Agr. Nat. Resour.

55 (2021) 485–495

AGRICULTURE AND
NATURAL RESOURCES
Journal homepage: http://anres.kasetsart.org

Research article

Utilizing rainfed supply and irrigation as a climate variability adaptation

solution for coastal lowland areas in Vietnam
Dang Truong Ana, Nguyen Van Hongb,*, Mai Phuong Ngocc
a
University of Science, Vietnam National University, Ho Chi Minh 70000, Vietnam
b
Sub-Institute of Hydrometeorology and climate change, Ho Chi Minh 70000, Vietnam
c
School of Social Sciences Education, Vinh University, Nghe An 70000, Vietnam

Article Info Abstract

Article history: Globally, the agricultural sector continues to be affected by the impacts of climate
Received 7 March 2021
variability. Lowland regions have frequent drought and saline intrusion, leading to
Revised 30 June 2021
Accepted 30 June 2021 irrigation water deficits as a part of climate variability. This study focused on determining
Available online 30 June 2021 the rainfed use efficiency (RUE) and further establishing a crop cultivation calendar
(CCC) for rice paddies across Kien Giang province, Vietnam. The crop yield was
Keywords:
simulated in response to the relevant environmental factors based on a crop model. The
Drought,
Lowland, model performance was appraised using calibration and validation procedures and was
Rainfed, considered suitable based on good values for statistical indicators (index of agreement
Saline instruction,
= 0.74–0.89, coefficient of determination = 0.73–0.92 and root mean square error
Yield losses
approximately 11–20%). The results indicated that the rainfed use efficiency and crop
yield of rice planting seasons notably improved when the CPC (baseline) was altered,
specifically, when the first cropping season was around the first week of November,
which increased RUE (14.2–17.9%) and the crop yield (1.5–5.9%) at three experimental
locations, while the second cropping season in the fourth week of March decreased RUE
by 8.8–10.9% and increased the crop yield by up to 6.9 t/ha. An alteration in the CCC
of rice paddies across the study area accounted for relevant environmental factors and
produced an effective adaptation solution.

Introduction consequence of climate variability, leading to irrigation water

deficits (IWDs) (Abbas and Mayo, 2020). Climate variability
In recent decades, agricultural production around the increases the frequency and intensity of drought and saline
world has been facing challenges from adverse cultivation intrusions and has seriously affected all aspects of people’s
conditions such as drought, heatwaves and saline intrusion as a lives, especially in the agricultural sector (Intergovernmental
Panel on Climate Change, 2014). Research Centers in Southeast
* Corresponding author. Asia (2016) reported that climate variability is one of
E-mail address: [email protected] (N.V. Hong)
online 2452-316X print 2468-1458/Copyright © 2021. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/),
production and hosting by Kasetsart University of Research and Development Institute on behalf of Kasetsart University.

https://doi.org/10.34044/j.anres.2021.55.3.19
486 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495

