Enabling modeling of waterlogging impact on wheat

Rogério de S. Nóia-Júnior a,b , Valentina Stocca a, Pierre Martre b, Vakhtang Shelia c,  
Jean-Charles Deswarte d, Jean-Pierre Cohan e, Benoît Piquemal e, Alain Dutertre e,  
Gustavo A. Slafer f,g, Zhentao Zhang h, Marijn Van Der Velde i, Yean-Uk Kim j, Heidi Webber j,k,  
Frank Ewert j,l, Taru Palosuo m, Ke Liu n,o, Matthew Tom Harrison n, Gerrit Hoogenboom d,  
Senthold Asseng a,*

a Technical University of Munich, Department of Life Science Engineering, Digital Agriculture, HEF World Agricultural Systems Center, Freising, Germany
b LEPSE, Univ Montpellier, INRAE, Institut Agro Montpellier, Montpellier, France
c Department of Agricultural and Biological Engineering, University of Florida, Gainesville, FL, USA
d ARVALIS, Villiers-le-Bâcle, France
e ARVALIS, Loireauxence, France
f Department of Agricultural and Forest Sciences and Engineering, University of Lleida – AGROTECNIO-CERCA Center, Lleida, Spain
g ICREA, Catalonian Institution for Research and Advanced Studies, Barcelona, Spain
h College of Resources and Environmental Sciences, China Agricultural University, Beijing, China
i European Commission, Joint Research Centre, Ispra, Italy
j Leibniz-Centre for Agricultural Landscape Research (ZALF), Müncheberg, Germany
k Brandenburg Technical University (BTU), Cottbus, Germany
l Crop Science Group, INRES, University of Bonn, Bonn, Germany
m Natural Resources Institute Finland (Luke), Helsinki, Finland
n Tasmanian Institute of Agriculture, University of Tasmania, Newnham, Launceston, Tasmania, Australia
o MARA Key Laboratory of Sustainable Crop Production in the Middle Reaches of the Yangtze River (Co-construction by Ministry and Province) /School of Agriculture, 
Yangtze University, Jingzhou, China

A R T I C L E  I N F O

Keywords:
NWheat
DSSAT
Excess of water
Grain number
Grain size
Wheat crop simulation model

A B S T R A C T

Most crop simulation models do not consider the effect of waterlogging despite its importance for crop perfor
mance. Here, we reviewed the impact of waterlogging during different wheat phenological stages on grain 
number per unit area, average grain size, and grain yield. Episodes of waterlogging from the onset of tillering to 
anthesis result in fewer, and during grain filling in lighter grains. To simulate such impacts, we implemented a 
new waterlogging module into the wheat crop simulation model DSSAT-NWheat, accounting for the effects of 
waterlogging on wheat root growth, biomass growth, and potential average grain size. The model incorporating 
the new waterlogging routine was tested using data from a controlled experiment, and it reasonably reproduced 
wheat yield responses to pre-anthesis waterlogging. A sensitivity analysis showed that the simulated impact of 
waterlogging on above ground biomass and roots, as well as leaf area index, grain number, and grain yield varied 
with phenological stages. The simulated crop was most sensitive to pre-anthesis waterlogging, consistent with 
experimental studies. The new waterlogging-enabled crop model is an initial attempt to consider the impact of 
excess rainfall and waterlogging on crop growth and final grain yield to reduce model uncertainties when 
projecting climate change impacts with increasing rainfall intensity.

1. Introduction

Waterlogging could be defined as the phenomenon in which pro
longed saturation of a soil with water inhibits or entirely negates oxygen 
availability to plant roots (Liu et al., 2021a; Kim et al., 2024). All direct 

and indirect impacts of waterlogging together are a major cause of crop 
yield losses across the world, estimated to affect 15–20 % of the global 
wheat cropping area each year (Sayre et al., 1994; Kaur et al., 2020). 
Crop production in southern Asia, many parts of Europe, Russia, China 
and southern Brazil may be more sensitive to waterlogging than to 

* Corresponding author.
E-mail address: senthold.asseng@tum.de (S. Asseng). 

Contents lists available at ScienceDirect

Field Crops Research

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

https://doi.org/10.1016/j.fcr.2025.110090
Received 29 October 2024; Received in revised form 6 June 2025; Accepted 26 July 2025  

Field Crops Research 333 (2025) 110090 

Available online 29 July 2025 
0378-4290/© 2025 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). 

https://orcid.org/0000-0002-4096-7588
https://orcid.org/0000-0002-4096-7588
mailto:senthold.asseng@tum.de
www.sciencedirect.com/science/journal/03784290
https://www.elsevier.com/locate/fcr
https://doi.org/10.1016/j.fcr.2025.110090
https://doi.org/10.1016/j.fcr.2025.110090
http://crossmark.crossref.org/dialog/?doi=10.1016/j.fcr.2025.110090&domain=pdf
http://creativecommons.org/licenses/by/4.0/


drought (Zampieri et al., 2017; Liu et al., 2023). Waterlogging from high 
rainfall has damaged 33.9 million hectares (Mha) of India’s arable area 
between 2015 and 2022 (Kulkarni et al., 2021). In Europe, excessive 
precipitation and waterlogging have been identified as major factors 
limiting wheat yield (Ceglar et al., 2020; Nóia Júnior et al., 2023a). In 
southern South America, prolonged waterlogging events have also 
caused severe yield losses, including a 40 % reduction in Brazil in 2017 
and recurrent flooding across over 30 % of Argentina’s Pampas Region 
in recent decades (Nóia Júnior et al., 2021; Kuppel et al., 2015). Despite 
the large worldwide damage of waterlogging on crops, studies on 
waterlogging impact on grain production are scarce.

The process of adapting agricultural crops to extreme weather 
events, such as waterlogging, demands sophisticated understanding of 
how crop genotypes respond to different weather conditions, at different 
phenological stages for different management options (Githui et al., 
2022; Liu et al., 2020a). For this, it is necessary to have controlled ex
periments in greenhouse and field conditions, with physiological, 
morphological and growth measurements. These experiments may be 
repeated in different years and locations, making them costly and often 
unfeasible, particularly if a thorough insight into the solution space is 
sought (Liu et al., 2020b, 2023; Githui et al., 2022).

Crop models codify the gathered knowledge and understanding on 
crop growth processes and contribute to the understanding of the 
physiological mechanisms of plant response to abiotic stresses. They are 
widely applied for simulating crop growth and physiology in different 
environments, locations and years, as well as for climate change sce
narios, assessing extreme weather impacts (Yan et al., 2022; Kassie et al., 
2016). Indeed, many studies that project the impacts of climate change 
on agriculture used crop simulation models (Asseng et al., 2015, 2013; 
Liu et al., 2016; Zhao et al., 2017; Liu et al., 2019; Asseng et al., 2019). 
Yet, the majority of crop models do not consider the impact of water
logging on crop growth (Liu et al., 2020b; Githui et al., 2022).

