Evapotranspiration Analysis and Irrigation PDF

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/223159378

Evapotranspiration analysis and irrigation requirements of


quinoa (Chenopodium quinoa) in the Bolivian highlands

Article  in  Agricultural Water Management · February 2003


DOI: 10.1016/S0378-3774(02)00162-2

CITATIONS READS

116 801

3 authors, including:

Magali Garcia Sven-Erik Jacobsen


Universidad Mayor de San Andres 211 PUBLICATIONS   7,091 CITATIONS   
40 PUBLICATIONS   1,347 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Linking local and scientific knowledge to reduce climate related risk View project

SWUP-MED EU 7th Framework Project View project

All content following this page was uploaded by Magali Garcia on 15 March 2018.

The user has requested enhancement of the downloaded file.


Agricultural Water Management 60 (2003) 119–134

Evapotranspiration analysis and irrigation


requirements of quinoa (Chenopodium quinoa)
in the Bolivian highlands
Magalı́ Garciaa, Dirk Raesb, Sven-Erik Jacobsenc,*
a
Universidad Mayor de San Andrés, Facultad de Agronomia, La Paz, Bolivia
b
Department of Land Management, Katolieke Universiteit Leuven, Leuven, Belgium
c
International Potato Center (CIP), Apartado 1558, Lima 12, Peru
Accepted 21 October 2002

Abstract

Quinoa (Chenopodium quinoa Willd.) is a staple crop of the Altiplano of the high Andean region,
up to 4000 m above sea level, which provides a highly nutritive food for the local and urban
population, under adverse climatic conditions of frequent droughts, night frost and poor soils, and is
regarded as having a great potential for export. With the aim of increasing production and
productivity, new irrigation systems are required. This is causing problems, because the standard
FAO formula for calculating reference evapotranspiration ET0 (Penman–Monteith) has not been
calibrated for altitudes above 3000 m, so that the crop coefficient Kc values for quinoa must be
adapted from other crops such as barley and wheat. Additionally, in cases where water availability for
quinoa does not reach the levels of crop evapotranspiration ETc, the reaction to drought stress and
water use efficiency has not been quantified.
In order to estimate irrigation requirements of quinoa in a representative site of the Bolivian
Altiplano, the crop water requirement, the crop coefficient (Kc), the yield response factor (Ky), and the
relative yield, was derived from lysimeter and field data. The Penman–Monteith formula for the
conditions of a representative site of the Bolivian Altiplano was calibrated in relation to grass
evapotranspiration data, used as reference crop, from lysimeters. The results showed that an excellent
agreement exists between measured and calculated ET0 values when an estimate of dew point
temperature, derived from minimum air temperature, was used in the calculation instead of recorded
humidity data. The Kc for quinoa, obtained from lysimeter data, varied over the growing season being
0.5 in the initial growth stage, 1.00 in the mid-season stage and 0.70 at harvest. The seasonal yield

*
Corresponding author. Present address: Department of Agricultural Sciences, Royal Veterinary and
Agricultural University, Agrovej 10, Taastrup DK-2630, Copenhagen, Denmark. Tel.: þ45-3528-3388;
fax: þ45-3528-3384.
E-mail addresses: magali.garciacardena@agr.kuleuven.ac.be (M. Garcia), dirk.raes@agr.kuleuven.ac.be
(D. Raes), seja@kvl.dk (S.-E. Jacobsen).

0378-3774/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 3 7 7 4 ( 0 2 ) 0 0 1 6 2 - 2
120 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

response factor (Ky) of 0.67 was estimated from data on grain yield and crop evapotranspiration
obtained in lysimeters. Relative yields obtained from lysimeters, sparsely irrigated and non-irrigated
fields, indicate that it may be more profitable to irrigate larger extensions of land using the maximum
water use efficiency approach than to irrigate with the same total amount of water to a much smaller
area applying maximum water use.
# 2002 Elsevier Science B.V. All rights reserved.

Keywords: Reference evapotranspiration; Crop coefficient; Crop evapotranspiration; Yield response factor;
Water use efficiency