the main challenges to human well-being in the 21st century. 2020) as well as WSS (Belder et al., 2004; Pascual and Wang,
A report by the Vietnamese Ministry of Natural Resources 2017; Hasan et al., 2019) are considered effective adaptation
and Environment showed that Kien Giang province is faced solutions and are broadly ongoing in many parts of the world
a lack of IWDs in lowland farming areas due to saline intrusion (Arimi, 2014; Bai and Xiao, 2020). These practice solutions
and drought events in 2016 (Dang, 2021; Lee and Dang, can help to reduce the negative ICV as well as contributing to
2018). Approximately 86,200 ha of the rice area in Kien maintaining crop yield and increasing profits.
Giang province was affected by drought and salinity with total Thus, the main objective of the current research focused
damage of approximately VND 2.35 billion during the drought on determining the RUE as well as establishing the CCC for
stages of 2014–2016 (Asia Pacific Network, 2010). lowland rice paddies across Kien Giang province based on the
In the context of IWDs for lowland agriculture due to Aquacrop model of the Food and Agriculture Organization of
the impacts of climate variability (ICV), studies on rainfed the United Nations (FAO). This model was applied to detect
use efficiency (RUE) have been widely deployed around the the interactions between environmental factors for a typical
world (Deryng et al., 2014; Bai and Xiao, 2020). For example, case to highlight the role of environmental factors for irrigation
in South and Southeast Asia, Zeigler and Puckridge (1995) activities as well as providing insight on links between CCC,
conducted a study on improving sustainable productivity in RUE and crop yield.
rainfed lowland rice. They reported that rice cultivation regions
are suitable for the application of water-saving solutions Materials and Methods
(WSSs). In central-northern India, Bouman and Tuong (2001)
deployed a study on WSSs for irrigating lowland rice and Study area
reported that irrigation water input reduced approximately 23%
with a crop yield decrease of 6%. Belder et al. (2004) assessed Kien Giang province is located southwestern Vietnam,
the effectiveness of WSSs on lowland rice and indicated (9°23’50”N–10°32’30”N and 104°26’40”E–105°32’40”E),
that using irrigation water could be reduced by up to 15% as shown in Fig. 1. Kien Giang province has diverse terrain
without affecting crop yield. Yang et al. (2007) established with a long coastline and many islands, mountains and rivers,
a WSS to increase crop yield in lowland paddies and their with the costal territory being relatively flat, with elevations in
results identified a WSS with savings of 24.5% to 29.2% the range 0.5–2.5 m above mean sea level (Nhan et al., 2011;
and increased rice yield of 7.4–11.3%. In the lower Mekong Son et al., 2014). The province has approximately 575.7 ha
Basin, Mainuddin et al. (2013) implemented a study on the of agricultural land of which 354.0 ha is in rice production,
rainfed rice yield under future climate scenarios. Their results accounting for 61.5% of the total agricultural land area (Trang,
suggested that altering the CCC could lead to increased crop 2016). Rice production is commonly based on irrigation from
yield. In Vietnam, Deb et al. (2015) investigated the ICV on the rivers combined with abundant rainfed supply (Mainuddin
rice yield in the lowland paddies of Ca Mau province and their et al., 2013), where the rainfed use efficiency is considered
results indicated that altering the CCC enhanced the crop yield as that part of the rainfall which is effectively utilized by the
in the context of climate variability. plants after subtracting lost rainfed volume due to runoff and
In recent years, the lowland rice paddies in Kien Giang percolation (Inthavong et al., 2014).
province have been exposed regularly to the potential risks of The climate in the province is affected from November to
climate change, leading to a decline in crop yield (Trang, 2016; April by the northeast monsoon circulation in the dry season and
Vu et al., 2018). Agricultural production has been strongly from May to October by the southwest monsoon circulation in
affected by climatic variables because the growth phases of the wet season (Auffhammer et al., 2011; Nguyen et al., 2015).
rice cultivation crops are closely involved with meteorological The area has a tropical monsoon climate, being hot and humid
factors (Asia Pacific Network, 2010; Nhan et al., 2011). with an average annual rainfall up to 2,146 mm and an average
Kien Giang is a coastal cultivation area where agricultural annual temperature varying from 26.4°C to 28.0°C. In general,
production is vulnerable to climate variability due to saline the climate has basic advantages as there are few natural
intrusion and drought, leading to a lack of IWDs. disasters such as tropical depressions and abundant sunshine
In the context of water scarcity for lowland agriculture, radiation, heat and rainfall (Table 1), so it is very favorable for
applying short-duration rice varieties (Kima et al., 2014; many kinds of plants (Food and Agriculture Organization of the
Won et al., 2020), altering the CCC (Sacks et al., 2010; Ding, United Nations, 2016; Kontgis et al., 2019).
 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495 487

Fig. 1 Location of study area in southwestern Vietnam

Table 1 Features of annual average rainfall (AAR) and standard deviation (SD) at rainfall observation stations in study area
Station AAR (mm) SD (mm) Latitude (N) Longitude (E)
Ha Tien 2677 608.2 10o22’00 105o30’14”
Kien Luong 2687 118.6 10°18'12” 104°38'18”
Rach Gia 2127 356.2 10°01´18” 105°05´26”
Chau Thanh 2481 281.5 09o53’00” 105o11’17”
Tan Hiep 2835 776.8 10o06’55” 105o17’00”

Agricultural production activities

In the province, farmers commonly plant single, double

or sometimes triple-cropped rice (Fig. 2A) depending on the
topographical characteristics and the supply of irrigation water
from the rivers to supplement the abundant local rainfed supply
(Kontgis et al., 2019). Growing and harvesting periods of crops
depend on local weather conditions and forecasts of saltwater
intrusion (Fig. 2B). In general, the CCC of the first cropping
season begins from the second week of November and the
second cropping season starts from the third week of March, Fig. 2 Rice cultivation regions in the study area: (A) rice cultivation systems
while the total cultivation area in the third cropping season is (Nguyen et al., 2015); (B) saline intrusion levels (Tran et al., 2019)
not substantial (Nguyen et al., 2015) and was not considered in
this study. crop models (Ding et al., 2020) but still balances simplicity,
accuracy and robustness (Shrestha et al., 2016). The model
Model description is easy to apply to practical problems based on integrating
four modules based on climate, soil, crop and management
The AquaCrop model was developed by FAO for simulating (Greaves et al., 2016).
attainable crop yield under different irrigation conditions In the AquaCrop model, the yield response to water
based on full irrigation and rainfed, supplemental and deficit expresses the relationship between crop yield and water stress
irrigation (Deb et al., 2015). The advantage of the AquaCrop because of insufficient irrigation water (rainfed or irrigation)
model is that it requires a comparatively small number of during the crop growing phases. Yield response to irrigation
parameters and intuitive input variables compared with other water is defined by Equation 1:
488 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495