Waterlogging is frequently disregarded in crop models, with only 
one-third of wheat crop models accounting for its effects (Nóia Júnior 
et al., 2023b). Even when considered, these crop models usually only 
focus on how it affects carbohydrate accumulation, radiation use effi
ciency, transpiration, root activity, and leaf area index (LAI) (Liu et al., 
2020c; Githui et al., 2022; Nóia Júnior et al., 2023b). In some crop 
models, the magnitude of simulated impact varies according to the 
phenological phase in which the waterlogging occurs (e.g., De San 
Celedonio et al., 2014). The wheat crop model in APSIM, for example, 
considers that the highest photosynthetic and roots activity reduction 
due to waterlogging occurs before wheat grain filling, with no impact 
during grain filling (Liu et al., 2021b). The EPIC model considers direct 
waterlogging effects only on photosynthesis and LAI (Githui et al., 
2022), with indirect impacts on other growth mechanisms. In these crop 
models, wheat growth is penalized via lower photosynthesis due to its 
direct reduction or lower light capture from decreased leaf area with 
waterlogging. Simulated waterlogging may indirectly impact nitrogen 
uptake via reduced root growth, resulting in a simulated yield penalty. 
However, these models have only been tested in a few situations with 
waterlogged wheat, and it is not clear how waterlogging affects yield 
and its components, and by extension, yield. Some wheat crop models do 
not correctly simulate the relationship between number of grains per 
unit area and average grain size (or grain unit weight), resulting in 
overestimation of simulated yields, particularly in seasons with poor 
grain setting (causing low grain number per unit area) and grain filling 
(causing low average grain size) (Nóia Júnior et al., 2023c). The rela
tionship between number of grains per unit area and average grain size 
is usually studied in controlled environments with few measured data 
(Calderini et al., 2001; Fischer, 1985; Asseng et al., 2017), making it 
difficult to improve wheat crop models.

The inability of most crop models to simulate waterlogged wheat 
growth potentially underestimates the projected impacts of climate 
change on agriculture (Van der Velde et al., 2020; Webber et al., 2020). 
To improve the representation of waterlogging in crop models, in this 

study we aimed to improve the waterlogging module in the wheat crop 
simulation model DSSAT-NWheat. For this, we reviewed published ar
ticles from 2008 to 2021 that studied the impacts of waterlogging during 
different wheat phenological stages on grain number per unit area, 
average grain size, and grain yield. Based on the general relationships of 
crop performance and waterlogging, a new waterlogging module was 
developed. This module was tested using data from a controlled 
waterlogging experiment in Lleida, Spain and from wheat field experi
ments in Loireauxence, France with side-by-side drained and undrained 
soils.

2. Material and methods

2.1. Review of waterlogging impacts on wheat grain number, average 
grain size and grain yield

We assessed 17 peer-reviewed articles published between 2008 and 
2021 which quantified the impacts of waterlogging on wheat grain 
yield, grain number or grain size in controlled waterlogging treatments 
(Supplementary Table S1). Relevant articles were found by using key
words ‘waterlogging’, ‘flooding’ or ‘excess of water’ and ‘wheat’ in Web 
of Science and Google Scholar Databases. In the assessed articles, 
waterlogging affected different wheat phenological stages, from seed
ling emergence to onset of grain filling and with waterlogging duration 
varying from 2 to 58 days (Fig. 3d) with fully waterlogged soil (i.e., no 
variations in soil moisture along the soil profile depth). From the 
assessed articles, we collected the average grain yield, average grain size 
(or grain unit dry mass or thousand grain dry mass) and grain number 
per unit area per each waterlogging treatments and their control. With 
these values, we calculated the relative change of each treatment in 
relation to the control and averaged according to the wheat phenolog
ical stage. Experiments with waterlogging starting at seven different 
wheat phenological stages were collated, namely at wheat seedling 
emergence, onset of tillering, jointing, booting, heading, anthesis and 
onset grain filling. The experiments were conducted in seven countries 
from three continents (Fig. 1).

2.2. Modeling waterlogging on wheat with DSSAT-NWheat

The DSSAT-NWheat model has been evaluated and used to simulate 
wheat growth and development across many environments around the 
world, including different seasonal ranges of air temperatures, soil 
water, nitrogen and air CO2 concentrations (Kassie et al., 2016). Here, 
we further-developed the waterlogging module of DSSAT-NWheat 
(Shelia et al., 2019), originally developed for APSIM Wheat (Asseng 
et al., 1997), by adding the soil waterlogging direct impact on carbo
hydrate accumulation and potential grain size in addition to the already 
existing impact on root growth. The latest versions of DSSAT (4.7 and 
4.8) available for free download do not include the Shelia et al. (2019)
waterlogging module, which is only available upon request from the 
authors of this paper. As a result, the default DSSAT-Nwheat available to 
users does not account for the impact of waterlogging on wheat growth. 
Here, we developed a new waterlogging module in DSSAT-Nwheat 
version 4.7, as presented here. Our new waterlogging module will be 
implemented in DSSAT-Nwheat version 4.8 and made freely available to 
the public. However, until that time, the module will also be available 
upon request from the authors.

2.2.1. Modeling soil waterlogging
The water flow module of DSSAT-NWheat was described in detail by 

Asseng et al. (1997) and was not further-modified in this study. The 
vertical water flow in a soil at or near to saturation (when soil water 
content is in between soil drained upper limit (or soil field capacity) and 
saturated soil water content) is controlled by soil saturated hydraulic 
conductivity (SSKS) and the soil saturated macro-flow water conduc
tivity, which is defined as the saturated drainage rate (SLDR) in 

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

2 



DSSAT-SBuild (Supplementary Figure S4, DSSAT-SBuild is the software 
to build soil profile in DSSAT). SSKS (in cm h− 1) is a quantitative mea
sure of a soil’s ability to transmit water vertically between layers under 
saturated conditions, following a multi-layer cascading model driven by 
a hydraulic gradient. However, only a portion of the water, defined by 
the SLDR parameter (ranging from 0 to 1), can move downward each 
day. Therefore, the SSKS value is applied only to this SLDR-defined 
fraction, representing the rate at which saturated water flows from 
one soil layer to the next. The remaining water fills the current soil layer 
up to saturation (e.g. with SLDR = 0.95, only 5 % of the water above soil 
drained upper limit in a day remains in the current soil layer). In 
DSSAT-SBuild the SLDR is defined based on soil drainage classes, 
varying from soils with very excessive drainage (soil drainage = 0.95) to 
very poor drainage (0.01). The drainage classes and their respective 
SLDR value are shown in Supplementary Table S2.

Waterlogging is indicated with an aeration deficit factor (AF) which 
is calculated daily for each rooted soil layer. The aeration deficit factor is 
derived from experimental data and incorporated into crop models, 
which simulate its effects on root development and overall crop growth 
(Asseng et al., 1997; Lizaso, 1993). The AF is calculated as follows: 

AFi =
SATi − SWCi

SATi − DULi
(1) 

Where, SATi is saturated soil water content, SWC is the soil water con
tent and DUL is the soil drained upper limit (or soil field capacity), in a 
soil rooted layer i. AFi is only calculated when soil layer is at or near to 
saturation, i.e., SWC > DUL.

2.2.2. Waterlogging impacts on roots
Waterlogging inhibits root growth and functioning (Colmer and 

Voesenek, 2009), which may lead to root death (Herzog et al., 2016). 
The module that accounts for the waterlogging impacts on roots, was 
previously implemented by Shelia et al. (2019). Our waterlogging 
modeling module was developed to consider the negative impacts of 
waterlogging on wheat growth at different soil depths (Malik et al., 

2001). When AFi in a rooted soil layer falls below the default threshold 
of 0.6 for three consecutive days, the model reduces the effective rooting 
depth to 5 cm below the deepest non-saturated layer. All roots located in 
waterlogged layers below this point are considered ‘dead’, meaning they 
no longer contribute to water or nitrogen uptake. In DSSAT-NWheat, 
only ‘active’ roots are used to simulate these processes. The purpose of 
this module is to capture the loss of root function under prolonged 
saturated conditions, regardless of the depth of the affected soil layer. 
Therefore, it is recommended to set the soil layer thickness between 
10 cm (more sensitive to waterlogging) and 15 cm (less sensitive to 
waterlogging) when creating a soil profile in DSSAT-SBuild. These depth 
thresholds are fixed in DSSAT-Nwheat and are not adjusted based on 
specific cultivars. Once waterlogging ends, dead roots remain 
non-functional and do not regain activity. However, the model allows 
for the regrowth of new active roots into soil layers that were previously 
affected by waterlogging.