1. Introduction

The food security of the rural inhabitants of the high Andes is tenuous because most are
subsistence farmer families with small land holdings, and it is estimated that almost half of
the children suffer from chronic malnutrition (Ministerio de Educación, 1994). The
principal source of protein in the Andes is quinoa (Chenopodium quinoa Willd.), but
often the supply is not sufficient to meet the nutritional demands of the population. Quinoa,
an annual crop species related to well-known agricultural crops such as sugar beet (Beta
vulgaris), spinach (Spinacia oleracea) and amaranth (Amaranthus sp.), is a 1–1.5 m high
plant terminating in a panicle consisting of small flowers, each of which produce one seed
of 2.5 mg and a diameter of 1 mm. Quinoa, consumed in bread, soups, biscuits, drinks, etc.
has a high protein content, but what is more important is a high protein quality with a high
level of the essential amino acids, and a wide range of vitamins (A, B2, E) and minerals (Ca,
Fe, Cu, Mg, Zn) (Repo-Carrasco et al., 2001). Some farmers rely on quinoa as a cash crop,
which has become increasingly important as the worldwide demand for quinoa is
increasing. Quinoa is regarded as one of the most drought-resistant crops existing, in
addition to a high level of tolerance to frost and saline soils, diseases and pests (Jacobsen
and Mujica, 2001; Mujica et al., 2001).
Because of quinoa’s nutritional qualities, its natural adaptation to harsh conditions, its
increasing marketability, and its cultural acceptability, increased quinoa production is
regarded as the single most important prospect for addressing the food security problems of
the Andes, and therefore it has been selected by FAO as one of the crops destined to offer
food security over the 21st century (FAO, 1998).
Quinoa is one of the main crops of the region, adapted to growth in dry environments and
under the generally harsh, climatic conditions (Vacher, 1998; Altieri, 1997; Jacobsen et al.,
1999). Water shortage is a major constraint to plant production due to the combined effect
of low rainfall, a relatively high evapotranspiration rate and soils with a low water retention
capacity. In dry years, water shortage may cause considerable yield reductions. Despite the
adaptation of quinoa to adverse conditions, better than most other crops, it was demon-
strated by Garcia (1992, 1999) that a small water application may result in considerable
yield increments. However, the precise amount of water necessary for obtaining a
reasonable yield have yet not been determined.
For the determination of irrigation requirements the standard method to estimate ET0
(Penman–Monteith) has not been properly examined for conditions above 3000 m a.s.l.,
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 121

such as under the extreme climatic conditions of the Altiplano. Additionally, the crop
coefficient (Kc) for quinoa has yet not been determined, so that Kc values for other crops
such as barley and wheat are often used for the calculations (Michel and Garcia, 1997).
Research on irrigation requirements and yield response to different levels of watering are
lacking, which makes the dimensioning of irrigation systems difficult (Mungai and Stigter,
1993). Therefore, there is an urgent need for obtaining an adequate method to determine
irrigation requirements of quinoa under Altiplano conditions.
Water requirement calculations for irrigation of a given crop under rainfall conditions of
a given locality entail (i) determining reference evapotranspiration (ET0), (ii) calculating
crop evapotranspiration (ETc ¼ Kc  ET0 ), (iii) subtracting rainfall from ETc in order to
obtain net irrigation requirements, and (iv) obtaining net irrigation requirements by
analyzing soil water balance. The net irrigation requirements of quinoa and the quinoa
Kc will be determined based on water use efficiency, through a detailed analysis of
reference and crop evapotranspiration.

2. Materials and methods

The experiments were conducted in the Bolivian Altiplano in Patacamaya (Fig. 1). The
trial was sown in November and harvested in April. The four growth phases defined by
Doorenbos and Pruitt (1986) were easily recognized for quinoa.
The data was obtained from lysimeter studies and field measurements in Patacamaya,
Bolivia (178150 S, 678550 W, 3786 m a.s.l.), which is a representative site of the Bolivian
Altiplano. Quinoa evapotranspiration (ETc) was determined under optimal growing
conditions in 1990, using two drainage lysimeters with dimensions of
2:0 m  2:0 m  0:9 m (Choquecallata, 1993). The lysimeters, which were located in
the middle of a cultivated area of 1500 m2, were established in 1985, 5 years prior to the
start of the experiment. Observations of drainage and soil water content were performed
every 3 days, and soil water balance was computed, including data for irrigation and
rainfall. The soil within the lysimeter was maintained permanently at field capacity, which
was controlled by a Solo Nardeux neutronprobe. Irrigation and drainage were controlled
using calibrated pots. Quinoa was irrigated every 5 days when it did not rain, resulting in
154 mm irrigation and 156 mm rain, in total 310 mm of water applied during the growing
period.
Grass evapotranspiration was determined in the same site in 1992 for the evaluation of
reference evapotranspiration (ET0) (Canqui, 1993). Ryegrass (Lollium perenne) was
established in the lysimeters under the same conditions as quinoa for 12 months. The
height of the grass was maintained at 12 cm during the whole period. In summer, that is the
rainy season from October to April, the grass had to be cut almost each week, but in winter
the frequency of cuts was lower due to a reduced growing rate of the crop. When necessary,
the grass was irrigated to avoid drought stress.
To determine the evapotranspiration of quinoa under modest irrigation applications and
rainfed conditions, Garcia (1994) determined, by means of a weekly soil water balance, the
adjusted crop evapotranspiration (ETcadj ). With exception of the water regime, growth
conditions and crop management in the field were similar to the conditions of quinoa in the
122 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

Fig. 1. Distribution of the cultivated areas of quinoa in South America (from Garcia, 1994).