(1) The crop yield (Y) is estimated as shown in Equation 7:

Y = fHIHI *B (7)
where Y x and Y are the maximum and actual crop
yields, respectively, ETx and ET are the maximum and where fHI is a multiplier which considers the stresses
actual evapotranspiration levels, respectively, and Ky is the that adjust the HI from its reference value. The adjustment of
proportionality factor between relative crop yield decline and the HI to irrigation water deficits depends on the timing and
relative reduction in evapotranspiration. ET in Equation 1 is extent of stress during the crop cycle (Steduto et al., 2012).
defined by Equation 2: In the AquaCrop model, water use efficiency (WUE) is based
on the relationship between crop yield and evapotranspiration
ET = E + Tr (2) (Steduto et al., 2012), as shown in Equation 8:

where E is the soil evaporation and Tr is the crop transpiration (8)

and Tr is obtained by multiplying the evaporating power of
the atmosphere by the crop transpiration coefficient (KcTr), WUE is considered as an indicator to assess the performance
water stress (Ks) and temperature stress (KsTr) as shown in of a system and it is applied to identify the management
Equation 3: strategies. According to Steduto et al. (2012), WUE is useful
under conditions of scarcity of water resources.
Tr = Ks * KsTr(KcTr,xCC*)ETo (3) The procedure for simulating WUE and crop yield are
illustrated in Fig. 3.
where, KcTr is calculated as shown in Equation 4:
Meteorological and soil data collection
KcTr = KcTr,x CC* (4)
Semiautomatic weather stations (Fig. 1) belonging to the
where KcTr,x is the proportional factor that integrates all the Southern Regional Hydrometeorological Center, Vietnam
effects of characteristics that distinguish the crop transpiration collected input data for the climate module based on: daily
from the grass reference surface and CC* is the fractional values of maximum (Tx) and minimum (Tn) air temperature,
canopy cover that is continuously adjusted to the simulated sunshine duration (SD), wind speed (WS) at 2 m height
canopy development. and relative humidity (RH) during 2011–2019 (Table 2).
In Equation 3, reference evapotranspiration (ETo) is defined Daily reference evapotranspiration (ETo) was obtained by
using the FAO Penman-Monteith equation (Greaves et al., applying the Penman-Monteith equation.
2016) as shown in Equation 5:

(5)

where Rn is the radiation at the soil surface, G is the soil
heat flux density, T is the average daily temperature, u2 is the
wind speed at 2.0 m height, es is the saturation vapour pressure,
ea is the actual vapour pressure, Δ is the slope of the vapour
pressure curve and γ is the psychrometric constant.
The aboveground biomass production (B) of the crop cycle
is defined by multiplying WP* by for that day, as
shown in Equation 6:

(6)
Fig. 3 Flow chart of AquaCrop model simulation
 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495 489

Table 2 Meteorological variables observed in study area for planting crop seasons
Year Tn (oC) Tx (oC) TR (mm) RH (%) SD (hr/d) WS (m/s) ETo (mm/d)
2011 20.3 33.4 2418.8 83 6.5 2.8 5.3
2012 20.7 33.6 1941.5 80 6.3 2.9 5.8
2013 20.2 33.1 2299.2 81 6.4 3.1 5.6
2014 21.1 33.7 1941.7 79 6.9 2.8 5.7
2015 19.5 33.2 1528.6 78 7.1 2.6 5.9
2016 19.4 32.7 2268.4 80 7.2 2.5 5.5
2017 19.0 32.9 2499.5 82 6.7 2.9 5.4
2018 20.7 32.7 2561.3 82 6.5 2.7 5.1
2019 18.0 32.9 2434.6 81 6.9 3.1 5.2
Average 19.9 33.2 2210.4 80 6.7 2.8 5.3
Tn, Tx = minimum and maximum air temperature, respectively; TR = total rainfall; RH = relative humidity; SD = sunshine duration, WS = wind speed;
ETo = reference evapotranspiration