2.2.3. Waterlogging impact on carbohydrate accumulation
Waterlogging causes stomata to close, reducing CO2 within the 

leaves, which restricts photosynthesis and carbohydrate accumulation 
(Else et al., 2001; Jitsuyama, 2017). The waterlogging effect on carbo
hydrate accumulation is based on a wheat root aeration index (AFroot), 
calculated as follows: 

AFroot =

⎡

⎢
⎢
⎣

∑m

i=1
(AFi )

m

⎤

⎥
⎥
⎦

PAF

(2) 

Where, PAF (Parameter of roots aeration index) is set to 3.0 (default), m 
is the deepest soil layer with roots, i is the rooted layer. The value of m is 
also dynamically affected by waterlogging impacts on roots, as 
described in Subsection 2.2.2.

Waterlogging affects daily carbohydrate accumulation in DSSAT- 
NWheat according to two new wheat cultivar parameters. First, the 

Fig. 1. Waterlogging experiments with wheat. Spatial distribution of reviewed experiments on waterlogging impact for wheat (green circles). A total of 17 
published studies were used here to summarize how waterlogging affects wheat. A list of all studies is shown in Supplementary Table S1. The black triangles represent 
experiments that were used to test the new waterlogging module for DSSAT-NWheat with a waterlogging experiment in an outdoor controlled environment from 
Marti et al. (2015), and a field experiment with drainage and undrained fields in La Jaillière, Loireauxence, France.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

3 



cultivar sensitivity to AFroot , considering the AFroot threshold value at 
which waterlogging impacts starts to affect carbohydrate accumulation 
(WLSI). Second, the maximum reduction of carbohydrate accumulation 
under waterlogging (WLMI). Both WLSI and WLMI vary from 0 to 1. A 
wheat cultivar has maximum sensitivity to waterlogging when WLSI is 1 
and WLMI is 0, and maximum tolerance when WLSI is 0 and WLMI is 1 
(Fig. 2). The default values of WLSI and WLMI in the DSSAT-NWheat 
ecotype parameters file are 0.2 and 0.1, respectively (Fig. 2), but they 
can be calibrated according to the waterlogging sensitivity of each 
cultivar.

2.2.4. Waterlogging impact on potential grain size
Waterlogging before and during anthesis affects potential grain size 

(Marti et al., 2015). In wheat, average grain size is most sensitive to 
biotic or abiotic stress between booting and the beginning of effective 
grain filling (Calderini et al., 2021a, 2001) (Supplementary Figure S5). 
DSSAT-NWheat, computes daily water deficit, nitrogen, heat and 
waterlogging stress factors, which impact potential growth, according to 
Eqs. 3 and 4. These equations are part of the original DSSAT-NWheat 
model. Our contribution consisted solely of incorporating the water
logging stress term. Therefore, the placement and relative treatment of 
each stress component, including the separate multiplicative application 
of heat stress, follow the pre-existing model logic and are not the subject 
of methodological discussion in this study.   

Daily wheat accumulation of carbohydrate

= Potential carbohydrate accumulationxplant stress (4) 

Wheat grain number per unit area is closely related to growing 
conditions before and shortly after anthesis (Fischer, 1985), when the 
number of fertile florets is determined and when fertile florets set grains 
(Slafer et al., 2015). This is also the period when potential grain size is 
set (Calderini et al., 2021a; Acreche and Slafer, 2006). DSSAT-NWheat 
estimates grain number in the first day of grain filling (phenological 

stage 5 in DSSAT-NWheat), based on the stem (above-ground biomass 
not considering leaves) growth from start-of-stem-growth (phenological 
stage 2) to anthesis (phenological stage 4) (For further details on the 
simulation of wheat grain number per unit area by DSSAT-Nwheat, 
Asseng et al. (2017)). However, since daily stem growth depends on 
carbohydrate accumulation (and our newly introduced waterlogging 
stress factor now affects this process, as shown in Eqs. 3 and 4) water
logging events occurring before or during these stages will indirectly 
influence the simulated grain number. Additionally, we introduced a 
waterlogging stress factor that directly affects potential grain size. In 
DSSAT-NWheat, the original potential average grain size was set to 
55 mg of dry mass per grain (mg grain-1, this is a crop parameter named 
MXGWT in the wheat ecotype file of DSSAT-NWheat, with a default 
value of 55). In our new approach, the potential wheat grain size is now 
based on a weighted crop stress factor between phenological stage 2–4, 
as defined in Eqs. 5 and 6, after Calderini et al. (2001): 

Potential grain sizestress = 0.25 × Plant stressstage2 + 0.35

× Plant stressstage3 + 0.40

× Plant stressstage4 (5) 

Potential grain size = 55 × Potential grain sizestress (6) 

Minimum potential grain size (MPGS) is as a default value within 
DSSAT-NWheat to 35 mg grain− 1, MXGWT and MPGS can be adjusted to 

different cultivars if supported by empirical evidence. In Eq. 5, plant 
stress corresponds to the average of the daily stress values calculated by 
DSSAT-NWheat during each phenological stage used in the equation.

2.3. Evaluation of the adapted waterlogging module

The 17 peer-reviewed articles we considered primarily served to 
demonstrate the observed impacts of waterlogging on wheat grain 
number per unit area, average grain size, and overall grain yield. Due to 
the lack of openly available data, such as detailed soil information, 
weather conditions, and planting and harvesting dates—required as 
inputs for simulation with DSSAT-Nwheat—we tested the waterlogging 
module with data from only one experiment from these articles, con
ducted in Lleida, Spain. Additionally, we performed a further validation 
using a field experiment from Loirauxence, La Jaillière, in France. The 
experiments used to test our module are described below.

2.3.1. Outdoor controlled waterlogging experiment in Lleida, Spain
To test the performance of the new waterlogging module we simu

lated the waterlogging experimental setup of Marti et al. (2015) with 
DSSAT-NWheat. This experiment was conducted with wheat plants 
grown outdoors in the campus of University of Lleida, Spain (Latitude 
41◦37’47”, Longitude 0◦35’47”) in columns (84 mm of diameter and 
1.25 m deep) filled with a loamy sand soil with 81 % sand content. The 
soil was waterlogged through the blocking of the bottom of the columns, 
to prevent water drainage. Waterlogging condition was imposed by 
saturating the soil with water 1 cm above the soil surface. The water
logging treatments were applied before anthesis, during 4–24 days 
(Supplementary Figure S1). The experiments were free from nutritional 
and water deficit stress and pests and diseases.

To simulate this experiment, we reproduced the sandy soil from 
Marti et al. (2015) with a depth of 1.25 m, divided into 11 layers with 
thicknesses ranging from 5 to 20 cm. To create the waterlogging con
dition of the experiment, we set SSKS of the bottom layer at 1 cm h− 1, 
which remains unchanged for the whole growing season. This resulted 

Fig. 2. Reduction of carbohydrate accumulation as a function of root 
aeration factor (AFroot) in DSSAT-NWheat. WLSI and WLMI are two new 
wheat cultivar parameters in DSSAT-NWheat, representing when waterlogging 
impacts starts to affect wheat carbohydrate accumulation (WLSI – Water Logging 
Impact Start), and the maximum reduction of carbohydrate accumulation under 
waterlogging (WLMI – Water Logging Maximum Impact). The values shown are 
the default in the code, but they may be parameterized according to 
wheat cultivar.

Plant stress = min(water deficitstress, nitrogenstress, ozonestress, waterloggingstress) ×heatstress (3) 

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

4 



in undrained of excess water from the bottom of the soil and caused 
excess water to slowly return to top layers of soil surface. In the simu
lation setup, the water lost by evapotranspiration was added daily 
through irrigation, and in the phenological phase when waterlogging 
was applied (Supplementary Figure S1), excess irrigation was applied to 
saturate the entire soil profile. The wheat cultivar in this experiment was 
Soisson, and the parameters were calibrated to simulate the observed 
phenology and wheat yield in the control treatment without water
logging (Table 1), using the automatic time-series calibration method 
for DSSAT wheat models (Röll et al., 2020).

