179

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

Quantifying Cardinal Temperatures in Quinoa

(Chenopodium quinoa) Cultivars

Abstract — González, Juan A.; Sebastián E. Buedo; Marcela Bruno; Fernando E. Prado.

2017. “Quantifying Cardinal Temperaturas in Quinoa (Chenopodium quinoa) Cultivars”. Lilloa

54 (2). Seed germination and plant growth are affected by temperature. This study was

conducted to evaluate the effect of temperature on seed germination of ten quinoa cultivars

under a temperature gradient between 8 ºC and 50 ºC. The time course of germination was

fitted using a logistic function. Aborted seeds and non-germinated seeds were also analyzed

as function of temperature gradient. Cardinal temperatures were estimated by regression of

the inverse time to 50% germination (germination rate) against temperature gradient. The

minimum (T

min

), optimum (T

opt

) and maximum (T

max

) temperatures for seed germination were

determined using both linear (bilinear model) and polynomial (quadratic and cubic models)

regressions. Based on T

opt

estimated from bilinear and cubic models, quinoa seeds can be

grouped into two subgroups: one represented by Kancolla, Chucapaca, Kamiri, Robura and

Sajama cultivars with values of T

opt

£ 33 ºC, and other represented by CICA, Sayaña, Amilda,

Ratuqui and Samaranti with T

opt

values  33 ºC, respectively. Percentages of maximum cumu-

lative germination calculated from the quadratic model were closely similar to those obtained

in germination trials.

Keywords: Germination, germination models, abnormal germination, optimum temperatures.

Resumen — González, Juan A.; Sebastián E. Buedo; Marcela Bruno; Fernando E. Prado.

2017. “Cuantificación de las temperaturas cardinales en cultivares de quinoa (Chenopodium

quinoa)”. Lilloa 54 (2). La germinación de las semillas y el crecimiento de las plantas son afec-

tados por la temperatura. Este estudio fue diseñado para evaluar el efecto de la temperatura

sobre la germinación de diez variedades de quinoa bajo un gradiente entre 8 ºC y 50 ºC. El

curso de la germinación en función del tiempo fue ajustado utilizando la función logística. Las

semillas abortadas y no germinadas también fueron analizadas en el gradiente de temperatu-

ra. Las temperaturas cardinales fueron estimadas por regresión de la inversa del tiempo de

germinación al 50 % (velocidad de germinación) contra el gradiente de temperatura. La tempe-

ratura mínima (T

min

), la óptima (T

opt

) y la máxima (T

max

) de germinación fueron determinadas

utilizando regresiones lineales (modelo bilineal) y polinómicos (modelos cuadráticos y cúbicos).

Basados en la T

opt

estimada a partir de los modelos bilineales y cúbicos las variedades de

quinoa estudiadas pueden ser divididas en dos subgrupos: uno representado por Kancolla,

Chucapaca, Kamira. Robura y Sajama con un valor de T

opt

de £ 33 ºC, y otro representado

por CICA, Sayaña, Amilda, Ratuqui y Samaranti con una T

opt

 33 ºC respectivamente. Los

porcentajes de germinación máxima obtenida a partir del modelo cuadrático utilizada fueron

muy cercanos a aquellos obtenidos en las pruebas de germinación.

Palabras claves: Germinación, modelos de germinación, germinación anormal, temperatura

óptima.

Recibido: 27/12/16 – Aceptado: 07/08/17

Cuantificación de las temperturas cardinales en cultivares

de quinoa (Chenopodium quinoa)

González, Juan A.

; Sebastián E. Buedo

; Marcela Bruno

;

Fernando E. Prado

Instituto de Ecología, Fundación Miguel Lillo, Miguel Lillo 251, (T4000JFE) San Miguel de Tucumán,

Argentina.

Facultad de Ciencias Naturales e IML, Cátedra de Fisiología Vegetal, Universidad Nacional de Tucumán,

Miguel Lillo 205, (4000) Tucumán, Argentina.

* Autor corresponsal: jagonzález@lillo.org.ar

180

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

INTROD UCTIO N

Seed germination is affected by different

environmental factors, being the tempera-

ture one of the most important to reach the

successful germination (Bewley and Black,

1994). Temperature influences seed germi-

nation through its impact on both physico-

chemical processes (e.g. water uptake) and

metabolic processes (e.g. enzyme-catalyzed

reactions). At low temperatures enzymes are

unable to adopt the active conformational

state to catalyze reactions, while at high tem-

peratures they are precipitated and cannot

catalyze reactions (Dixon and Webb, 1979).

Germination responses of a given seed frac-

tion to temperature can be characterized in

terms of cardinal temperatures: minimum

or base temperature (T

min

), defined as the

temperature below which germination rate is

zero, optimum temperature (T

opt

), defined as

the temperature at which germination rate is

maximal, and maximum temperature (T

max

defined as the temperature above which ger-

mination rate is zero (Yan and Hunt, 1999).

Cardinal temperatures are species-specific or

cultivar-specific and even species origin- or

cultivar origin-specific (Kamkar et al., 2012).

Hence cardinal temperatures can be consid-

ered as important factors to predict germi-

nation performance of new cultivars and/or

provenances of a determined crop in a par-

ticular environment.

Quinoa (Chenopodium quinoa Willd.) has

high tolerance to extreme conditions such

as salinity, drought, low temperatures and

high solar radiation (Risi and Galwey, 1984;

González et al., 2015). Its grain contains

high quality protein with a good provision of

minerals (Prado et al., 2014) and better ami-

no acid profile (González et al., 2011) than

other crops such as wheat, rice, barley and

maize (Abugoch James, 2009; Vega-Gálvez

et al., 2010). According to FAO nutritional

reference values, quinoa is considered today

as a promissory alternative crop, especially

for marginal regions where the growth of

traditional cereals and legumes are limited

(Choukr-Allah et al., 2016). In South Amer-

ica the growth range of quinoa extends from

sea level in Chile and lowland regions of Peru

to over 3800 m altitude in highland regions

of Bolivia and Ecuador (Bazile et al., 2013).