lysimeters, because the experiments were conducted in the same location and the same
year. Drainage was determined in the field by the method of Hillel (1984), demonstrating
that it could be considered as negligible. The water table is typically 30 m deep in this area,
and therefore groundwater contribution will have no effect. The soil water content was
measured every 3 days with the neutron probe, using the calibrated equations for the
specific soil of Patacamaya. The plots were irrigated four times during the growing period,
when the rainfall was sparse, and following the normal irrigation scheme in the area. The
rainfed plots relied on the natural precipitation.
For all experiments, the meteorological data was obtained from a climate station located
50 m from the plots (Table 1). Monthly grass evapotranspiration obtained with lysimeters
(Canqui, 1993), was compared to the ET0 calculated using the Penman–Monteith formula
for Patacamaya (Allen et al., 1991, 1994):
0:408DðRn  GÞ þ gð900=ðT þ 273ÞÞu2 ðes  ea Þ
ET0 ¼ (1)
D þ gð1 þ 0:34u2 Þ
where ET0 is the reference evapotranspiration (mm per day), Rn the net radiation at the crop
surface (MJ m2 per day), G the soil heat flux density (MJ m2 per day), T the mean daily
air temperature at 2 m altitude (8C), u2 the wind speed at 2 m altitude (m s1), es the
saturation vapor pressure (kPa), ea the actual vapor pressure (kPa), es  ea the saturation
vapor pressure deficit (kPa), D the slope vapor pressure (kPa 8C1) and g is the psycho-
metric constant (kPa 8C1).
This equation is based on data from Allen et al. (1998), which states that the reference
evapotranspiration closely resembles an extensive surface of green, well-watered grass of
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 123

Table 1
Climatic conditions in Patacamaya (1989–1990 for quinoa and 1992 for ryegrass)

Month Tmin (8C) Tmax (8C) Rainfall (mm) RH (%)

1989–1990
November 2.8 18.8 16.0 58
December 4.4 19.2 27.5 61
January 5.5 16.8 70.0 74
February 3.8 17.6 52.0 69
March 3.5 17.4 14.0 65
April 1.7 17.0 20.0 63
1992
January 2.8 17.8 79.4 75.8
February 2.4 18.2 27.2 70.5
March 2.5 17.3 63.0 70.2
April 1.1 15.6 6.1 65.4
May 1.3 14.3 0.0 44.2
June 4.5 13.2 13.4 39.2
July 4.3 14.1 0.0 47.1
August 3.8 15.6 1.5 55.2
September 1.5 14.5 4.5 52.4
October 1.8 15.6 5.8 55.1
November 2.5 15.2 55.2 57.2
December 2.3 16.2 65.2 56.1

uniform height, actively growing and completely shading the surface. The differences
between the results (grass and Penman–Monteith ET0) were analyzed.
For every week during the growth period, the crop coefficient (Kc) was calculated by the
following equation:
ETc
Kc ¼ (2)
ET0
where ETc (crop evapotranspiration under optimal conditions) was obtained from the
lysimeter data, and ET0 (reference evapotranspiration) was calculated by means of the
corrected FAO Penman–Monteith. The Kc values found were used to estimate Kc for quinoa
according to the method proposed by Doorenbos and Pruitt (1986) for the initial, mid and
late crop growth seasons.
The effect of drought stress was estimated with the yield response factor (Ky):
   
Ya ETadj
1  Ky 1  (3)
Ym ETc
where Ya is the actual yield under non-standard conditions, and Ym is the maximum
expected crop yield under maximum ETc (Doorenbos and Khassam, 1979). Ky is the
standard indicator of the sensitivity to drought stress for the crop. Ky values below 1
indicate a drought-resistant crop, and values above 1 a crop not well adapted to drought.
The relative yield (Ya/Ym) was also calculated for the sparsely irrigated and the rainfed
fields to determine the water use efficiency in the analyzed points.
124 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

The net irrigation requirements (Inet) to fulfill both the crop evapotranspiration and the
adjusted evapotranspiration were determined using 25 years historical rainfall data. The
75% 10-daily dependable rainfall (PP75%), that is rainfall with a probability of 75%, was
calculated and used in the water balance:
Inet ¼ ET  PP75% (4)

3. Results

3.1. Reference evapotranspiration (ET0)

Calculated and measured (lysimeter) grass ET0 are plotted in Fig. 2. Due to the variation
of climate in the Altiplano, grass ET0 exhibits large variations between the dry and the
rainy season (Table 1). The highest rate of water demand occurred during the months of
maximum solar radiation, which is in the rainy summer. Vacher et al. (1994) reported that
the reduction in measured ET0 during the winter season could be attributed not only to
reduced atmospheric demand and net radiation, but also to the possible reduction of the
metabolic activity of grass, due to decreasing minimum temperatures (Tmin).
The ET0 values from the Penman–Monteith equation were obtained by applying
available climatic data (solar radiation, air temperature, air humidity and wind speed),
whereas the calculated ET0 showed less variation than those obtained from lysimeters
(Fig. 2).

3.2. Crop evapotranspiration (ETc)

Measured crop evapotranspiration (ETc) data for quinoa from lysimeters are seen in
Table 2. Minimum water requirements occurred during the first 2 weeks, when only few
leaves contributed to the transpiration, and most ETc was evaporation from the soil. Water
consumption increased from week 5–8, mainly due to water use by the plants during the
vegetative stage. Typical of a grain crop, maximum water requirements occurred during the
flowering and ripening stages (Mariscal, 1992). During the next 4 weeks (8–12), water use
was stable until it decreased from week 13 (seed filling).