2.3.2. Field experiment from Loirauxence, La Jaillière in France
ARVALIS data from field experiments (located in Loireauxence, 

research station of La Jaillière, France) with measured grain and sea
sonal accumulated water drainage and run-off (Supplementary 
Figure S2), with side-by-side drained and undrained soils from 1990 to 
2014 were used to test the performance of the new waterlogging module 
implemented in DSSAT-NWheat.

The experiments were rainfed and sown between October and 
November with sowing dates provided by Arvalis. Each year of culti
vation involved the use of a different cultivar, and the Arvalis experi
ments we simulated range from 1991 to 2014. In each year of the 
experiment, a single cultivar was sown across the entire area, encom
passing both drained and undrained soils. To minimize the need for 
calibration specific to field, year, and cultivar, we categorized the cul
tivars into three groups based on the years they were grown, which we 
refer to as ’Generic Cultivars’ We calibrated Generic Cultivar 1 (pre
dominantly derived from the wheat cultivar Soisson, representing 38 % 
of all data during this period) for the years 1991–2003. Generic Cultivar 
2 (primarily derived from the wheat cultivar Altigo, representing 32 % 
of all data during this period) was calibrated for the years 2004–2012. 
Finally, Generic Cultivar 3 (mostly derived from the wheat cultivar 

Ascott, representing 50 % of all data during this period) was calibrated 
for the years 2013–2014. It is important to note that we did not test 
Generic Cultivar 3 for the non-drained fields with occurrences of 
waterlogging due to a lack of data.

We calibrated the three generic cultivars using the same water
logging parameters, as the available data were insufficient to assess 
cultivar-specific sensitivities to waterlogging. Although Generic Cultivar 
1 is primarily represented by Soisson (32 % of the group’s data, and also 
used in the controlled experiment described in subsection 2.3.1, where it 
was calibrated with different crop parameters within the model), other 
cultivars with varying waterlogging sensitivities were also included. The 
goal of this study was to test the new waterlogging module in the DSSAT- 
NWheat model using field data from both drained and undrained soils, 
rather than to determine the waterlogging sensitivity of individual cul
tivars. Identifying cultivar-specific responses would require a larger 
dataset with multiple cultivars tested under identical conditions within 
the same year. However, since only one cultivar was tested per year in 
this experiment, such comparisons were not possible. Therefore, our 
analysis did not focus on determining waterlogging sensitivity among 
cultivars from Loirauxence and La Jaillière, France. Also for this reason, 
we used the term ’generic cultivars’.

The crop protection programs were similar to local farm practices 
and included fungicide, herbicide, and insecticide applications to pre
vent any damage to the crop. Phosphorous and potassium fertilizers 
were applied during autumn if needed to prevent late-season shortage of 
these nutrients affecting nitrogen uptake, yield and grain quality. The 
calibrated cultivar parameters used in this experiment are given in 
Table 1. For the cultivar Soisson, we adjusted the crop parameters 
MXFIL (potential kernel growth rate) and MPGS (minimum potential 
grain size) based on observed data (Fig. 4). For the other three generic 
cultivars, no observed data on grain size or grain number were available; 
thus, we used generic parameters with default values from DSSAT- 
NWheat and applied them uniformly across genotypes. We followed 
the same approach for other crop parameters. Since more detailed 
observed data were available for the Soisson cultivar, its calibration was 
more specific. For the remaining cultivars, in the absence of detailed 
information, we kept the recommended default values.

Soil attributes used in the simulations are given in Supplementary 
Table S3. Daily weather data with maximum and minimum temperature 
(◦C), rainfall (mm) and solar radiation (MJ m− 2 d− 1) were from a 
weather station located on the experimental site. Arvalis also provided 
observed input nitrogen data applied over each season for each cultivar.

2.3.3. Sensitivity analysis
To determine the potential impact of waterlogging on simulated 

wheat growth with the waterlogging module in DSSAT-NWheat, we 
conducted a sensitivity analysis. The sensitivity analysis was carried out 
by simulating waterlogging starting in different phenological stages and 
with six different waterlogging durations, from 4 to 24 days. This 
analysis was carried out under the same soil and climate and manage
ment conditions as the outdoor controlled waterlogging experiment in 
Lleida, Spain described in the subsection 2.3.1. The waterlogging con
ditions were created by setting SSKS of the bottom soil layer at 1 cm h− 1, 
which prevents water from draining from the soil and causes excess 
water to return to topsoil layers in the surface. Surface runoff may occur 
depending on soil conditions, following the default soil water balance 
structure implemented in DSSAT, which includes a runoff routine based 
on the Soil Conservation Service curve number method (Porter et al., 
1999). In the simulation setup, water lost by evapotranspiration was 
replenished daily through irrigation. During the phenological phase 
when waterlogging was applied, additional irrigation was used to 
saturate the entire soil profile, ensuring waterlogging across all soil 
layers. Simulated waterlogging was applied during the following 
phenological stages: seedling emergence, 2-leaves, 3-leaves, onset of 
tillering, onset of stem elongation, anthesis, and onset of grain filling.

Table 1 
NWheat genetic coefficients for the winter wheat cultivars Soisson (Spain 
experiment) and Generic cultivar 1, cultivar 2 and cultivar 3 (France 
experiment).

Genetic 
Coef.

Definition Soisson 
cv.

Generic 
cultivar 
1

Generic 
cultivar 
2

Generic 
cultivar 
3

VSEN Sensitivity to 
vernalization scale 
1–4

4.43 4.43 3.70 3.70

PPSEN Sensitivity to 
photoperiod

3.36 3.36 4.00 4.00

P1 Thermal time from 
seedling emergence 
to the end of the 
juvenile phase [◦C d]

446 400 400 400

P5 Thermal time (base 
0◦C) from beginning 
of grain filling to 
maturity [◦C d]

754 500 580 600

PHINT Phyllochron interval 100 100 110 110
GRNO Coefficient of kernel 

number per stem 
weight at the 
beginning of grain 
filling [kernels g− 1 

stem− 1]

32 29 22 28

MXFIL Potential kernel 
growth rate [mg 
kernel− 1 d− 1]

3.20 1.60 1.60 1.60

WLSI WLSI – Water 
Logging Impact Start

0.2 0.2 0.2 0.2

WLMI WLMI – Water 
Logging Maximum 
Impact

0.1 0.3 0.3 0.3

MPGS Minimum potential 
grain size

20 35 35 35

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

5 



2.4. Data analysis

Data and statistical analyses were conducted using the statistical 
software program R (R Core Team, 2017). To evaluate the predictive 
performance of the waterlogging module in DSSAT-NWheat we 
computed the relative root mean squared error (rRMSE), the coefficient 
of determination (r2) and the Willmott agreement index (d) (Willmott 
et al., 1985), based on the estimated wheat yield at the tested location 
together with the corresponding observed yield. The equations for the 
indices used to evaluate the model performance are presented in 
Table 2.

3. Results

3.1. Review of waterlogging impacts on wheat yield components

Waterlogging from the onset of tillering to anthesis results in a 25 
± 10 % decrease in grain number per unit area in the 17 reviewed ar
ticles reporting wheat waterlogging experiments (Fig. 3). The highest 
variation in waterlogging impacts on wheat grain number per unit area 
(i.e., highest s.d. shown in Fig. 3a) are shown when waterlogging 
occurred at the onset of tillering and heading (Fig. 3a). Grain number 
was less affected when waterlogging occurred at the onset of grain 
filling, with a 5 ± 5 % decrease.