Depending on genotype characteristics and

phenological growth stage, quinoa can toler-

ate a wide range of temperature (from –8 °C

to 38 °C) and relative humidity conditions

(from 40% to 88%) (Jacobsen et al., 2005;

Bazile et al., 2013). Seed germination and

seedling development of quinoa cultivars are

influenced by environmental conditions, and

are strongly dependent on temperature (Ja-

cobsen and Bach, 1998; Bertero, 2001; Bois

et al., 2006). González and Prado (1992)

found that low germination percentages

of seeds of the Sajama cultivars could be

explained by an interactive effect between

low temperature and soil salinity. In addi-

tion, Jacobsen and Bach (1998) reported

that maximum germination percentage of

the Olav cultivar occurs between 30 °C and

35 °C. This temperature range is significantly

higher than the temperature range recorded

in the Bolivian Altiplano during quinoa seed-

time (Bois et al., 2006). Beyond the body

of knowledge on quinoa physiology, the ef-

fect of temperature on germination of major

commercial cultivars still remains unclear.

Germination response tends to be linear in

a limited range of temperatures. In most

cultivated species the linear range of tem-

peratures varies between 10 and 30 °C (Bon-

homme, 2000), but at extreme thresholds of

both low and high germination temperatures

the germinative response tends to be curvi-

linear (Hardegree, 2006). In this way, the

estimation of cardinal temperatures can be

made through linear and curvilinear models

as long as limitations intrinsic to each model

are recognized. The aim of this work was to

analyze the effect of constant temperatures,

on germinative traits of ten quinoa (Cheno-

podium quinoa Willd.) cultivars in order to

determine: i) differences in cardinal tem-

peratures determined by curvilinear models

(second-and third-order regression equa-

tions) and bilinear model (linear regression

equation), and ii) intracultivar variability in

the percentage of germinated seeds, aborted

seeds and non-germinated seeds at different

181

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

temperatures. Alternative temperatures ex-

periment was not considered according the

results obtained by Strenske et al., (2017).

Because the estimation of cardinal tempera-

ture is highly dependent on the statistical

model, bilinear and curvilinear regressions

(White et al., 2015) were used.

MATERI A LS A ND M E THODS

Seed material

Seed of 10 quinoa genotypes (Amilda,

Chucapaca, CICA, Kamiri, Ratuqui, Robura,

Sajama, Samaranti and Sayaña), representa-

tive of the most common available commer-

cial cultivars, were evaluated in this study.

Seeds were collected from plants grown dur-

ing the 2014-2015 growing season at En-

calilla Experimental Station (Amaicha del

Valle, Tucumán, Argentina, 22°31’S latitude,

65°59’W longitude and 1980 m asl). After

harvest, the seeds were bulked, cleaned

manually, placed in paper bags, and stored

at 6 °C until the beginning of experiments.

ermination teStS

Seeds were surface-sterilized in a 2% so-

dium hypochlorite solution for 2 min, washed

twice with distilled water and dried on filter

paper. Sterilized seeds were placed on filter

paper in 5 cm glass Petri dishes (50 seeds per

dish). Petri dishes were added with 2 mL of

distilled water and transferred to tempera-

ture-controlled cabinet under darkness condi-

tion and relative humidity of 70%. Selected

temperatures were 8 °C, 15 °C, 20 °C, 25 °C,

30 °C, 35 °C, 40 °C, 45 °C, and 50 °C.

In lab, germinated seeds (radicle protru-

sion  2 mm in length) were counted every

2 h during a 24-h period and cumulative

germination percentage was plotted against

time. We chose 24 h as maximum germina-

tion period because under laboratory condi-

tions the total germination process of qui-

noa seeds occurs in a very short time period

(~12-14 h) (González and Prado, 1992).

Aborted seeds (hypocotyl emergence without

radicle protrusion) (Prado et al., 2000) were

also counted. Seeds without radicle and/or

hypocotyl emergence after a 24-h incubation

period were considered as non-germinated

seeds. For each cultivar and each tempera-

ture five 50-seed replicates were done.

From germination curves, the time to

50% germination was determined by fitting

a logistic model of cumulative germination

percentage (G) against time (t), as described

by Eq. 1 and 2 (Covell et al., 1986; Dumur

et al., 1990):

Where,

y: germination percentage in each mea-

sured time;

e: base of natural logarithm;

t: time to each germination percentage;

a and b: regression coefficient con-

stants.

ield temperature

meaSurement

The choice of the mentioned temperature

range was made according to the time that

quinoa is sown in Argentinean Northwest.

Normally quinoa is sown in November. So

field temperature data (air and soil) were

obtained for this month by a device that

was put in a place where quinoa usually

was grown in Tucumán province (Encalilla,

Amaicha del Valle). Temperature was record-

ed each hour, during November 2014, using

a thermocouple with temperature range be-

tween -20 ± 1 ºC and +70 ± 1 ºC (Hobo H8

RH/Temp family Data Logger; Onset Com-

puter Corp., Bourne, MA, USA). Soil and air

temperature were registered at -2 cm and

150 cm respectively.

ardinal temperatureS

Since the estimation of cardinal temper-

atures is highly dependent on the statisti-

cal model used to describe the germination

process (White et al., 2015), both bilinear

and curvilinear models were used to esti-

y = ———— Eq. (1)

1 + be

-at

t = ln ———————–– Eq. (2)

– a

(G – y) / (y · b)