Fig. 2. Reference evapotranspiration ET0, estimated from lysimeters (ET0 grass), and calculated from Penman–
Monteith formula (ET0 P–M).
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 125

Table 2
ETc (mm per day) of quinoa determined from drainage lysimeters in Patacamaya during the 1990 growth season
(Choquecallata, 1993)

Period Weeks after sowing ETc (mm per day) Phenological phase

15 December to 3 January 1–3 2.80 Emergence


5 January to 10 January 4 2.77 Four true leaves
11 January to 22 January 5–6 3.12 Vegetative stage
26 January to 31 January 6–7 3.32 Bud formation
3 February to 11 February 8 4.21 Flowering
13 February to 25 February 9–10 4.54 Deflowering
3 March to 8 March 11–12 4.71 Milky grain
10 March to 20 March 13–14 4.08 Pasty grain
23 March to 5 March 14–15 3.25 Physiological maturity
Average 3.6
Total (mm per season) 450

3.3. Adjusted evapotranspiration (ETcadj ) under soil-moisture deficit, and the


yield response factor (Ky)

In the field plots used to calculate ETcadj , the water balance showed that with 154 mm of
irrigation and 156 mm of rainfall the total applied water was 310 mm, from which 306 mm
were consumed by the crop and 4 mm remained in the soil (Table 3). Crop evapotranspira-
tion (ETc) under optimum water application was compared to the adjusted crop evapo-
transpiration (ETcadj ) under limited irrigation applications (Fig. 3). At the beginning of the

Table 3
Soil water balance per week to determine adjusted evapotranspiration

Weeks after Rainfall Irrigation Difference in soil ETcadj ETcadj


sowing (mm) (mm) water content (mm) (mm per week) (mm per day)

3 – – 10.4 10.4 1.6


4 44 – 23.1 20.9 2.9
5 7 – 14.2 21.2 3.0
6 10 39 25.8 23.2 3.3
7 7 – 16.0 23.0 3.3
8 21.5 – 4.5 26.0 3.7
9 5.5 39 18.6 25.9 3.7
10 21 – 5.8 26.8 3.8
11 4.5 39 16.5 27.0 3.9
12 2.5 – 22.2 24.7 3.5
13 9.5 – 8.1 17.6 2.5
14 4.1 39 16.9 26.2 2.2
15 4.0 – 13.5 17.5 2.2
16 13 – 2.4 15.4 2.2
Total 153.6 156.0 3.8 305.8
126
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134
Fig. 3. Crop (ETc) and adjusted evapotranspiration (ETcadj ) for quinoa.
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 127

growth season, ETc and ETcadj were similar, due to low water consumption and sufficient
rainfall to eliminate the need for irrigation. By weeks 8–9, atmospheric water demand
increased due to the low atmospheric humidity and sparse rainfall, which in the case of
ETcadj , where irrigation was applied only in four occasions, was insufficient to cover the
crop’s potential water consumption.

4. Discussion

4.1. Reference evapotranspiration

The relatively stable values of the Penman–Monteith calculations do not correspond


with the large variation in radiation and temperature on the Altiplano. When compared
to the lysimeter data, the Penman–Monteith formula seems to underestimate ET0
values during the summer growth season, and to overestimate ET0 values during
periods of low physiological activity (winter), with a low correlation (R2 ¼ 0:5375).
The degree of underestimation during summer is in some cases as much as 1.3 mm per
day or 30%, which creates a monthly underestimation of up to 40 mm. The over-
estimation of ET0 during the winter season is due to vegetation dormancy. This season
is of relatively low importance in terms of irrigation needs, since no agricultural
activities occur. However, for the summer season, some corrections of the input data,
that is the climatic records, are required to estimate ET 0 with the Penman–Monteith
formula.
The main factors influencing ET0 are solar radiation, air humidity, air temperature and
wind speed. It may be adequate to perform an analysis of their behavior during the
observed period to find a possible source of error. Solar radiation was directly measured
and can possibly be left out, and wind speed and temperature followed normal patterns
for the area (Michel and Garcia, 1997), so they should also be eliminated as sources of
error. Vapor pressure deficit, however, was estimated from temperature and relative
humidity, the latter reported to be a very unstable parameter. Allen et al. (1998), Mota
(1989) and Stigter et al. (1986, 1989) have recommended avoiding the use of this
parameter, because it could lead to errors in the calculation of ET0 as high as 15%.
Furthermore, WMO (1983) states that the absolute accuracy of humidity data
from hygrographs in routine use remains difficult to evaluate as long as the stations
are not frequently visited for calibration purposes. In the Altiplano, the instability may
be further enhanced, due to the dry conditions and to the high variation in relative
humidity during the day. For that reason, alternative methods to estimate air humidity
were tested.
The use of dew point temperature (Tdew) for the determination of air humidity is highly
recommended. A correction of the relative humidity while estimating Tdew will be
proposed. The relation between vapor saturation pressure es and temperature (T) is
expressed by
 