Episodes of waterlogging resulted in lower average grain size 
compared with the control treatments with no waterlogging (Fig. 3b). 
Differently from the wheat grain number, the average grain size was 
more sensitive to waterlogging when it occurred during the onset of 
grain filling, with a 30 ± 5 % decrease. Waterlogging during seedling 
emergence had almost no impact on average grain size. With the com
bined effect of waterlogging on average grain number and grain size per 
unit area, grain yield was more sensitive to waterlogging when it 
occurred during the onset of tillering, anthesis and the onset of grain 
filling, with on average 37 ± 10 %. Waterlogging impacts on grain yield 
were lower during seedling emergence.

3.2. DSSAT-NWheat performance to simulate the impacts of pre-anthesis 
waterlogging on wheat yield components

We tested the improved DSSAT-NWheat to simulate the impacts of 
waterlogging on yield and yield components, with an outdoor controlled 
waterlogging experiment in Lleida, Spain (described in subsection 
2.3.1). First, the simulations were done with the default DSSAT-NWheat 
without any improvements to simulate waterlogging impacts. As part of 
this, the simulated wheat yield was 9.3 t ha− 1, regardless of the duration 
of pre-anthesis waterlogging (Fig. 4). Second, the simulations were 
performed with DSSAT-NWheat with added modules for simulating the 
impacts of waterlogging on wheat roots (R) and carbohydrate accumu
lation (CA). In this case, the simulated wheat yield shows a decline due 
to pre-anthesis waterlogging, of almost 3 t ha− 1 with 24 days long 
waterlogging period (Fig. 4a). This yield decline is mainly due to the 
decreased simulated grain number (Fig. 4b), with increased grain size 

compensating for it (Fig. 4c). The simulated grain size increased up to 
57 % compared to the simulations with DSSAT-Wheat without any im
provements to simulate waterlogging impacts. To better simulate the 
impacts of waterlogging on yield components, we added an impact of 
waterlogging on potential grain size (GS). Potential grain size responses 
to stress are known (Calderini et al., 2021b, Reynolds et al., 2022) but 
are often not represented in crop models, which typically simulate a 
compensatory effect—larger grain size when grain number is low, and 
vice versa (Fig. 4c). However, this compensation may occur for potential 
grain yield but is not always observed, especially when both grain 
number and potential grain size are simultaneously affected 
(Supplementary Material Figure S7). Simulated potential grain size in 
DSSAT-Nwheat is therefore now affected by stress around anthesis.

Considering decreased rooting depth, carbohydrate accumulation 
and potential grain size under waterlogging conditions, the new 
improved DSSAT-NWheat satisfactorily simulated wheat yield loss due 
to waterlogging, with lower grain number and grain size (Fig. 4).

With the new improved DSSAT-NWheat the rRMSE between simu
lated and observed waterlogged grain yield, grain number per unit area 
and grain size was 18 % (Table 3). For grain yield and grain number r2 

was 0.68, and for average grain size r2 was 0.30. The accuracy of the 
model, represented by the index d, was 0.86 for grain yield, 0.79 for 
grain number per unit area and 0.38 for average grain size. Even with 
these improvements, the DSSAT-NWheat tends to underestimate the 
average grain number and to overestimate the average grain size, 
particularly when waterlogging occurred from 4 to 12 days (Fig. 4b and 
c).

The improved DSSAT-NWheat model have demonstrated increased 
sensitivity in simulating the impacts of waterlogging on wheat yield, 
particularly during the anthesis and grain filling stages (Fig. 5b and c). 
Notably, there has been an escalation in loss estimates when the dura
tion of waterlogging exceeded 4 days, compared to the previous version 
of DSSAT-NWheat, which included the waterlogging module developed 
by Shelia et al. (2019) (Fig. 5).

We further evaluated the new improved DSSAT-NWheat water
logging model for the field experiment with side-by-side drained and 
undrained soils conducted from 1990 to 2014 in Loireauxence, La Jail
lière (44), France (Fig. 6). The DSSAT-NWheat simulated drained water 
with r2 of 0.72, rRMSE of 44 % and d of 0.64. Runoff water was simu
lated with r2 of 0.9, rRMSE of 107 % and d of 0.78. The simulations of 
grain yield showed a similar performance with r2 slightly increasing 
from 0.05 to 0. 11 when comparing grain yield simulated with DSSAT- 
NWheat with no waterlogging module (Fig. 6c) to with DSSAT- 
NWheat with waterlogging module (Sim. NWheat improved R + CA +
GS) (Fig. 6d), respectively. In addition, the rRMSE slightly increased 
from 19 % to 20 %, and the d-index from 0.41 to 0.43 when comparing 
grain yield simulated with DSSAT-NWheat with no waterlogging mod
ule (without simulations of the impact of waterlogging on wheat 
growth) (Fig. 6c) to with DSSAT-NWheat with waterlogging module 
(with simulations of the impact of waterlogging on wheat growth) 
(Fig. 6d).

3.3. Sensitivity analysis

The capacity of the model to simulate the impacts of waterlogging on 
wheat growth variables and yield components is shown through the 
sensitivity analysis, with waterlogging occurring at different phenolog
ical stages and with different durations (Fig. 7). For winter wheat 
(cultivar Soisson), simulated waterlogging starts to impact yield com
ponents during the onset of tillering, except for root biomass. Simulated 
root biomass is less impacted by waterlogging when it occurred in the 
early phenological stages (Fig. 7b). The highest impacts on root biomass 
occur during the onset of grain filling, with a reduction of up to 60 % 
after 24 days of waterlogging compared to control treatment with no 
waterlogging. The biomass at maturity and the grain number per unit 
area are affected from the onset of tillering to the onset of grain filling, 

Table 2 
Statistical indexes and errors used for evaluating the DSSAT-NWheat 
performance.

Statistical indexes and errors Formula*

Willmott agreement index (d)
d = 1 −

∑n
i=1(Simi − Obsi)

∑n
i=1 (|Simi − Obs| + |Simi − Obs|)2

Relative root mean squared error 
(rRMSE)

rRMSE =
̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
1
n
×
∑n

i=1
(Simi − Obsi)2/Obs × 100

√

* Sim i and Obsi are the simulated and observed wheat yield; n is the number 
of observations; and Obs and Est are the average of Obsi and Sim i, respectively.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

6 



with the highest impact occurring at the onset of stem elongation 
(Fig. 7a and d). At this stage, both biomass at maturity and grain number 
per unit area decrease by 50 % with 24 days of waterlogging.

Leaf area index is impacted by waterlogging from 3-leave-stage to 
the onset of stem elongation, with the highest sensitivity at the onset of 
tillering. The average grain size decreased with waterlogging from the 
onset of stem elongation to the onset of grain filling, and it declines more 
than 50 % with 24 days of waterlogging in the onset of grain filling. As a 
result of the decline of all wheat yield components under waterlogging 
conditions, grain yield can decrease by up to 65 % compared to the 
control (with no waterlogging), with waterlogging occurring from the 
onset of stem elongation to the onset of grain filling.

4. Discussion

4.1. Modeling performance of the new waterlogging module in DSSAT- 
NWheat

Oxygen availability is restricted in waterlogged soils, suppressing 
root respiration and causing decreased root activity (Van Veen et al., 
2014). Plants can temporarily maintain energy production during 
waterlogging via glycolysis and ethanol fermentation, but prolonged 
waterlogging leads to the accumulation of toxic metabolites and 
increased reactive oxygen species, resulting in root cell death (Pan et al., 
2021). This inhibits root growth and nutrient transport to shoots, and 
can lead to nutrient deficit stress (Herzog et al., 2016). Reduced root 
function and increased nutrient leaching from excessive soil water due 
to waterlogging may lead to nutrient deficit stress (Salazar et al., 2014; 
Kaur et al., 2020). Even in the absence of nutrient deficiencies, water
logging limits root water conductivity, causing stomatal closure and 
reduced crop growth (Else et al., 2001; Jitsuyama, 2017).