The soil properties at the experimental locations were 180 kg/ha, respectively. In addition, OM2517 is known as
collected from field surveys and analyzed based on United a high yielding variety, with good pest resistance and is planted
States Department of Agriculture software (Gerakis and Baer, widely in the area. Detailed information about the amount of
2000; Saxton and Rawls, 2006). The results from the standard irrigation, observed rainfed supply and fertilizer rates (N, P, K)
laboratory procedures for soil sample analysis were classified are provided in Table 4.
as silty clay loam to silt loam and the percentages of each soil
are provided in Table 3. The average volumetric water content Results and Discussion
for a 0.6 m soil profile depth at saturation, field capacity
(FC), total available soil moisture (FC-WP), total available Model performance assessment
water (TAW), permanent wilting points (PWP), hydraulic
conductivity (K) and physicochemical properties (K+, Na+, Most studies apply calibration and validation procedures to
Ca2+, Mg2+ and pH) are presented in Table 3. appraise the performance of a model (Greaves and Wang, 2016;
With the subtropical climate in the study area, rice was Dang, 2021). In the current study, as a first step, the calibration
grown annually based on single-crop, double-crop rice procedure was conducted to compare the simulated model
paddies and a few areas of triple-crop rice paddies (Fig. 2A). and observed crop yield during the period 2000–2010 and
In general, the CCC of the first cropping season was the second step for validation was also carried out in the same
around 18–30 November and was harvested 20–28 February, way as the calibration procedure for the period 2011–2019.
while the second cropping season started 20–31 March The calibration and validation procedures were evaluated based
and was harvested 20–25 June. The short-duration variety on the indication of the index of agreement (d), the coefficient
OM2517 was grown with 85 d to 95 d length for the growth of determination (R2) and the root mean square error (RMSE),
cycle with the planting density varying from 160 kg/ha to as presented in Dang (2021).

Table 3 Soil profile characteristic of soil samples in 0.6 m depth in the study area
Three investigated locations
Determination
Ha Tien Kien Luong Rach Gia
Sand (%) 22 28 24
Silt (%) 61 54 51
Clay (%) 27 18 25
Soil features Silty clay loam Silt loam Silty clay loam
FC (% vol) 39 41 43
PWP (% vol) 24 15 17
K (mm/d) 15 2.0 2.7
K+(ppm) 19 17 21
Na+(ppm) 28 16 31
Ca2+(ppm) 45 38 29
Mg2+(ppm) 25 28 34
pH 5.6 5.8 5.4
FC = field capacity; PWP = permanent wilting point; K = hydraulic conductivity
490 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495

Table 4 Collected information on irrigation, rainfed supply and fertilizer rates for growth and development stages of rice plants across the study area
Location Crop Total observed rainfed supply Total irrigated water Fertilizer rate applied (kg/ha)
(mm) (mm)
I II III IV I II III IV I II III IV
Ha Tien First 30.8 202.8 34.9 23.4 50 20 60 50 40 60 50 30
Second 47.9 70.5 189.2 302.8 40 40 0 0 50 65 50 35
Kien Luong First 38.9 198.6 45.9 33.7 40 20 50 40 50 45 45 40
Second 53.2 85.4 177.5 287.6 40 40 10 0 60 60 50 40
Rach Gia First 27.5 193.2 16.3 16.9 60 10 70 60 40 50 50 30
Second 41.2 68.3 209.7 334.1 45 55 10 0 50 50 40 50
First = winter-spring crop season; Second = summer-autumn crop season; I = initial stage of rice crop; II = developmental stage of rice crop; III = middle
stage of rice crop; IV = late stage of rice crop.

Model calibration The index of agreement (d) is a measure of relative error

in the simulated results of the model and is a dimensionless
The AquaCrop model was calibrated based on the simulated quantity, with the model being considered a more perfect
results for observed crop yield during two planting crop seasons simulation as the value approaches 1 (Heng et al., 2009; Zeleke
at three different locations, covering the study area during the et al., 2011). The coefficient of determination (R2) is a measure
period 2000–2010. The results produced high correlations with of the mean square to the observed variance. The coefficient
values for d, R2 and RMSE in the ranges 0.77–0.89, 0.81–0.92 of determination designates the overall deviation between
and 11–20%, respectively, for the two planting crop seasons at observed data and simulated results from the overall deviation
the experimental locations across the study area (Figs. 4A–4B). between observed data with its average value. The simulated
Specifically, at Ha Tien station, the values for R2, d and results are considered good if the value of R2 is close to 1
RMSE between the simulated model and observed crop yield (Heng et al., 2009; Babel et al., 2018). The RMSE represents
were d = 0.83–0.86, R2 = 0.85–0.87 and RMSE = 11–17%, a measure of the overall deviation between observed data and
while at Kien Luong station, the corresponding values were simulated results. It also is considered as a synthetic indicator
0.77–0.83, 0.81–0.85 and 13–19% and at Rach Gia station, the of absolute model uncertainty and the performance of a model
corresponding values were 0.85–0.89, 0.87–0.92 and 12–20% is considered good if the value of RMSE approaches zero
(Table 5). (Dang, 2021). The calibrated results of the high d and R2 values
and the low percentage of RMSE of the models for the three
observed locations across the study area supported the model
calibration was acceptable.