182

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

mate minimum (T

min

), optimum (T

opt

), and

maximum (T

max

) temperatures by regressing

the germination rate (GR), calculated as the

inverse of time to reach 50% germination,

against the temperature gradient. In the bi-

linear model Eq. 3 and 4 were employed to

describe the response of germination to sub-

optimal and supraoptimal temperatures:

y = a

+ b

x (x < T

) Eq. (3)

y = a

+ b

x (x > T

) Eq. (4)

where; a

, b

, a

, and b

are regression

parameters from which the three cardinal

temperatures can be derived. Germination

rate (GR) is the y-value and T

opt

is the x-

value of the breakpoint between the two

lines (Kakani et al., 2002). In the curvilin-

ear model Eq. 5 and 6 were used in order

to model and accurately determine cardinal

temperatures:

y = a + bx + cx

Eq. (5)

y = a + bx + cx

+ dx

Eq. (6)

where; a, b, c, and d are intercept, first,

second and third-order coefficients, respec-

tively. T

min

and T

max

were determined by

lower and upper points were regression

curve intersect the x-axis. The peak of re-

gression curve was deemed as T

opt

(White

et al., 2015).

Difference between T

max

and T

min

known

as temperature adaptability range (TAR)

(Seepaul et al., 2011), showed the germina-

tion ecological range of cultivars:

TAR = T

max

– T

min

reGreSSion analySiS oF Cumulative

aborted SeedS and Cumulative

non

-Germinated SeedS

Based on results of ANOVA analysis, sec-

ond-order regression equations to describe

both the percentage of maximum cumula-

tive aborted seeds and maximum cumula-

tive non-germinated seeds as function of the

temperature gradient, were established for

all cultivars (SAS, 2010).

tatiStiCS

Data are means of three independent

experiments. To detect differences in evalu-

ated parameters (cumulative germination

percentage, cumulative aborted seed per-

centage and non-germinated seed percent-

age) among cultivars, data were processed

by analysis of variance (ANOVA) and means

were compared using the Tukey’s test at P

< 0.05.

RESULTS

air and Soil temperature

According to our field measurements

(Encalilla, Amaicha del Valle, Tucumán,

Argentina at 1,995 m asl) maximum soil

temperature, registered during November

2014, at – 2 cm, was 50,7 º C and minimum

one was 8.6 º C (Fig. 1). Low temperatures

can induce inhibition in the germination of

quinoa seeds due to embryo death as dem-

onstrated by Rosa et al (2004). It is known

that protein synthesis and activation is af-

fected and seed reserves start to deteriorate

(Bove et al., 2001). On the other hand, stress

induced by high temperature increases ab-

normal germination in many crop species

(Pineda Mejia, 1999). Soil temperature may

be an important data to take into account

because water uptake by seeds is a function

of temperature (Sigstad and Prado, 1999)

(see discussion).

ermination time CourSe

The germination time course of quinoa

seeds shows typical sigmoidal curve with a

triphasic pattern: phase I between 0 h and

4-6 h, phase II between 4-6 h and 8-10 h,

and phase III between 8-10 h and 16 h (Fig.

2). In general, all cultivars showed a simi-

lar pattern of cumulative germination with

a fairly constant percentage between 20 °C

and 35 °C. Below 15 °C and above 45 °C

the germination was sharply reduced. Over

50% of maximum germination was reached

by all cultivars between 20 °C and 40 °C,

but in Sajama and Samaranti cultivars values

higher 50% of maximum germination were

found between 15 °C and 45 °C. Maximum

183

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

cumulative germination was achieved dur-

ing the phase III and ranged between 72.5%

(Kancolla) and 90.8% (Chucapaca) (mean

= 84.1%).

aximum Seed Germination reSponSe

temperature

Among linear and polynomial regressions

used to analyse the effect of temperature on

cumulative germination, the quadratic model

best described the response of maximum cu-

mulative germination to temperature (mean

= 0.927) (data not shown). Maximum

germination values derived from quadratic

model ranged between 68.3% (Kancolla)

and 92.5% (Chucapaca) (mean = 83.3%),

while those derived from bilinear model fluc-

tuated between 78.3% (Kancolla) and 100%

(Chucapaca and Sajama) (mean = 92.2%)

(Table 1). Maximum cumulative aborted

seeds and non-germinated seeds are shown

in Fig. 3. Aborted seeds strongly increased

at highest temperatures, being significantly

higher in Robura cultivar with a maximum

value of 53% at 50 °C. High percentages of

aborted seeds were also observed in Sama-

ranti, Kamiri, Sayaña and Ratuqui cultivars.

The lowest percentage of aborted seeds (8%)

occurred in Kancolla cultivar. High values of

the aborted seeds/germinated seeds ratio

(A/G) were observed at low and high tem-

peratures, but were significantly higher in

these last (data not shown). Non-germinated

seeds also occurred along the temperature

gradient. The percentage of non-germinated

seeds was significantly lower than aborted

seed percentage in all cultivars. Maximum

percentage of non-germinated seeds was

28% and occurred in Sajama cultivar at 50

°C. Based on second-order regression equa-

tions, positive and significant curvilinear

relationships between temperature gradient

and both aborted seeds and non-germinated

seeds were found in all cultivars. Values of

ranged between 0.94 (CICA) and 0.98

(Amilda and Kamiri) for aborted seeds and

between 0.84 (Ratuqui) and 0.97 (Sajama)

for non-germinated seeds.

emperature adaptability

ranGe

(tar)

The value of temperature adaptability

range derived from quadratic model varied

between 47 ºC (Kancolla) and 56.6 ºC (Sa-

maranti) (mean = 53.7 ºC), whereas TAR

value derived from cubic model fluctuated

between 52 ºC (CICA) and 58.7 ºC (Sa-

maranti) (mean = 55.7 ºC). Temperature

adaptability range derived from bilinear

model ranged between 63 ºC (Kancolla) and

67.1 ºC (Ratuqui) (mean = 65 ºC), higher

than those estimated from both quadratic

and cubic models (Table 2).

ardinal temperatureS

Quadratic and cubic models (second-

and third-order polynomial regressions)

Fig. 1. Air and soil temperature (A and B respectively) in Encalilla (1995 m asl, Tucumán,

Argentina) during November 2014. (Max): maximal daily temperature, (Min): minimum daily

temperature.