17:27  T
es ¼ 0:611 exp (5)
T þ 237:3
128 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

To be saturated, the air must be cooled to the dew point (Tdew). Therefore, actual vapor
pressure ea is the saturation vapor pressure at Tdew, which can be calculated according to
Allen et al. (1998):
 
17:27  Tdew
ea ¼ 0:611 exp (6)
Tdew þ 237:3
In many cases, however, data for Tdew are difficult to obtain. Therefore, when humidity data
are either lacking or they are a source of error, the estimated actual vapor pressure (ea) can
be obtained by assuming that Tdew is close to the daily minimum temperature (Tmin). This is
the case at sunrise when air temperature is close to Tmin, and relative humidity is close to
100%. If Tmin is used to represent Tdew, then the equation becomes
 
17:27  Tmin
ea ¼ 0:611 exp (7)
Tmin þ 237:3
According to Allen et al. (1998), the Tmin  Tdew relationship is maintained in locations
where the cover crop of the agricultural field station is well watered, but in arid regions, the
air might not be saturated at Tmin, and Tmin is higher than Tdew. In this situation, Tdew could
be more accurately estimated using values of Tmin–3 8C in Eq. (7). Patacamaya is classified
as an arid region, and very rarely in summer the dew point is reached. Tdew data from two
meteorological stations in the area (Puno and Desaguadero) were compared with Tmin.
Given the similarities between Tdew and Tmin–3 8C in summer, the climatic input for the
calculation of ET0 was adjusted by estimating ea from the dew point, by assuming that Tdew
was equal to Tmin–3 8C. The correlation (R2 ¼ 0:8576) between the lysimeter data and the
calculated ET0 by using the adjusted climatic data improved significantly (Fig. 4).
It seems advisable to use measured or estimated Tdew data to determine ET0 with the
Penman–Monteith equation in arid or semiarid zones. However, the current calibration was
performed using only one representative point of the central part of the Altiplano, hence
further studies for other points of the Altiplano are required.

4.2. Crop evapotranspiration (ETc) and crop coefficient (Kc)

The total water use computed from the soil water balance was 450 mm per season, which
is slightly lower than the 500–650 mm per season, calculated for cereals such as oats and
barley in the Altiplano.
The calculated crop coefficients (Kc), derived from ET0 and ETc (Eq. (3)), are plotted in
Fig. 5. The Kc values are then adapted and defined following the development stages as in
Doorenbos and Pruitt (1986) and Allen et al. (1998): initial stage (Kc ini), crop devel-
opment stage, mid-season stage (Kc mid) and the late-season stage (Kc end).
At the beginning of the season (from emergence until 10% of ground cover) Kc is about
0.5. According to Doorenbos and Pruitt (1986), this value is a function of the frequency of
irrigation and atmospheric demand (ET0). The value of 0.5 corresponds well to the ET0
value of 4.9 mm per day with 4 days irrigation frequency.
The stage of vegetative growth is a dynamic growth period with a Kc value increasing to
a maximum of about 1.0 (Kc mid) at the beginning of the mid-season stage (flowering and
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 129

Fig. 4. Corrected reference evapotranspiration ET0, using ea from Eq. (7) (ET0 grass), compared to ET0 from
Penman–Monteith (ET0 P–M).

seed filling period) when it is a relatively constant value. During the late season, the Kc
decreases and reaches a value of 0.7 (Kc end) at harvest.
Kc values for mid season in cereals are generally higher (about 1.15) than those observed
here for quinoa (Allen et al., 1998; Tyagi et al., 2000). Low Kc values mean reduced water
requirements according to Eq. (2), that is the evapotranspiration of quinoa is less than that
of cereals. Irrigation projects intent mainly to supply water in quinoa’s most drought-
sensitive phase (mid season), and therefore the above data indicate that estimating
irrigation requirements for quinoa using cereal Kc would result in an overestimation of
the water demand by 15%.

Table 4
Crop coefficient (Kc) for different growth stages of quinoa, derived from calculated ET0 and measured
(lysimeters) ETc.

Growth stage Kc ini Vegetative Kc mid Kc end

Kc 0.52 Dynamic 1.00 0.70

Fig. 5. Crop coefficient Kc for quinoa derived from lysimeter data.


130 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

The crop coefficients can be defined for the first, third, and fourth (initial, mid, and late
season) growth periods, and used to construct the Kc curve (Fig. 5). The standardized Kc
values according to the FAO guidelines are presented in Table 4 (Doorenbos and Pruitt,
1986; Allen et al., 1998).