Crop models consolidate scientific knowledge of crop growth pro
cesses and enhance understanding of plant responses to abiotic stresses, 
such as waterlogging. However, they often simplify these processes and 
omit detailed physiological and biochemical responses (boote et al., 
2013; Nóia-Júnior et al., 2025a). For example, the current models, 
including the waterlogging module introduced in this study, do not ac
count for the plant’s ability to cope with toxins (both elemental, such as 
excessive Fe or Mn, and organic compounds, such as phenolics and 
carboxylic acids) produced during hypoxic conditions (Yemelyanov 
et al., 2023). Additionally, reduced nutrient availability due to limited 
mitochondrial energy production for active nutrient uptake, and 
oxidative stress induced by waterlogging (León et al., 2021), are also 
factors that have not been integrated into the model. This level of 
complexity is often omitted because the inclusion of more physiological 
processes increases the need for additional crop parameters and inputs, 
which require new field experiments for calibration. Beyond a certain 
point, adding such complexity can increase model prediction error 
(Reynolds and Acock, 1985). In this study, we implemented a new 
waterlogging module into the wheat crop simulation model 
DSSAT-NWheat, accounting primarily for the effects of waterlogging on 
wheat root growth, carbohydrate accumulation, and potential average 
grain size. Many of the physiological and biochemical factors mentioned 
above were not considered in the module, in line with common sim
plifications in crop models.

We evaluated the new waterlogging module in DSSAT-NWheat based 

(caption on next column)

Fig. 3. Observed waterlogging impact on wheat grain number, average 
grain size, and grain yield. The observed impacts of waterlogging applied in 
different phenological stages on (a) grain number per unit area, (b) average 
grain size, and (c) grain yield. (d) Observed waterlogging average duration in 
each phenological phase. The 17 peer-reviewed articles are shown in Supple
mentary Table 1. The number of experiments (n) per for each phenological 
stage is shown in (a). Points are means and the vertical bars represent ± 1 s.d. 
the black dashed lines represent the control treatments in (a-c), and the results 
are presented as a percentage difference from the control.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

7 



on its ability to simulate yield under waterlogged conditions, using data 
from an outdoor controlled experiment with measured pre-anthesis 
waterlogging impacts and from an undrained wheat field.

In the first test set, conducted in an outdoor controlled experiment in 
Lleida, Spain, the improved DSSAT-NWheat model demonstrated good 
performance in simulating the impacts of pre-anthesis waterlogging on 
wheat yield and yield components. The model was tested under pre- 
anthesis waterlogging conditions, and the simulations showed a 
decline in yield due to longer waterlogging periods. For example, with a 
24-day pre-anthesis waterlogging period, yield declined by about 3 t ha- 
1, primarily reduced growth resulting in a reduction in grain number. 
The rRMSE for grain yield was 18 %, with an r² of 0.69. In comparison, 
an experiment using APSIM achieved enhanced simulation accuracy for 
waterlogging effects on wheat carbohydrate accumulation, with an r² of 
0.70 and an rRMSE of 11 % (Liu et al., 2023). While the model per
formed reasonably well, it tended to underestimate grain number and 
overestimate grain size, particularly when waterlogging occurred for 
4–12 days. Despite these issues, the improved DSSAT-NWheat model 
showed an increased sensitivity in simulating the impacts of pre-anthesis 
waterlogging on yield.

In contrast, the second test set, conducted in Loireauxence, France, 
presented additional challenges. The lack of data on key variables, such 
as soil moisture and water table depth limited the crop model’s ability in 
simulating these conditions. For this field, we only had information 

indicating that one plot was drained while the other was not, restricting 
the setting of parameters for this location. Additionally, expected field- 
scale factors, such as plant diseases, soil variability, and microclimatic 
variations within the field—which could further impact observed and 
simulated responses—were not reported. We also used generic cultivar 
parameters when simulating this experiment, which ignores potential 
genotypic differences in adaptive traits for waterlogging tolerance, such 
as root porosity, radial oxygen loss, and root architecture (Pedersen 
et al., 2021). These factors led to limited model performance in this 
second test, although the new waterlogging module demonstrated its 
capability to improve grain yield simulations for this field. This is 
particularly relevant for regions such as northern France—where the test 
site is located—where excess water is often the main constraint to wheat 
productivity (Nóia-Júnior et al., 2025b). In such environments, yield 
losses caused by heavy rainfall are common but are typically not 
captured by crop models (Nóia-Júnior et al., 2025a), making simulations 
more challenging. Despite these limitations, our new routine still led to a 
slight improvement in simulations, indicating that yield was penalized 
in undrained fields due to waterlogging.

While the waterlogging module provides useful insights, several 
model deficiencies became apparent, particularly the need for more 
detailed calibration data and an understanding of key environmental 
and genetic factors that influence plant responses to waterlogging. A full 
exploration of these drivers of model error is needed. Future studies 
should focus on improving parameterization for different soil types, 
incorporating genotypic variation, and including additional physiolog
ical processes, such as nutrient uptake limitations under anaerobic 
conditions. The model’s simplification of waterlogging effects may also 
overlook important physiological responses, such as the production of 
toxic metabolites in hypoxic environments and oxidative stress. In our 
new module, we considered only the direct impacts of waterlogging on 
selected aspects of wheat growth, including root function, potential 
grain size, and carbohydrate accumulation. Other processes possibly to 
be affected by waterlogging, such as biomass partitioning and 
phenology (Nóia Júnior et al., 2023a; Garcia-Vila et al., 2025), were not 
included. Likewise, indirect effects such as reduced nitrogen availability 
due to increased leaching or limited However, indirect effects such as 
reduced nitrogen availability due to increased leaching or limited root 

Fig. 4. Performance of the improved DSSAT-NWheat in simulating pre-anthesis waterlogging impacts on wheat. Performance of DSSAT-NWheat in simu
lating pre-anthesis waterlogging impacts on wheat (a) grain yield (b) grain number and (c) grain size. Data (points) are from an outdoor controlled experiment where 
wheat plants where waterlogged for 0–24 days before anthesis (Marti et al., 2015). Simulations (lines) were done with DSSAT-NWheat with no waterlogging module 
(Sim. NWheat), with the waterlogging module accounting only for the impacts of waterlogging on wheat roots and carbohydrate accumulation (Sim. NWheat 
improved R + CA) and with the waterlogging module accounting for the impacts of waterlogging on wheat roots, carbohydrate accumulation and potential grain size 
(Sim. NWheat improved R + CA + GS). The acronyms R represents the waterlogging module with impacts on wheat roots, CA in carbohydrate accumulation and GS 
in potential grain size.

Table 3 
Performance of the improved DSSAT-NWheat in simulating pre-anthesis 
waterlogging impacts on wheat. Data are from an outdoor controlled experi
ment where wheat plants where waterlogged for 0–24 days before anthesis 
(Marti et al., 2015) with the DSSAT model with a waterlogging module affecting 
wheat roots, carbohydrate accumulation potential grain size. Statistical indices 
are the relative root mean square error (rRMSE), the coefficient of determination 
(r2) and the Willmott agreement index (d) (Willmott et al. 1985).

Wheat yield components rRMSE (%) r2 d

Grain yield 18 0.69 0.86
Grain number 18 0.68 0.79
Grain size 18 0.30 0.38

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

8 



uptake—caused by shallower rooting under waterlogged conditions are 
included, but were not evaluated due to lack of data. These limitations 
should be acknowledged, and further data from field experiments are 
essential to reduce model uncertainty and improve performance in 
diverse conditions. In addition, our new waterlogging module would 

benefit from further validation using controlled experiments in which 
waterlogging is the only isolated factor affecting yield at phenological 
stages beyond pre-anthesis, as was the case in the present study. Such 
experiments are rare, and when they do exist, they are often not avail
able in open data repositories, limiting their use for modeling (Nóia 
Júnior et al., 2023a). Although the module is designed to affect crop 
growth at any stage, it was validated using controlled experiments only 
for pre-anthesis waterlogging impacts. Nevertheless, the sensitivity 
analysis results shown in Fig. 7 are consistent with the findings from a 
literature review of multiple studies, as summarized in Fig. 3.