184

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

and bilinear model (linear regression),

used to estimate minimum, optimum and

maximum temperatures for germination of

quinoa cultivar seeds, are shown in Fig. 4

and 5. Minimum temperatures for seed ger-

mination derived from quadratic and cubic

models varied between 4.2 ºC ( and Sama-

ranti) and 5.8 ºC (Kancolla and CICA) for

the former and –2.2 ºC (Sayaña) and 0 ºC

(Amilda) for the latter, respectively (Fig. 4

and 5). The minimum temperature derived

from the bilinear model fluctuated between

–2 ºC (Sajama) and –0.2 ºC (Sayaña) (Fig.

6 and Table 3). Minimum temperatures ob-

tained from both cubic and bilinear models

are correlated to remarkable adaptation of

quinoa to harsh climatic conditions of the

Andean regions (Bois et al., 2006). Optimum

temperatures estimated from the quadratic

model varied between 29.8 ºC (Kancolla)

and 32.5 ºC (Ratuqui) and from the cubic

model between 32 ºC (Kancolla) and 36 ºC

Fig. 2. Cumulative germination of 10 quinoa cultivars at different times. Values are means

± standard error of 5 replicates (250 seeds altogether). For each cultivar large and short

bars represent maximum and minimum values of SE.

185

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

Fig. 3. Maximum cumulative aborted seeds and non-germinated seeds of 10 quinoa cultivars,

at different temperatures with the fitted quadratic equations. Different lowercase letters on

bars denote significant differences in maximum cumulative values of aborted seed. Different

uppercase letters denote significant differences in maximum cumulative values of non-germi-

nated seeds. Values are means ± standard error of 5 replicates.

186

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

Fig. 4. Germination rate of ten quinoa cultivars incubated at different temperatures with the

fitted quadratic equations. Points are the observed data, solid lines are quadratic models.

Error bars represent one standard error of mean.

187

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

(Samaranti). The optimum temperature, es-

timated from the intercept of sub and su-

praoptimal temperature-response functions

(bilinear model), ranged between 31.8 ºC

(Kamiri) and 34.7 ºC (Amilda), slightly low-

er than the values derived from the cubic

model. Maximum temperatures ranged be-

tween 52.9 ºC (Kancolla) and 61 ºC (Saya-

ña), and between 51.5 ºC (CICA) and 56 ºC

(Robura and Samaranti) for quadratic and

cubic models, respectively. The maximum

temperature estimated from the bilinear

model varied between 61.2 ºC (Kancolla

and Kamiri) and 65.7 ºC (Robura), slightly

higher those obtained from polynomial re-

gressions (Table 3).

As a first approximation and based on

optimum temperatures derived from cubic

and bilinear models, two groups of quinoa

seeds can be distinguished: one represented

by Kancolla, Kamiri, Robura and Sajama

cultivars with T

opt

values between 31.8 ºC

and 33.2 ºC and other represented by CICA,

Sayaña, Amilda Ratuqui and Samaranti with

opt

values between 33 ºC and 36 ºC, respec-

tively. No clear segregation was observed

between T

opt

values derived from quadratic

model. Probably a more detailed experiment

is necessary to give a conclusion in relation

to these points.

DISCU S SION

Germination is of great importance be-

cause represents the first step in the plants’

life. Crop success has been ascribed, among

other factors, to high germinative fitness of

seeds, strongly temperature dependent (No-

nogaki et al., 2010). Most of cereal crops

exhibit maximum germination between 20

ºC and 30 ºC, but at lower and higher tem-

perature germination values significantly

decrease. In fact, longer exposure of seeds

to sub- and supraoptimal germination tem-

peratures may lead to increased exposure

to soil pathogens, which can cause decrease

of seedling emergence (Berti and Johnson,

2008). In contrary, crops with broader range

of germination temperature can utilize soil

minerals earlier compared with other spe-

cies (Luna et al., 2012), and then will have

better opportunities to grow and develop in

different agroecological regions. Our results

showed that, seven quinoa cultivars exhibit

their highest germination percentages be-

tween 20 ºC and 40 ºC whereas two cultivars

(Sajama and Samaranti) also showed high

germination percentages between 15 ºC and

45 °C. The quadratic model was significantly

more accurate than the bilinear model to ex-

plain relationships between temperature and

cumulative germination for all quinoa culti-

vars. Furthermore the predicted maximum

cumulative germination percentages from

the bilinear model were higher than those

predicted for the quadratic model (Table

1). According with Yan and Hunt (1999)

the overestimation of maximum cumulative

germination from the bilinear model is be-

cause such estimation is established from the

intercept of two linear regressions, but the

Table 1. Maximum cumulative germination obtained in experimental trials and predicted by

quadratic and bilinear models. Values are means ± standard error of 5 replicates.