5. Crop water yields

From the data obtained for ETc and ETcadj (Fig. 3) and yield, Ky is calculated by means of
Eq. (3) and shown in Table 5. The seasonal Ky factor for quinoa under limited irrigation
conditions is 0.67, which is lower than for any other crop (Doorenbos and Khassam, 1979;
Allen et al., 1998). The lowest seasonal Ky values reported by Allen et al. (1998) are those
for groundnut and cotton crops, with 0.70 and 0.85, respectively. The derived Ky coefficient
is in accordance with results reported by Vacher (1998), who states that quinoa presents a
strategy to support the negative effects of drought stress through very low leaf water
potentials, and an optimization of the water use efficiency by a low gas exchange.
Furthermore, it should be noted that in the case of the sparsely irrigated field, yield
reduction in quinoa was due to stress during the flowering and seed filling periods, the most
drought-sensitive periods for most crops (Vazquez, 1995). In other experiments, least
drought resistance was reported during flowering (Mujica and Jacobsen, 1999; Jacobsen
et al., 1999).
The relative yield obtained in the plots (Table 5) shows that although quinoa is a
highly drought-resistant crop, the yield reduction in relation to drought stress may not be
completely linear. From the rainfed plots it was found that yield is severely reduced
under very low water supply, and that the water use efficiency is also reduced reaching
28% of the potential yield with 40% of the potential water consumption. After a certain
threshold value, the addition of relatively small amounts of water through supplemen-
tary irrigation would lead to significant increments in yield because in this case with
only 67% of the potential water consumption the crop would produce almost 80% of
the potential yield. This corresponds well with the findings of Andersen et al. (1996)
and Jensen et al. (2000), who demonstrated that leaf elasticity of quinoa is low,
having the same effect as described above. It also agrees with the model for water

Table 5
Crop evapotranspiration (ETc), relative evapotranspiration (ETcadj /ETc), yield, relative yield (Ya/Ym) and Ky for
quinoa in a well-watered lysimeter, in a sparsely irrigated and a rainfed field

Crop parameter Water supply

Optimal (lysimeter) Limited irrigation (field) Rainfed (field)


ET (mm per season) 450 305 183
ETcadj /ETc (%) – 67.8 40.1
Yield (kg ha1) 3700 2900 1053
Ya/Ym (%) 100 78.4 28.5
Ky 0.67 –
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 131

use efficiency of quinoa proposed by Vacher (1998), who showed that quinoa optimizes
carbon gains with a minimization of water loss, and it maintains high leaf water use
efficiency to compensate for the decrease in stomatal conductance. However, the water
use efficiency decreases when irrigation levels reach the potential water consumption
with an increment of only 20% in yield using 30% more water than in the moderately
irrigated fields. Based on the results presented, more studies are required to accurately
determine the specific characteristics of water consumption and yield in quinoa. This is
consistent with the results of Flynn (1990), who on sandy–loam soil in Colorado, USA,
found maximum yield efficiency with 208 mm of available water, better than 128, 307,
or 375 mm.

5.1. Net irrigation requirements (Inet)

The data on the effect of drought stress indicate that to maximize water use efficiency,
irrigation projects for quinoa should be designed to meet less than the full water demand,
also considering the few water resources available in the South American Andes. Table 6
and Fig. 6 show that some irrigation, especially in flowering and ripening phases, is desired
to ensure crop production in dry years or during drought spells. Comparing the irrigation
requirements to fulfill ETc and ETcadj , it is found that the water supply can be reduced with
about 100 mm, mainly concentrated from 80 to 130 days after sowing, when irrigating only
to the ETcadj . The corresponding net irrigation requirement for both levels is plotted in
Fig. 6, and the 10-daily balance is presented in Table 6. It is likely that in terms of a cost-
benefit analysis, a decrease in water supply of 100 mm (1000 m3 ha1) would not produce
a significant yield decrease.
In view of the above data, it seems valid to suggest that in the Altiplano, instead of trying
to irrigate small extensions of land intensively, irrigation projects should be based on a
limited water supply. This would increase the potential area for irrigation and the efficiency

Table 6
The 10-daily water balance for the determination of net irrigation requirements of quinoa to fulfill both ETc and
ETcadj

10-Daily periods ETc ETcadj 75% 10-Daily Net irrigation Net irrigation
after sowing dependable rainfall requirements (ETc) requirements (ETcadj )

3 28 26 0.0 28.0 26.0


4 29 28 9.0 20.0 19.0
5 31 29 23.4 7.6 5.6
6 33 30 9.5 23.5 20.5
7 37 32 8.1 28.9 23.9
8 43 35 2.7 40.3 32.3
9 47 37 9.2 37.8 27.8
10 47 35 0.7 46.3 34.3
11 42 21 1.1 40.9 19.9
12 36 21 0.0 36.0 21.0
13 34 20 0.0 34.0 20.0
Total 343.3 250.3
132 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

Fig. 6. Net irrigation requirements (Inet) to reach crop (ETc) and adjusted crop evapotranspiration (ETcadj ).

of the system, since maximum production per unit of water applied would be achieved
(Garcia, 1999). This suggestion, based on field experimental results, is in accordance with
the farmer’s conceptualization of irrigation, which is seen as an insurance for production
during periods of low rainfall. The effect of the proposed deficit irrigation is not believed to
affect negatively yield due to eventual rising salt levels in the quinoa cultivation area,
because quinoa is extremely tolerant to soil salinity (Jacobsen et al., 2001), and because of
the low groundwater level.