Fig. 4 Relationship between simulated model and observed crop yield at three experimental locations across study area: (A) first cropping season;
(B) second cropping season, where dotted lines are plots of y = x, d is the index of agreement, R2 is the coefficient of determination and RMSE
is the relative root mean squared error.
 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495 491

Table 5 Performance results of AquaCrop model in calibration and validation processes based on statistical analysis
Location Crop Calibration Validation
d R2 RMSE d R2 RMSE
Ha Tien First 0.86 0.87 11% 0.84 0.82 13%
Second 0.83 0.85 17% 0.78 0.78 18%
Kien Luong First 0.83 0.85 13% 0.82 0.74 15%
Second 0.77 0.81 19% 0.75 0.80 19%
Rach Gia First 0.89 0.92 12% 0.79 0.76 14%
Second 0.85 0.87 20% 0.74 0.73 17%
First = winter-spring crop; Second = summer-autumn crop; d = index of agreement; R2 = coefficient of determination; RMSE = relative root mean
squared error

Model validation The results indicated that the RUE would notably improve
if the CCC of both the first and second cropping seasons
Model validation was also deployed by comparing the were altered compared to the CCC (Figs. 5A–5B). In the first
simulated results and observed crop yields for two planting cropping season, the total RUE would increase on average
crop seasons at three experimental locations covering 16.6%, 17.9% and 14.2% at Ha Tien, Kien Luong and Rach Gia
the study area in the period 2011–2019. The validation process stations, respectively, (Table 6) when the CCC is commences
of the AquaCrop model run with the calibrated parameters in the first week of November while the corresponding values
produced good results with averages of d = 0.78, R2 = 0.77 for the second cropping season would decrease by 8.8%, 10.8%
and RMSE = 15% (Table 5). Specifically, the statistical and 8.9% (Table 6) when the CCC commences in the fourth
indicators between the simulated results and observed week of March.
crop yields for both planting crop seasons had values of Specifically, the RUE increased steadily in the initial,
d = 0.78–0.84, R 2 = 0.78–0.82 and RMSE = 13–18% at development, middle and late stages of the first cropping
Ha Tien station, values of d = 0.75–0.82, R2 = 0.74–0.80 season with increases in the ranges at the three experimental
and RMSE = 15–19% at Kien Luong station and values locations being 10.6%–14.4%, 12.8%–17.5%, 17.0–23.1% and
of d = 0.74–0.79, R 2 = 0.73–0.76 and RMSE = 14–17% 29.0%–35.0%, respectively, (Table 6). This is essential for the
at Rach Gia station (Table 5). The goodness of fit between normal growth of rice plants because the regular distribution of
the simulated results and observed crop yield based on the rainfed supply for the growth and development stages of rice
validation process confirmed the applicability of the AquaCrop plants will help to avoid a lack of supply of irrigation water
model for determining RUE and crop yields across the study during the crop’s growth cycle, especially during the grain
area. filling or harvesting stages, as this would seriously affect crop
yield.
Crop cultivation periods and rainfed use efficiency In the second cropping season, when the CCC was adjusted
forward 1 wk hydrologically, there were only minor changes
For years, the rice paddies have been sown without in RUE compared to the corresponding baseline values with
considering adverse environmental conditions such as the 6.8–9.7% and 10.0–11.0% for initial and middle stages,
effects of irrigation water salinity, drought during the growth respectively, while there were major increases in RUE in the
stage and high intensity rainfall during flowering, grain filling development stage and a major decrease in the late stage of the
stages and the harvesting stage, resulting in reduced crop yield. rice plants (Figs. 6A–6B). This was most suitable for the grain
According to Deb et al. (2015), increasing salinity in irrigation filling process as well as harvesting activities, by avoiding
water during the growing and developing stages of rice can a high concentration of rainfed supply during the harvest stage,
lead to decreased crop yield. which could easily result in crop losses. According to Dang
(2021), the crop yield would significantly reduce if the rice
plants were subjected to daily high intensity rainfed supply
during the flowering and harvesting phases.
492 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495

Fig. 5 Comparison of rainfed and rainfed use efficiency at three experimental locations after adjusting crop cultivation calendar: (A) in first cropping
season; (B) in second cropping season