Cultivar

72.5 ± 6.4

90.8 ± 8.3

78.7 ± 8.5

80.4 ± 8.6

90.0 ± 9.5

85.7 ± 8.1

80.4 ± 6.9

90.2 ± 10.0

85.1 ± 8.8

87.2 ± 9.2

Kancolla

Chucapaca

Robura

Kamiri

Sajama

Amilda

Ratuqui

CICA

Sayaña

Samaranti

Germination trial (%) Quadratic (%) Bilinear (%)

68.3 ± 7.1

92.5 ± 9.4

80.8 ± 7.6

78.3 ± 6.4

90.0 ± 8.3

86.7 ± 9.2

80.0 ± 8.0

89.2 ± 8.9

80.0 ± 8.9

87.5 ± 8.8

78.3 ± 8.2

100.0 ± 9.3

89.2 ± 9.6

88.3 ± 8.5

100.0 ± 11.0

93.3 ± 8.2

89.2 ± 9.3

97.5 ± 8.7

87.5 ± 9.0

99.2 ± 9.9

188

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

real response of germination to temperature

is curvilinear and usually smoother. In ad-

dition, Hardegree (2006) showed that near

the optimum temperature there is a distinct

plateau (piece-wise regression model) in-

stead of the sharp peak that arises from the

confluence of two straight lines. Nine of ten

quinoa cultivars show values of temperature

adaptability range (TAR) higher than 50 ºC

(Table 2), which allow assuming that quinoa

seeds may have a great potential to success-

fully germinate under field conditions, and

thus will have higher probability of success

than most common cereals in extremely

harsh climatic conditions.

As a consequence of the thermal meta-

bolic dynamics (Essemine et al., 2010) seed

germination rapidly increases between the

base (minimum) temperature and optimum

temperature (suboptimal range), but at tem-

peratures higher than the optimum tempera-

ture (supraoptimal range), the germination

decreases sharply. According to Sigstad &

Prado (1999) water uptake by Robura and

Sajama varieties at 25 ºC was very strong in

the first 15 minutes and a thermal power (=

heat production rate) decreased during the

first 90 minutes of the germination process.

After 105 to 120 minutes, the heat produc-

tion showed an increase which was more

evident between 310 and 390 minutes of

imbibition, when radicule protrudes. From

a thermal point of view, the imbibitions and

heat production may be affected by low

temperature. Obviously this hypothesis may

be tested in lab, but in our experiment the

germination at low temperature (8ºC) only

5 varieties out of 10 achieved a radicule pro-

trusion in the first 12 hs. On the contrary,

high temperatures affect biochemical reac-

tions of seed germination metabolism. In

most species from temperate regions, when

the temperature increases beyond the opti-

mum, the rate of vital processes decreases

and finally ceases at a temperature of  60

°C (K’opondo et al., 2011). Thermal dena-

turation of proteins, membrane dysfunction,

oxidative stress, and decrease of metabolic

efficiency have been proposed as factors that

produce the decline of germination under

supraoptimal temperatures (Hasanuzzaman

et al., 2013). According to Pineda Mejia

(1999) the stress induced by high tempera-

tures increases the percentage of abnormal

germinated seeds and dead seeds in many

crop species. In agreement with this find-

ing the highest percentages of aborted and

non-germinated seeds were observed under

supraoptimal temperatures in all quinoa

cultivars (Fig. 3). Since, the occurrence of

aborted and non-germinated seeds showed

great variability among quinoa cultivars.

Probably it can be better explained by an

exponential function instead of a linear tem-

perature-response. In this case we can as-

sume that the interactive effect of metabolic

and genetic factors instead of the thermal

factor «per se» is the main responsible of the

occurrence of abnormalities during the qui-

noa germination. However the mechanism

underlying abnormal germination occurring

in quinoa seeds at the supraoptimal tem-

Table 2. Temperature adaptability range (TAR) for 10 quinoa cultivars derived from quadrat-ic,

cubic and bilinear models. Values are means ± standard error of 5 replicates.

Cultivar

47.0 ± 4.2

51.8 ± 5.2

55.9 ± 6.0

51.5 ± 4.9

55.7 ± 5.6

55.0 ± 5.0

55.6 ± 6.2

52.4 ± 5.5

55.8 ± 6.2

56.6 ± 5.8

Kancolla

Chucapaca

Robura

Kamiri

Sajama

Amilda

Ratuqui

CICA

Sayaña

Samaranti

Quadratic (ºC) Cubic (ºC) Bilinear (ºC)

56.3 ± 6.2

56.8 ± 5.4

56.8 ± 6.1

54.8 ± 5.8

54.3 ± 5.0

53.3 ± 4.8

57.2 ± 6.5

52.0 ± 6.4

56.4 ± 5.8

58.7 ± 6.5

63.0 ± 5.8

66.7 ± 7.2

66.4 ± 6.0

62.4 ± 6.2

66.8 ± 7.1

63.0 ± 6.5

67.1 ± 5.9

65.7 ± 6.4

63.7 ± 5.7

65.5 ± 6.8

189

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

Fig. 5. Germination rate of ten quinoa cultivars incubated at different temperatures with the

fitted third-order equations.Points are the observed data, solid lines are cubic models. Error

bars represent one standard error of mean.

190

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

Fig. 6. Germination rate of ten quinoa cultivars incubated at different temperatures with the

fitted linear equations. Points are the observed data, solid lines are bilinear models. Error

bars represent one standard error of mean.

191

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

perature gradient needs further investiga-

tion. The higher percentage of aborted and

non-germinated seeds observed at supraop-

timal temperatures could also explain the

ancestral cropping practice applied by the

Andean farmers, based on the need to sow

more than 100 quinoa seeds per hole to get

a maximum of 10-15 plants. Supporting this

assumption, both strong rises of the diurnal

temperature and wide daily thermal ampli-

tudes as consequence of the intense solar ra-

diation and low atmospheric water content,

are recorded frequently in highland arid re-

gions of Bolivia, Peru, Chile and Argentina,

where the quinoa cultivation is an ancestral

practice (National Research Council, 1989).

Caution is required when analysing relation-

ships between daily thermal amplitude and

quinoa germination in the field.