6. Conclusions

It was demonstrated that in the Altiplano, irrigation projects must be designed taking
into consideration:
Calculation of the reference evapotranspiration (ET0) by means of the Penman–
Monteith equation is only accurate when the input data is of high quality. The use
of dew point temperature to estimate air humidity is highly recommended. For the dry
conditions of the high Andean region, an excellent agreement between measured
(lysimeter) and calculated ET0 was obtained when assuming that Tdew was equal to
Tmin–3 8C during the summer season.
The derived Kc values for quinoa did not correspond to values for cereals confirming the
inadequacy of calculating irrigation requirements for quinoa using cereal crop coeffi-
cients.
The seasonal yield response factor Ky for quinoa was low (0.67), indicating that a minor
drought stress does not result in a large yield decrease. The actual overestimation of
water requirements in the Andean region has increased the costs of irrigation projects
and, in many cases, led to their rejection, since the potential gain in yield was insufficient
to justify the cost. It is recommended supplying a limited amount of water to a larger
area, instead of supplying the same amount of water to a limited number of well-watered
quinoa plots.
Experiments should be conducted to confirm values for Ky for quinoa, as well as to
estimate the fraction of the plant water, which can be absorbed by quinoa without
producing stress (p).
M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134 133

References

Allen, R.G., Howell, T.A., Pruitt, W.O., Walte, I.A., Jensen, M.E., 1991. Lysimeters for evapotranspiration and
environmental measurements. In: Proceedings of the International Workshop on Lysimetry. American
Society of Civil Engineering, Honolulu, HI.
Allen, R.G., Smith, M., Pereira, L., Perrier, A., 1994. An Update for the Definition of Reference
Evapotranspiration. International Commission on Irrigation and Drainage Bulletin. John Wiley and Sons
Ltd. 43 (2), 1–34.
Allen, R.G., Pereira, L., Raes, D., Smith, M., 1998. Crop evapotranspiration. FAO Irrigation and Drainage Paper
No. 56. Rome, Italy.
Altieri, M., 1997. Agroecologia y Desarrollo Sostenible. Curso auspiciado por el Consorcio Latinoamericano de
Desarrollo Sostenible. Facultad de Agronomia. UMSA, La Paz, Bolivia.
Andersen, S.D., Rasmussen, L., Jensen, C.R., Mogensen, V.O., Andersen, M.N., Jacobsen, S.-E., 1996. Leaf
water relations, photosynthesis and gas exchange of Chenopodium quinoa during drought in lysimeters. In:
Proceedings of the COST-Workshop, RVAU, Copenhagen, 22–24 February 1996, pp. 117–122.
Canqui, E., 1993. Determinación de la evapotranspiración potencial (ETP) por lisimetria en la Estación
Experimental de Patacamaya. Ing. Agr. Thesis, Oruro, Bolivia.
Choquecallata, J., 1993. Evapotranspiración maxima de quinua. Ing. Agr. Thesis, Santa Cruz, Bolivia.
Doorenbos, J., Khassam, A.H., 1979. Yield response to water. FAO Irrigation and Drainage Paper No. 33.
Rome, Italy.
Doorenbos, J., Pruitt, W.O., 1986. Crop water requirements. FAO Irrigation and Drainage Paper No. 24. Rome,
Italy.
FAO (Food and Agriculture Organization of the United Nations), 1998. Under-utilized Andean food crops. Latin
America and the Caribbean. Rome, Italy.
Flynn, R.O., 1990. Growth characteristics of quinoa and yield response to increase soil water deficit. M.Sc.
Thesis, Colorado State University, Fort Collins.
Garcia, M., 1992. Physiological responses of quinoa (Chenopodium quinoa) to drought in the Bolivian
Altiplano, Thesis Agronomist Engineer. UMSA, La Paz, Bolivia.
Garcia, M., 1994. Modeling soil–plant–water relationships of Andean crops in the Bolivian Altiplano. M.Sc.
Thesis, Agricultural University of Wageningen, The Netherlands.
Garcia, M., 1999. Rainfall and evapotranspiration analysis of the Bolivian Altiplano. M.Sc. Thesis,
Interuniversity Programme of Water Resources Engineering, Katolieque Universiteit Leuven, Belgium.
Hillel, D., 1984. Soil and Water Principles and Processes. Academic Press, NY, p. 288.
Jacobsen, S.-E., Mujica, A., 2001. Quinua: cultivo con resistencia a la sequı́a y otros factores adversos. In:
Jacobsen, S.-E., Mujica, A., Portillo, Z. (Eds.), Memorias, Primer Taller Internacional sobre Quinua—
Recursos Geneticos y Sistemas de Producción. CIP, UNALM, Lima, Peru, 10–14 May 1999, pp. 175–180.
Jacobsen, S.-E., Quispe, H., Mujica, A., 2001. An alternative crop for saline soils in the Andes. In: Scientist and
Farmer—Partners in Research for the 21st Century. CIP Program Report 1999–2000, pp. 403–408.
Jacobsen, S.-E., Nuñez, N., Stølen, O., Mujica, A., 1999. Que sabemos sobre la resistencia de la quinua a la
sequı́a? In: Jacobsen, S.-E., Mujica, A. (Eds.), Fisiologı́a de la Resistencia a Sequı́a en Quinua
(Chenopodium quinoa Willd.). CIP, Lima, Peru, pp. 65–69.
Jensen, C.R., Jacobsen, S.-E., Andersen, M.N., Nuñez, N., Andersen, S.D., Rasmussen, L., Mogensen, V.O.,
2000. Leaf gas exchange and water relations of field quinoa (Chenopodium quinoa Willd) during soil drying.
Eur. J. Agron. 13, 11–25.
Mariscal, A., 1992. Agroclimatologia. Universidad Autónoma Tomas Frias, Potosı́, Bolivia.
Michel, T., Garcia, M., 1997. Zonificación Agroclimática del Altiplano Paceño. Servicio Nacional de
Meteorologia e Hidrologia. La Paz, Bolivia.
Ministerio de Educación, 1994. I Censo Nacional de Talla en Escolares 1993. Ministerio de Educación,
UNICEF, PMA, y FONCODES. Lima, Peru, pp. 19–25.
Mota, F., 1989. Meteorologia Agrı́cola. Biblioteca Ruara, Libraria Nobel S/A.
Mujica, A., Jacobsen, S.-E., 1999. Resistencia de la quinua a la sequı́a y otros factores abioticos adversos, y su
mejoramiento. In: Jacobsen, S.-E., Mujica, A. (Eds.), Fisiologı́a de la Resistencia a Sequı́a en Quinua
(Chenopodium quinoa Willd.). CIP, Lima, Peru, pp. 71–78.
134 M. Garcia et al. / Agricultural Water Management 60 (2003) 119–134