Table 6 Comparison of rainfed use efficiency before and after adjusted crop cultivation calendar corresponding to growth and development stages of rice
planting crops across study area
Stage First cropping season Second cropping season
Ha Tien Kien Luong Rach Gia Ha Tien Kien Luong Rach Gia
Rainfed use efficiency before adjusting crop cultivation calendar
Initial 21.6 27.7 19.6 39.1 44.2 35.3
Growth 144.1 154.3 150.1 56.6 65.5 52.4
Middle 26.6 33.0 11.7 146.3 128.0 151.2
Late 15.7 29.3 14.7 222.0 199.9 232.2
Total 208.0 244.3 196.1 464.0 437.6 471.1
Rainfed use efficiency after adjusting crop cultivation calendar
Initial 23.8 31.0 22.3 45.7 49.5 40.5
Growth 165.7 180.5 168.1 63.4 69.6 56.3
Middle 31.9 38.6 14.4 160.9 139.5 167.8
Late 21.2 37.8 19.1 151.0 125.9 155.6
Total 242.6 288.0 224.0 421.0 399.5 441.3
I (%) 10.6 12.7 14.4 6.8 9.7 9.2
II (%) 15.3 17.5 12.8 22.6 16.9 28.5
III (%) 20.4 17.0 23.1 10.0 9.0 11.0
IV (%) 35.0 29.0 30.0 -32.0 -37.0 -33.0
Change in total RUE (%) +16.6 +17.9 +14.2 -8.8 -10.8 -8.9
I, II, II and IV = percentage change in initial, growth, middle and late stages, respectively, of rainfed use efficiency (RUE) before and after adjusted crop
cultivation calendar; change in total RUE = change in total rainfed use efficiency
 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495 493

Fig. 6 Trend in crop yield increase at three experimental locations (Ha Tien, Kien Luong and Rach Gia) across study area corresponding to appropriate
crop cultivation calendars for (A) first cropping season and (B) second cropping season

Crop cultivation periods and crop yield daily rainfed supply of approximately 1.6 mm, 4.6 mm, 2.4
mm and 4.6 mm for the initial, growth, middle and late stages,
The results showed that the crop yield of rice cultivation respectively, and enhanced the optimal crop yield by up to
paddies improved substantially if the flowering and harvesting 7.1 t/ha at the Ha Tien, Kien Luong and Rach Gia experiment
stages of rice crops receive an appropriate amount of daily stations (Table 7).
rainfed supply. Specifically, for the first cropping season, When the CCC was delayed by one hydrological week,
a shift to two hydrological weeks earlier resulted in the rice the rice paddies received high daily rainfed supply during
cultivation paddies receiving an average daily rainfed supply the development and middle stages but received lower daily
of approximately 1.3 mm, 5.7 mm, 0.9 mm and 1.7 mm for rainfed supply in the late stage, which suited harvest activities
the initial, growth, middle and late stages, respectively, and as well as reducing yield losses due to flooding and rice plants
a maximum increase in crop yield of approximately 5.9% falling over due to high intensity rainfed supply.
(Fig. 6A). The results showed that a shift to two hydrological Studies on crop yield optimization solutions in the Plain of
weeks earlier in the CCC was consistent with actual Reeds and Ca Mau Peninsula regions of Vietnam by Dang (2021)
environmental factors in the study area. Similarly, for the and Deb et al. (2015) reported that the crop yield of the second
second cropping season, a delay of one hydrological week cropping season declined if the grain filling and harvesting
resulted in the rice cultivation paddies receiving an average stages were subjected to high intensity rainfed supply.

Table 7 Crop yield corresponding to different crop cultivation times for two planting crop seasons at three experimental locations
Period Ha Tien Kien Luong Rach Gia
Crop yield Increased rate Crop yield Increased rate Crop yield Increased rate
(t/ha) (%) (t/ha) (%) (t/ha) (%)
First crop season
Current 6.8 - 6.5 - 6.6 -
280 7.0 2.9 6.7 3.1 6.8 3.0
270 7.3 5.9 6.9 6.2 6.9 4.5
260 7.0 2.9 6.6 1.5 6.7 1.5
Second crop season
Current 6.3 5.9 6.2
85 6.7 6.1 6.2 5.6 6.6 6.5
95 6.9 4.6 6.5 4.3 6.9 4.2
105 7.1 3.0 6.7 2.9 7.1 3.4
Current = current crop yield, where 280, 270, 260, 85, 95 and 105 d are crop planting times according to calendar dates.
494 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495