Cardinal temperatures i.e. T

min

, T

opt

and

max

, are important parameters to evaluate

the success of seed germination and seedling

establishment (Saeidnejad et al., 2012), and

then its determination becomes important

when a new crop will be introduced in a

given region. Regarding to quinoa, Bois et

al. (2006) in a study of ten quinoa cultivars

from the Andean region (Bolivian Altiplano)

found a linear positive correlation between

germination and temperature in a tempera-

ture gradient between 2 ºC and 20 ºC. Val-

ues of T

min

reported in this work ranged be-

tween –1.97 ºC and 0.24 ºC that were closely

similar to reported here, derived from both

cubic (–2.2 ºC to 0 ºC) and bilinear (–2.0

ºC to –0.5 ºC) models, respectively (Table

3). A value close to 0 ºC has been recently

communicated for the T

min

of Chenopodium

pallidicaule, another Chenopodiaceae species

native from the Andean region (Rodriguez et

al., 2016). However, values of T

min

derived

from the quadratic model were significantly

higher than values reported by Bois and co-

workers as well as those derived from cubic

and bilinear models, indicating that these

latter seem to be more reliable to establish

the T

min

of quinoa seed germination. None-

theless, Jacobsen and Bach (1998) using a

linear model reported a T

min

value of 3 °C

for a Danish quinoa cultivar. Significant dif-

Table 3. Estimates of minimum temperature (Tmin), optimum temperature (Topt) and maximum temperature (Tmax), for germination of 10

quinoa cultivars using quadratic, cubic and linear functions. Values are means ± standard error of 5 replicates.

Cultivar

5.8 ± 0.4

4.2 ± 0.3

4.3 ± 0.4

4.5 ± 0.3

4.5 ± 0.4

4.8 ± 0.5

5.2 ± 0.5

5.8 ± 0.6

5.2 ± 0.5

4.2 ± 0.5

Kancolla

Chucapaca

Robura

Kamiri

Sajama

Amilda

Ratuqui

CICA

Sayaña

Samaranti

Quadratic Cubic Bilinear

min

(ºC) T

opt

(ºC) T

max

(ºC) T

min

(ºC) T

opt

(ºC) T

max

(ºC) T

min

(ºC) T

opt

(ºC) T

max

(ºC)

29.8 ± 1.8

30.0 ± 3.1

32.0 ± 2.0

30.5 ± 3.2

32.2 ± 3.6

32.0 ± 3.0

32.5 ± 2.2

32.0 ± 3.7

33.2 ± 2.9

32.5 ± 3.4

52.9 ± 4.3

56.0 ± 5.0

60.2 ± 6.3

56.0 ± 6.0

60.2 ± 5.9

59.8 ± 6.3

60.8 ± 6.0

58.2 ± 4.9

61.0 ± 6.4

60.8 ± 5.2

–1.8 ± (–0.2)

–2.0 ± (−0.3)

–0.8 ± (−0.1)

–0.3 ± (–0.0)

–1.5 ± (–0.2)

0.0 ± 0.0

–1.5 ± (–0.2)

–0.5 ± (–0.1)

–2.2 ± (–0.3)

–0.5 ± (–0.1)

32.0 ± 3.6

33.0 ± 2.9

33.0 ± 3.3

32.5 ± 3.6

33.2 ± 2.5

35.0 ± 4.1

35.2 ± 3.0

33.8 ± 2.9

35.0 ± 3.1

36.0 ± 4.0

54.5 ± 4.1

54.8 ± 5.5

56.0 ± 4.9

54.5 ± 6.1

52.8 ± 3.9

53.3 ± 5.0

55.7 ± 5.0

51.5 ± 4.8

54.2 ± 5.7

56.0 ± 5.7

–1.8 ± (–0.2)

–2.0 ± (–0.2)

–0.7 ± (–0.1)

–1.2 ± (–0.3)

–2.0 ± (–0.2)

–1.2 ± (–0.1)

–1.6 ± (–0.2)

–0.5 ± (–0.1)

–0.2 ± (–0.0)

–1.0 ± (–0.2)

32.0 ± 2.8

32.0 ± 3.4

32.7 ± 3.4

31.8 ± 2.8

33.2 ± 3.3

34.7 ± 3.3

34.5 ± 4.1

33.5 ± 3.2

34.6 ± 3.0

34.0 ± 3.0

61.2 ± 6.7

64.7 ± 6.7

65.7 ± 5.9

61.2 ± 4.7

64.8 ± 6.8

61.8 ± 5.3

65.5 ± 6.3

65.2 ± 5.9

63.5 ± 6.8

64.5 ± 6.1

192

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

ferences between field and controlled condi-

tions in T

min

values were also reported for

quinoa cultivars (Bertero et al., 1999, cited

by Bois et al., 2006). Disparity in T

min

values

within the same species occurring under both

field and controlled conditions appears as a

common trait in most plants, but it cannot

be explained by a simple factor such as seed

size or seed water status (Wang et al., 2004;

Naim and Ahmed, 2015). Complex maternal

effects and the variation in individual seed

sensitivity to temperature are interacting and

trigger the disparity of germination response

at intraspecific level (Vange et al., 2004).

The T

opt

values calculated from linear

and polynomial regression models showed

variation among cultivars but in all models

the maximum germination occurred at T

opt

values from 29.8 ºC to 36 °C, being slightly

lower in the quadratic model. In agreement

with our results, Jacobsen and Bach (1998)

reported a value of T

opt

between 30 ºC and

35 ºC for maximum germination of different

populations of a Danish quinoa cultivar. By

contrast maximum temperature (T

max

) was

higher in the bilinear model compared with

both quadratic and cubic models. According

with Hardegree (2006) when the regression

line is forced to intercept the x-axis at supra-

optimal temperatures, an overestimation of

the T

max

occurs. By comparison of data, bilin-

ear (mean R

= 0.975) and cubic (mean R

= 0.914) models seem to have comparative

advantage over the quadratic model (mean

= 0.825) to estimate cardinal tempera-

tures of quinoa seed cultivars.