Mujica, A., Jacobsen, S.E., Izquierdo, J., 2001. Resistencia a factores adversos de la quinua. In: Mujica, A.,
Jacobsen, S.-E., Izquierdo, J., Marathee, J.P. (Eds.), Quinua (Chenopodium quinoa Willd.)—Ancestral
cultivo andino, alimento del presente y futuro. FAO, UNA-Puno, CIP, Santiago, Chile, pp. 162–183.
Mungai, D.N., Stigter, C.J., 1993. The use of weather and climate information in planning and management of
food security and production in Africa under conditions of increasing climate variability. In: Proceedings of
the First International Conference on Recent Climate Anomalies and Prediction in Africa, Journal of the
African Meteorological Society, February 1993, Nairobi.
Repo-Carrasco, R., Espinoza, C., Jacobsen, S.-E., 2001. Valor nutricional y usos de la quinua y la kañiwa. In:
Jacobsen, S.-E., Mujica, A., Portillo, Z. (Eds.), Memorias, Primer Taller Internacional sobre Quinua—
Recursos Geneticos y Sistemas de Producción, 10–14 May 1999, UNALM, Lima, Peru. CIP, pp. 391–400.
Stigter, C.J., Mjungu, Y.B., Hyera, T.M., 1986. Accuracies of humidity data from Tanzanian stations and their
implications for evaporation calculations. East Afr. Agric. Forestry J. 51 (3), 131–140.
Stigter, C.J., Hyera, T.M., Mjungu, Y.B., 1989. Errors in routine humidity data from Tanzanian stations and their
consequences. In: Proceedings of the fourth WMO Technical Conference on Instruments and Observing
Methods. Report No. 35. WMO/TD. No. 303, Geneva.
Tyagi, N.K., Sharma, D.K., Lutra, S.K., 2000. Determination of evapotranspiration and crop coefficients of rice
and sunflower with lysimeter. Agric. Water Manage. 45 (1), 41–54.
Vacher, J., 1998. Responses of two main crops, quinoa (Chenopodium quinoa Willd) and papa amarga (Solanum
juzepczukii Buk.) to drought on the Bolivian Altiplano: significance of local adaptation. Agric. Ecosyst.
Environ. 68, 99–108.
Vacher, J., Imaña, E., Canqui, E., 1994. Las caracterı́sticas radiativas y la evapotranspiración potencial en el
Altiplano Boliviano. Revista de Agricultura. Facultad de Ciencias Agrı́colas, Pecuarias, Forestales y
Veterinarias, Universidad Mayor de San Simón.
Vazquez, T., 1995. Evaluación de cuarenta genotipos de quinoa (Chenopodium quinoa Willd) al déficit hı́drico
en cuatro fases fenológicas. M.Sc. Thesis, UNA, Puno, Perú.
WMO (World Meteorological Organization), 1983. Guide to Meteorological Instruments and Methods of
Observation, 5th ed. WMO-No. 8, Geneva.

View publication stats

You might also like