The current research assessed the impacts of relevant Belder, P., Bouman, B.A.M., Cabangon, R.J., Quilang, E.J.P., Li, Y.,
environmental factors on rice cultivation paddies in Kien Spiertz, J.H.J., Tuong, T.P. 2004. Effect of water saving irrigation
on rice yield and water use in typical lowland conditions in Asia.
Giang province. The results indicated that the crop yield of two
Agric. Water Manag. 65: 193–210. doi.org/10.1016/j.agwat.2003.
planting crops can be considerably affected by environmental 09.002
factors as a part of climate variability. The applicability of Bouman, B.A.M., Tuong, T.P. 2001. Field water management to save
the AquaCrop model for simulating the crop features was water and increase its productivity in irrigated lowland rice. Agric.
confirmed through the good statistical test results supporting its Water Manag. 49: 11–30. doi.org/10.1016/S0378-3774(00)00128-1
robustness. The study revealed that a shift to two hydrological Dang, T.A. 2021. Grain yield optimisation in the plain of reeds in the
context of climate variability. Rev. Bras. Eng. Agric. Ambient. 25:
weeks earlier in the crop cultivation calendar of the first
591–596. doi.org/10.1590/1807-1929/agriambi.v25n9p591-596
cropping season enhanced rainfed use efficiency as well as Deb, P., Tran, D.A., Udmale, P.D. 2015. Assessment of the impacts of
the optimal crop yield while both rainfed use efficiency and climate change and brackish irrigation water on rice productivity
the crop yield of the second cropping season achieved optimal and evaluation of adaptation measures in Ca Mau province, Vietnam.
values if the crop cultivation calendar of the second cropping Theor. Appl. Climatol. 125: 641–656. doi.org/10.1007/s00704-015-
1525-8
season was delayed by about one hydrological week. The crop
Deryng, D., Conway, D., Ramankutty, N., Price, J., Warren, R. 2014.
cultivation calendar of the two rice cropping seasons needed
Global crop yield response to extreme heat stress under multiple climate
to be altered to adapt to adverse cultivation conditions as well change futures. Environ. Res. Lett. 9: 034011. doi.org/10.1088/1748-
as contributing to optimizing crop yields for rice cultivation 9326/9/3/034011
paddies in the study area. Ding, Y.M., Wang, W.G., Zhuang, Q.L., Luo, Y.F. 2020. Adaptation of
paddy rice in China to climate change: The effects of shifting sowing
date on yield and irrigation water requirement. Agric. Water Manag.
Conflict of Interest
228: 105890. doi.org/10.1016/j.agwat.2019.105890
Food and Agriculture Organization of the United Nations (FAO). 2016. El
The authors declare that there are no conflicts of interest. Niño event in Viet Nam: Agriculture, food security and livelihood need
assessment in response to drought and salt-water intrusion. Assessment
Acknowledgements Report. Ha Noi, Vietnam. http://www.fao.org/3/i6020e/i6020e.pdf,
8 July 2021.
Gerakis, A., Baer, B. 2000. Computer program for soil textural
The reviewers provided helpful comments.
classification. Soil Sci. Soc. Am. J. 63: 807–808. doi.org/10.2136/
sssaj1999.634807x
References Greaves, G.E., Wang, Y.M. 2016. Assessment of FAO AquaCrop model
for simulating maize growth and productivity under deficit irrigation
Abbas, S., Mayo, Z.A. 2020. Impact of temperature and rainfall on rice in a tropical environment. Water 8: 557. doi.org/10.3390/w8120557
production in Punjab, Pakistan. Environ. Dev. Sustain. 23: 1706–1728. Hasan, M.R., Nuruzzaman, M., Mamun, A.A. 2019. Contribution of
doi.org/10.1007/s10668-020-00647-8 rainwater to the irrigation requirement for paddy cultivation at
Arimi, K. 2014. Determinants of climate change adaptation strategies used Tanore Upazila in Rajshahi, Bangladesh. Air Soil Water Res. doi.
by rice farmers in South-Western Nigeria. J. Agric. Rural Dev. Trop. org/10.1177/1178622119837544
Subtrop. 115: 91–99. Heng, L.K., Hsiao, T., Evett, S., Howell, T., Steduto, P. 2009. Validating
Asia Pacific Network. 2010. Climate change in Southeast Asia and the FAO AquaCrop model for irrigated and water deficient field maize.
assessment on impact, vulnerability and adaptation on rice production Agron. J. 101: 488–498. doi.org/10.2134/agronj2008.0029xs
and water resource. Project Reference Number: CRP2008-03CMY- Intergovernmental Panel on Climate Change. 2014. Climate Change
Jintrawet. Chiang Mai, Thailand. http://startcc.iwlearn.org/doc/ 2014: The Synthesis Report of the Fifth Assessment Report of
Doc_eng_10.pdf, 8 July 2021. the Intergovernmental Panel on Climate Change. Cambridge
Auffhammer, A., Ramanathan, V., Vincent, J.R. 2011. Climate change, the University Press. Cambridge, UK. https://www.ipcc.ch/site/assets/
monsoon, and rice yield in India. Clim. Change. 111: 411–424. doi.org/ uploads/2018/05/SYR_AR5_FINAL_full_wcover.pdf, 8 July 2021
10.1007/s10584-011-0208-4 Inthavong, T., Fukai, S., Tsubo, M. 2014. Estimation of separate effects
Babel, M.S., Deb, P.L., Soni, P.Y. 2018. Performance evaluation of of water and nutrient limitation for rainfed lowland rice within
AquaCrop and DSSAT-CERES for maize under different irrigation and a province in the Mekong region. Field Crops Res. 163: 100–108. doi.
manure application rates in the Himalayan region of India. Agric. Res. org/10.1016/j.fcr.2014.03.020
8: 207–217. doi.org/10.1007/s40003-018-0366-y Kima, A.S., Chung, W.G., Wang, Y.M. 2014. Improving irrigated lowland
Bai, H., Xiao, D. 2020. Spatiotemporal changes of rice phenology in rice water use efficiency under saturated soil culture for adoption in
China during 1981–2010. Theor. Appl. Climatol. 140: 1483–1494. tropical climate conditions. Water 6: 2830–2846. doi.org/10.3390/
doi.org/10.1007/s00704-020-03182-8 w6092830
 D.T. An et al. / Agr. Nat. Resour. 55 (2021) 485–495 495