CONCL U SIONS

Our results showed that high and low

temperature affects not only the rate but

also the maximal germination. On the basis

of a cost/benefit analysis, data of this study

indicates that to perform quinoa cultiva-

tion in mountainous regions it is preferable

planning the crop on an altitudinal ecologi-

cal zoning base, that is to say, select one or

more cultivars for each altitudinal level. This

implies to know both thermal requirements

and germinative fitness (germination per-

centage, germination rate, cardinal temper-

atures, non-germinated seeds and aborted

seeds) of quinoa cultivars for each selected

site to choose more suitable cultivars before

starting the crop.

ACKNOW LEDGE M ENTS

This work was supported by the Agen-

cia de Promoción Científica y Tecnológica

(grant PICT 23153) and Fundación Miguel

Lillo (Miguel Lillo 251, T4000JFE, Tucumán,

Argentina).

REFER E NCES

Abugoch James L. E. 2009. Quinoa (Che-

nopodium quinoa Wil

ld.): composition,

chemistr y, nutritional, and functional

properties. Advances in Food Nutrition

Research 58: 1-31.

zile D., Fuentes F. F., Mujica A. 2013. His-

torical perspectives and domestication of

quinoa, in: Bhargava, A., Srivastava S.

(Eds.), Quinoa: Botany, Production and

es. CAB International, Wallingford,

UK, pp. 16-35.

rtero H. D. 2001. Effects of photoperiod,

temperature and radiation on the rate of

leaf appearance in quinoa (Chenopodium

quinoa Wi

ld.) under field conditions. An-

nals of Botany 87: 495-502.

rtero H. D., King R. W., Hall A. J. 1999.

Modelling photoperiod and temperature

responses of flowering in quinoa (Che-

nopodium quinoa Wi

ld.). Field Crops

Research 63: 19-34.

Berti M. T., Johnson B. L. 2008. Seed germi-

nation response of cuphea to tempera-

ture. Industrial Crops and Products 27:

17-21.

wley J. D., Black M. 1994. Seeds: Physiolo-

of Development and Germination, sec-

ond ed. Plenum Press, NewYork, USA.

is J. F., Winkel T., Lhomme J. P., Raffaillac

J. P., Rocheteau A. 2006. Response of

some Andean cultivars of quinoa (Che-

nopodium quinoa Wi

ld.) to tempera-

ture: effects on germination, phenology,

growth and freezing. European Journal

of Agronomy 25: 299-308.

Bonhomme R. 2000. Bases and limits to us-

ing ‘degree day’ units. European Journal

of Agronomy 13: 1-10.

Bove J., Jullien M., Grappin P. 2001. Func-

tional genomics in the study of seed

germination. Genome Biology, 3 (1): re-

views 1002.1-1002.5.

193

Lilloa 54 (2): 179–194, 7 de diciembre de 2017

Choukr-Allah R., Rao N. K., Hirich A., Shahid

M., Alshankiti A., Toderich K., Gill S.,

Butt K. U. R. 2016. Quinoa for mar-

ginal environments: toward future food

and nutritional security in MENA and

central Asia regions. Frontiers in Plant

Science 7: 346.

Covell S., Ellis R. H., Roberts E. H., Sum-

merfield R. J. 1986. The influence of

temperature on seed germination rate in

grain legumes. Journal of Experimental

Botany 37: 705-715.

Dixon M., Webb E. C. 1979. Enzymes. Pren-

tice Hall Press, New Jersey, USA.

Dumur D., Pilbeam C. J., Craigon J. 1990.

Use of the Weibul function to calculate

cardinal temperatures in Faba bean.

Journal of Experimental Botany 41:

1423-1430.

Essemine J., Ammar S., Bouzid S. 2010.

Impact of heat stress on germination

and growth in higher plants: physiologi-

cal, biochemical and molecular repercus-

sions on mechanisms defence. Journal

of Biological Sciences 10: 565-572.

González J. A., Prado F. E. 1992. Germina-

tion in relation to salinity and tempera-

ture in Chenopodium quinoa (Willd.).

Agrochimica 36: 101-108.

Gonzalez J. A., Konishi Y., Bruno M., Valoy

M., Prado F. E. 2011. Interrelationships

among seed yiel d , to t al protein and

amino acid composition of ten quinoa

(Chenopodium quinoa) cultivars from two

different agroecological regions. Journal

of the Science of Food and Agriculture

92: 1222-1229.

González J. A., Eisa S., Hussin S., Prado

F. E. 2015. Quinoa: an Incan Crop to

Face Global Changes in Agriculture. In:

Murphy, K.S., Matanguihan, J. (Eds.).

Quinoa: Improvement and Sustainable

Production. Wiley-Blackwell, Hoboken,

NJ, USA, pp. 1-18.

Hardegree S. P. 2006. Predicting germination

response to temperature. I. Cardinal-

temperature models and subpopulation-

specific regression. Annals of Botany

97: 1115-1125.

Hasanuzzaman M., Vahar K., Fujita M. 2013.

Extreme temperature responses, oxida-

tive stress and antioxidant defense in

plants, in: Vahdati, K., Leslie, C. (Eds.),

Abiotic Stress – Plant Responses and

Applications in Agriculture. InTech, Croa-

tia, pp. 169-203.

Jacobsen S. E., Bach A. P. 1998. The influ-

ence of temperature on seed germina-

tion rate in quinoa (Chenopodium quinoa

Willd). Seed Science and Technology 26:

515-523.

