., Olkin, I., Bravata, D.M., 2012. Are organic
foods safer or healthier than conventional alternatives? A systematic review. Ann. In-
tern. Med. 157 (5), 348–366.
Testen, A.L., Backman, P.A., 2013. Plant Growth Promoting Characteristics of Bacillus Spe-
cies Associated with Chenopodium quinoa (Sustainable Agriculture and Natural Re-
source Management).
Tyedmers, P., 2000. Salmon and Sustainability: The Biophysical Cost of Producing Salmon
through the Commercial Salmon Fishery and the Intensive Salmon Culture Industry.
PhD Dissertation. University of British Columbia, Vancouver.
Vázquez-Rowe, I., Villanueva-Rey, P., Moreira, M.T., Feijoo, G., 2014. Edible protein energy
return on investment ratio (ep-EROI) for Spanish seafood products. Ambio 43 (3),
381–394.
Vázquez-Rowe, I., Golkowska, K., Lebuf, V., Vaneeckhaute, C., Michels, E., Meers, E.,
Benetto, E., Koster, D., 2015. Environmental assessment of digestate treatment tech-
nologies using LCA methodology. Waste Manag. 43, 442–459.
Vázquez-Rowe, I., Kahhat, R., Quispe, I., Bentín, M., 2016. Environmental profile of green
asparagus production in a hyper-arid zone in coastal Peru. J. Clean. Prod. 112,
2505–2517.
Vázquez-Rowe, I., Larrea-Gallegos, G., Villanueva-Rey, P., Gilardino, A., 2017a. Climate
change mitigation opportunities based on carbon footprint estimates of dietary pat-
terns in Peru. PLoS One 12 (11), e0188182.
Vázquez-Rowe, I., Kahhat, R., Lorenzo-Toja, Y., 2017b. Natural disasters and climate
change call for the urgent decentralization of urban water systems. Sci. Total Environ.
605, 246–250.
232 E. Cancino-Espinoza et al. / Science of the Total Environment 637–638 (2018) 221–232

VDI-Richtlinien, 1997. Cumulative energy demand: Terms, definitions, methods of calcu-
lation. VDI-Richtlinien, Düsseldorf.
Vega-Gálvez, A., Miranda, M., Vergara, J., Uribe, E., Puente, L., Martínez, E.A., 2010. Nutri-
tion facts and functional potential of quinoa (Chenopodium quinoa willd.), an ancient
Andean grain: a review. J. Sci. Food Agric. 90 (15), 2541–2547.
Willer, H., Lernoud, J., 2014. The world of organic agriculture: statistics and emerging
trends 2014. Research Institute of Organic Agriculture (FiBL). Frick/Ifoam, Bonn.
Williams, C.M., 2002. Nutritional quality of organic food: shades of grey or shades of
green? Proc. Nutr. Soc. 61 (1), 19–24.