Despite the challenges and limited data for traditional model eval
uation, the model utility remains significant for simulating waterlogging 
impacts on wheat growth. However, its current limitations restrict 
prediction accuracy across different contexts, as is typical with any crop 
simulation model, with or without a waterlogging module implemented 
(Rötter et al., 2018, 2011).

4.2. Simulated impacts of waterlogging on wheat yield components

The DSSAT-NWheat model uses the stem weight at the beginning of 
grain filling to simulate grain number per unit area. Once grain number 
per unit area is determined, DSSAT-NWheat simulates carbohydrate 
accumulation in grains, with the maximum grain size limited by the 
MXGWT parameter. With few grains, all the simulated carbohydrate 
assimilated during grain filling plus the carbohydrate remobilized from 
the shoot is distributed to the few grains, causing the simulation of 
heavy grains (i.e. high average grain size). Although this is the physio
logical concept within the DSSAT-NWheat, high average grain size may 
also be caused due to the non-development and growth of potential 
grains close to the labile florets, positioned more distally and with 
constitutively low grain size, which usually occur during limited grain 
set situations. In this case, these potential grains with low grain size 
would not succeed (would not grow and become grains), resulting in an 
increase in the average grain size of harvested grains (because only the 
bigger grains would be harvested) (Beral et al., 2022). Waterlogging 
accelerates flag leaf senescence (Li et al., 2012), affecting post-anthesis 
biomass production and translocation to grains, further impairing the 
filling of small grains. However, the current waterlogging module in 
DSSAT-NWheat does not directly simulate these effects, such as the 
non-development of small grains and accelerated flag leaf senescence. 
This relation between grain number per unit area and grain size was here 
demonstrated for waterlogging simulations before wheat anthesis, 
which resulted in few but heavier grains on average. To minimize the 
overestimation of grain size, we implemented an equation to limit the 
potential grain (Eqs. 5 and 6) size when abiotic stress occurs before 
anthesis similar to observations made experimentally (Liu et al., 2020a). 
This also improved the simulation of waterlogging impact on grain size.

Grain volume enlargement involves the coordinated expansion of the 
pericarp of maternal origin and the endosperm of the seed, and is almost 
complete when grain filling begins (Calderini et al., 2021a; Reynolds 
et al., 2022). The expansion of these tissues determines the potential 
grain size and proteins storage capacity (Herrera and Calderini, 2020). 
The growing conditions around anthesis are therefore critical for po
tential grain size determination, a time when simultaneously also grain 
number per unit area is set (Acreche and Slafer, 2006; Calderini et al., 
2021a; Slafer et al., 2023). In addition to the impacts of waterlogging on 
the potential grain size, grain size may also be affected by any other 
abiotic constraints at this developmental stage, like heat, drought, ozone 
and nitrogen deficit, all factors considered in DSSAT-NWheat. The 
combined impact of these factors on grain size determination still needs 
to be tested with field data.

4.3. Factors that affect the intensity of waterlogging impacts on wheat 
growth

The impacts of waterlogging on wheat growth varies particularly 

Fig. 5. Simulated waterlogging impacts on wheat with different water
logging modules approaches. Simulated impacts of waterlogging applied 
during (a) booting (b) anthesis and (c) grain filling on grain yield simulated 
using a waterlogging module implemented by Shelia et al. (2019) in 
DSSAT-Nwheat and the improved waterlogging module for DSSAT-Nwheat 
proposed in this study.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

9 



according to its duration and crop phenological stage. We showed that 
wheat may be less affected by waterlogging during the seedling emer
gence stage, and can drop drastically during anthesis and grain filling, 
either due to lower grain numbers or grain size (Fig. 3), which was also 
simulated with the DSSAT-NWheat (Fig. 7). In addition, waterlogging 
impacts on wheat and on any crop also vary according to the soil 
waterlogged depth, air temperature, soil type, mineral nutrition man
agement and genotype, which includes adaptive traits like root porosity, 
radial oxygen loss, and root architecture that allow some genotypes to 
better tolerate waterlogging (Herzog et al., 2016). Some of these factors 
can be indirectly considered via parameters, like a parameter on root 
survival length under waterlogging, if genetic variation is available for 
this. Other possible interactions are not included, like a low air tem
perature leading to slower soil oxygen depletion which could make the 
impacts of waterlogging on crop growth less severe (Trought and Drew, 
1982). However, more field experimental data are need to evaluate how 
critical these factors are under field conditions to be eventually included 
into a crop model.

The DSSAT-NWheat model, enhanced with a waterlogging module, 
represents a substantial advancement in simulating high-intensity 
rainfall impact on wheat growth and yields. Model simulated outputs 
correspond reasonably with established crop response curves, depicting 
reductions in grain number per unit area and grain size under water
logged conditions. However, the limited availability of comprehensive 
datasets for validating such model routines underscores a critical need 
for further detailed research on waterlogging impacts on wheat under 
field conditions, as previously noted (Rötter et al., 2011; Boote et al., 
2013).

Author contributions

All co-authors conceptualized the study. JCD, JPC, PM and SA su
pervised the study. VSt and VSh implemented the waterlogging module 
in DSSAT-NWheat. GS made available the data from the outdoor 
controlled waterlogging experiment in Lleida in Spain, and BP and AD 
the wheat data from field experiments with side-by-side drained and 
undrained soils in France. The grain number per unit area and average 
grain size data in France, was made available by JCD and JPC. RSNJ and 
Vst performed the formal analysis. RSNJ wrote initial draft, all co- 
authors assisted with writing and reviewed the manuscript.

CRediT authorship contribution statement

Zhentao Zhang: Writing – review & editing, Conceptualization, 
Methodology. Gustavo A. Slafer: Resources, Conceptualization, 
Writing – review & editing, Methodology. Senthold Asseng: Supervi
sion, Conceptualization, Writing – review & editing, Project adminis
tration. Alain Dutertre: Resources, Writing – review & editing. Gerrit 
Hoogenboom: Methodology, Writing – review & editing, Conceptuali
zation. Benoît Piquemal: Resources, Writing – review & editing. 
Matthew Tom Harrison: Writing – review & editing, Conceptualiza
tion, Methodology. Jean-Pierre Cohan: Methodology, Writing – review 
& editing, Conceptualization. Ke Liu: Conceptualization, Writing – re
view & editing. Jean-Charles Deswarte: Writing – review & editing, 
Conceptualization, Methodology. Taru Palosuo: Methodology, Writing 
– review & editing, Conceptualization. Vakhtang Shelia: Software, 
Conceptualization, Writing – review & editing, Methodology. Frank 

Fig. 6. Simulated versus observed drained water, runoff, and grain yield. Performance of the improved DSSAT-NWheat to simulate seasonal accumulated (a) 
drained and (b) runoff water and grain yield (c) before and (d) after the implementation of the waterlogging module (impact of waterlogging on R+CA+GS) in 
DSSAT-Nwheat. Data are from 1990 to 2012 in drained (open black symbols) and undrained field (closed red symbols) in Loireauxence, La Jaillière, France. Drained 
water in undrained fields were not measured. The acronym WL refers to waterlogging in (c) and (d). Different colors denote drained (open black symbols) and 
undrained (closed red symbols) fields in all plots, and different shapes denote cultivars in (c) and (d).