Jacobsen S. E., Monteros C., Christiansen

J. L., Bravo L. A., Corchera L. J., Mu-

jica A. 2005. Plant responses of quinoa

(Chenopodium quinoa Willd.) to frost at

various phenological stages. European

Journal Agronomy 22: 131-139.

KakaniV. G.,Prasad P. V. V.,Craufurd P. Q.,

Wheeler T. R. 2002. Response on in

vitro pollen germination and pollen tube

growth of groundnut (Arachis hypogaea

L.) genotypes to temperature. Plant, Cell

and Environment 25: 1651-1661.

Kamkar B., Al-Al a h m a d i M. J., Ma h d a v i -

Damghani A., Villalobos F. J. 2012.

Quantification of the cardinal tempera-

tures and thermal time requirement of

opium poppy (Papaver somniferum L.)

seeds to germinate using non-linear re-

gression models. Industrial Crops and

Products 35: 192-198.

K’Opondo F. B. O., Groot S. P. C., Van Rheen-

en H. A. 2011. Determination of tem-

perature and light optima for seed ger-

mination and seedling development of

spiderplant (Cleome gynandra L.) mor-

photypes from western Kenya. Annals

Biological Research 2: 60-75.

Luna B., Pérez B., Torres I., Moreno J. M.

2012. Effects of incubation temperature

on seed germination of Mediterranean

plants with different geographical dis-

tribution ranges. Folia Geobotanica 47:

17-27.

Naim A. H., Ahmed F. E. G. 2015. Variation

in thermal time model parameters be-

tween two contrasting chickpea (Cicer

arietinum) cultivars. Agricultural Scienc-

es 6: 1421-1427.

National Research Council. 1989. Lost Crops

of the Incas: Little-Known Plants of the

Andes with Promise for Worldwide Cul-

tivation. National Academy Press, Wash-

ington, DC, USA.

Nonogaki H., Bassel G. W., Bewley J. D.

2010. Germination still a mystery. Plant

Science 179: 574-581.

Pineda Mejia R. 1999. Effects of stress tem-

peratures of germination on polyamine

titers of soybean seeds. Ph.D. thesis,

Iowa State University, USA, pp. 97.

Prado F. E., Boero C., Gallardo M., González

J. A. 2000. Effect of NaCl on germina-

tion, growth, and soluble sugar content

in Chenopodium quinoa Willd. seeds. Bo-

tanical Bulletin of Academia Sinica 41:

27-34.

194

J. A. González et al.: Quantifying cardinal temperatures in Chenopodium quinoa cultivars

Prado F. E, Fernández-Turiel J. L., Tsarouchi

M., Psaras G. K., González J. A. 2014.

Variation of seed mineral concentrations

in seven quinoa cultivars grown in two

agroecological sites» Cereal Chemistry

91: 453-459.

Risi J., Galwey N. W. 1984. The Chenopo-

dium grains of the Andes: Inca crops for

modern agriculture. Advances in Applied

Biology 10: 145-216.

Rodriguez J. P., Jacobsen S. E., Sørensen

M., Andreasen C. 2016. Germination

responses of cañahua (Chenopodium

pallidicaule Aellen) to temperature and

sowing depth: a crop growing under

extreme conditions. Journal of Agrono-

my and Crop Science. DOI: 10.1111/

jac.12158.

Rosa M., Hilal M., González J. A., Prado F. E.

2004. Changes in soluble carbohydrates

and related enzymes induced by low

temperature during early developmental

stages of quinoa (Chenopodium quinoa

Willd.) seedlings. Journal of Plant Physi-

ology 161: 683-689

Saeidnejad A. H., Kafi M., Pessarakli M.

2012. Evaluation of cardinal tempera-

tures and germination responses of four

ecotypes of Bunium persicum under dif-

ferent thermal conditions. International

Journal of Agriculture and Crop Sciences

4: 1266-1271.

SAS Statistical Analysis Software for Win-

dows. Release 9.1.3., 2010. C a r y,

North Carolina, USA.

Seepaul R., Macoon B., Reddy K. R., Bald-

win B. 2011. Switchgra ss (Panicum

virgatum L.) intraspecific variation and

thermotolerance classification using in

vitro seed germination assay. American

Journal of Plant Sciences 2: 134-147.

Strenske A., Soares de Vasconcelos, E., Ege-

warth V. A, Michelon Herzog N. F., de

Matos Malavasi M. 2017. Responses

of quinoa (Chenopodium quinoa Willd.)

seeds stored under different germina-

tion temperaturas. Acta Scientiarum.

Agronomy Maringá 39: 83-88

Vange V., Heuch I., Vandvik V. 2004. Do seed

mass and family affect germination and

juvenile performance in Knautia arven-

sis? A study using failure-time methods.

Acta Oecologica 25: 169-178.

Vega-Gálvez A., Miranda M., Vergara J., Uribe

E., Puente L., Martínez E. A. 2010. Nu-

trition facts and functional potential of

quinoa (Chenopodium quinoa willd.), an

ancient Andean grain: a review. Journal

of the Science of Food and Agriculture

90: 2541-2547.

Wang R., Bai Y., Tanino K. 2004. Effect of

size and sub-zero imbibition-temperature

on the thermal time model of winterfat

(Eurotia lanata (Pursh) Moq.). Environ-

mental and Experimental Botany 51:

183-197.

White S. N., Boyd N. S., Van Acker R. C.

201 5 . Te m pe r a tu r e th r e sh o ld s a nd

growing-degree-day models for red sor-

rel (Rumex acetosella) ramet sprouting,

emergence, and flowering in wild blue-

berry. Weed Science 63: 254-263.

Yan W., Hunt L. A. 1999. An equation for

modeling the temperature response of

plants using only the cardinal tempera-

ture. Annals of Botany 84: 607-614.