Kontgis, C., Schneider, A., Ozdogan, M., Kucharik, C., Pham, V.D.T., Steduto, P., Hsiao, T.C., Raes, D. 2012. Crop Yield Response to Water.
Nguyen, H.D., Schatz, Z. 2019. Climate change impacts on rice FAO Irrigation and Drainage Paper no. 66. FAO. Rome, Italy.
productivity in the Mekong River Delta. Appl. Geogr. 102: 71–83. doi. Tran, D.D., van Halsema, G., Hellegers, P.J.G.J., Hoang, L.P. 2019.
org/10.1016/j.apgeog.2018.12.004 Long-term sustainability of the Vietnamese Mekong Delta in question:
Lee, S.K., Dang, T.A. 2018. Spatio‑temporal variations in meteorology An economic assessment of water management alternatives. Agric.
drought over the Mekong River Delta of Vietnam in the recent decades. Water Manag. 223: 105703. doi.org/10.1016/j.agwat.2019.105703
Paddy Water Environ. 17: 35–44. doi.org/10.1007/s10333-018-0681-8 Trang, T.H.L. 2016. Effects of climate change on rice yield and rice market
Mainuddin, M., Kirby, M., Hoanh, C.T. 2013. Impact of climate change on in Vietnam. J. Agric. Appl. Econ. 48: 366–382. doi.org/10.1017/
rainfed rice and options for adaptation in the lower Mekong Basin. Nat. aae.2016.21
Hazards. 6: 905–938. doi.org/10.1007/s11069-012-0526-5 Vu, D.T., Yamada, T., Ishidaira, H. 2018. Assessing the impact of sea
Nguyen, D.B., Clauss, K., Cao, S., Naeimi, V., Kuenzer, C., Wagner, W. level rise due to climate change on seawater intrusion in Mekong
2015. Mapping rice seasonality in the Mekong Delta with multi-year Delta, Vietnam. Water Sci. Technol. 77: 1632–1639. doi.org/10.2166/
Envisat ASAR WSM data. Remote Sens. 7: 15868–15893. doi.org/ wst.2018.038
10.3390/rs71215808 Won, P.L.P., Liu, H.Y., Banayo, N.P.M., et al. 2020. Identification and
Pascual, V.J., Wang, Y.M. 2017. Utilizing rainfall and alternate wetting characterization of high-yielding, short-duration rice genotypes for
and drying irrigation for high water productivity in irrigated lowland tropical Asia. Crop Sci. 60: 2241–2250. doi.org/10.1002/csc2.20183
paddy rice in southern Taiwan. Plant Prod. Sci. 20: 24–35. doi.org/10. Yang, J., Liu, K., Wang, Z., Du, Y., Zhang, J. 2007. Water-saving and
1080/1343943X.2016.1242373 high-yielding irrigation for lowland rice by controlling limiting values
Research Centers in Southeast Asia. 2016. The drought and salinity of soil water potential. J. Integr. Plant Biol. 49: 1445–1454. doi.org/
intrusion in the Mekong River Delta of Vietnam. Assessment Report. 10.1111/j.1672-9072.2007.00555.x
Ben Tre, Tra Vinh, Kien Giang, Vietnam. file:///C:/Users/User/ Zeigler, R.S., Puckridge D.W. 1995. Improving sustainable productivity
Downloads/CGIAR%20ASSESSMENT%20REPORT_Mekong_ in rice-based rainfed lowland systems of South and Southeast Asia.
June2.pdf, 8 July 2021. Geojournal 35: 307–324. doi.org/10.1007/BF00989138
Saxton, K.E., Rawls, W.J. 2006. Soil water characteristic estimates by Zeleke, K.T., Luckett, D., Cowley, R. 2011. Calibration and testing
texture and organic matter for hydrologic solutions. Soil Sci. Soc. Am. of the FAO AquaCrop model for canola. Agron. J. 103: 1610–1618.
J. 70: 1569–1578. doi.org/10.2136/sssaj2005.0117 doi.org/10.2134/agronj2011.0150
Shrestha, S., Deb, P., Bui, T.T.T. 2016. Adaptation strategies for rice
cultivation under climate change in Central Vietnam. Mitig. Adapt.
Strateg. Glob. Chang. 21: 15–37. doi.org/10.1007/s11027-014-9567-2