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

10 



Ewert: Writing – review & editing, Conceptualization, Methodology. 
Pierre Martre: Supervision, Writing – review & editing, Conceptuali
zation. Heidi Webber: Methodology, Writing – review & editing, 
Conceptualization. Valentina Stocca: Writing – review & editing, 
Methodology, Conceptualization, Software, Formal analysis. Yean-Uk 
Kim: Writing – review & editing, Conceptualization, Methodology. 
Rogério de S. Nóia Júnior: Writing – original draft, Methodology, Data 
curation, Writing – review & editing, Visualization, Formal analysis, 
Conceptualization. Van der Velde Marijn: Methodology, Writing – re
view & editing, Conceptualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial 
interests or personal relationships that could have appeared to influence 
the work reported in this paper.

Acknowledgements

The authors thank ARVALIS for the financial support of this study. R. 
S.N.J. acknowledges support from the Prince of Albert II of Monaco 
foundation through the IPCC Scholarship Program. The contents of this 
manuscript are solely the liability of R.S.N.J. and under no circum
stances may be considered as a reflection of the position of the Prince 
Albert II of Monaco Foundation and/or the IPCC. P.M. acknowledges 

support from the Agriculture and Forestry in the Face of Climate Change: 
Adaptation and Mitigation (CLIMAE) Meta-program of the French Na
tional Research Institute for Agriculture, Food and Environment 
(INRAE). F.E. acknowledges support from the Deutsche For
schungsgemeinschaft (DFG, German Research Foundation) under Ger
many’s Excellence Strategy – EXC 2070 – 390732324 (PhenoRob).

Appendix A. Supporting information

Supplementary data associated with this article can be found in the 
online version at doi:10.1016/j.fcr.2025.110090.

Data availability

Data will be made available on request.

References

Acreche, M.M., Slafer, G.A., 2006. Grain weight response to increases in number of 
grains in wheat in a Mediterranean area. Field Crops Res. 98, 52–59.

Asseng, S., Ewert, F., Martre, P., Rötter, R.P., Lobell, D.B., Cammarano, D., Kimball, B.A., 
Ottman, M.J., Wall, G.W., White, J.W., Reynolds, M.P., Alderman, P.D., Prasad, P.V. 
V., Aggarwal, P.K., Anothai, J., Basso, B., Biernath, C., Challinor, A.J., De Sanctis, G., 
Doltra, J., Fereres, E., Garcia-Vila, M., Gayler, S., Hoogenboom, G., Hunt, L.A., 
Izaurralde, R.C., Jabloun, M., Jones, C.D., Kersebaum, K.C., Koehler, A.-K., 
Müller, C., Naresh Kumar, S., Nendel, C., O’leary, G., Olesen, J.E., Palosuo, T., 
Priesack, E., Eyshi Rezaei, E., Ruane, A.C., Semenov, M.A., Shcherbak, I., Stöckle, C., 

Fig. 7. Sensitivity analysis of the impacts of waterlogging on winter wheat growth simulated by DSSAT-NWheat. Impacts of simulated waterlogging on wheat 
(a) biomass at maturity, (b) root biomass, (c) leaf area index, (d) grain number per unit area, (e) average grain size and (f) grain yield. The sensitivity analysis was 
carried out by simulating waterlogging starting in different phenological stages, from wheat seedling emergence to the onset of grain filling, with six different 
durations varying from 4 to 24 days as schematically shown in Supplementary Fig. S1. Sensitivity analysis of the impacts of waterlogging on spring wheat are shown 
in Supplementary Fig. S6.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

11 

https://doi.org/10.1016/j.fcr.2025.110090
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref1
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref1
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2


Stratonovitch, P., Streck, T., Supit, I., Tao, F., Thorburn, P.J., Waha, K., Wang, E., 
Wallach, D., Wolf, J., Zhao, Z., Zhu, Y., 2015. Rising temperatures reduce global 
wheat production. Nat. Clim. Change 5, 143–147.

Asseng, S., Ewert, F., Rosenzweig, C., Jones, J.W., Hatfield, J.L., Ruane, A.C., Boote, K.J., 
Thorburn, P.J., Rötter, R.P., Cammarano, D., Brisson, N., Basso, B., Martre, P., 
Aggarwal, P.K., Angulo, C., Bertuzzi, P., Biernath, C., Challinor, A.J., Doltra, J., 
Gayler, S., Goldberg, R., Grant, R., Heng, L., Hooker, J., Hunt, L.A., Ingwersen, J., 
Izaurralde, R.C., Kersebaum, K.C., Müller, C., Naresh Kumar, S., Nendel, C., 
O’leary, G., Olesen, J.E., Osborne, T.M., Palosuo, T., Priesack, E., Ripoche, D., 
Semenov, M.A., Shcherbak, I., Steduto, P., Stöckle, C., Stratonovitch, P., Streck, T., 
Supit, I., Tao, F., Travasso, M., Waha, K., Wallach, D., White, J.W., Williams, J.R., 
Wolf, J., 2013. Uncertainty in simulating wheat yields under climate change. Nat. 
Clim. Change 3, 827–832.

Asseng, S., Kassie, B.T., Labra, M.H., Amador, C., Calderini, D.F., 2017. Simulating the 
impact of source-sink manipulations in wheat. Field Crops Res. 202, 47–56. 〈htt 
ps://www.sciencedirect.com/science/article/pii/S0378429016301277〉 (Online). 

Asseng S., Keating B.A., Huth N.I. and Easthan J. 1997 Simulation of Perched 
Watertables in a Duplex Soil International Congress on Modelling and Simulation 
Proceedings (Hobert, Tasmania: The Modelling and Simulation Society of Australia) 
pp 538–43.

Asseng, S., Martre, P., Maiorano, A., Rötter, R.P., O’leary, G.J., Fitzgerald, G.J., 
Girousse, C., Motzo, R., Giunta, F., Babar, M.A., Reynolds, M.P., Kheir, A.M.S., 
Thorburn, P.J., Waha, K., Ruane, A.C., Aggarwal, P.K., Ahmed, M., Balkovič, J., 
Basso, B., Biernath, C., Bindi, M., Cammarano, D., Challinor, A.J., De Sanctis, G., 
Dumont, B., Eyshi Rezaei, E., Fereres, E., Ferrise, R., Garcia-Vila, M., Gayler, S., 
Gao, Y., Horan, H., Hoogenboom, G., Izaurralde, R.C., Jabloun, M., Jones, C.D., 
Kassie, B.T., Kersebaum, K.-C., Klein, C., Koehler, A.-K., Liu, B., Minoli, S., Montesino 
San Martin, M., Müller, C., Naresh Kumar, S., Nendel, C., Olesen, J.E., Palosuo, T., 
Porter, J.R., Priesack, E., Ripoche, D., Semenov, M.A., Stöckle, C., Stratonovitch, P., 
Streck, T., Supit, I., Tao, F., Van Der Velde, M., Wallach, D., Wang, E., Webber, H., 
Wolf, J., Xiao, L., Zhang, Z., Zhao, Z., Zhu, Y., Ewert, F., 2019. Climate change 
impact and adaptation for wheat protein. Glob. Chang Biol. 25, 155–173.

Beral, A., Girousse, C., Le Gouis, J., Allard, V., Slafer, G.A., 2022. Physiological bases of 
cultivar differences in average grain weight in wheat: scaling down from plot to 
individual grain in elite material. Field Crops Res 289, 108713.

Boote, K.J., Jones, J.W., White, J.W., Asseng, S., Lizaso, J., 2013. Putting mechanisms 
into crop production models (Online). Plant Cell Environ. 36, 1658–1672. https:// 
doi.org/10.1111/pce.12119.

Calderini, D.F., Castillo, F.M., Arenas-M, A., Molero, G., Reynolds, M.P., Craze, M., 
Bowden, S., Milner, M.J., Wallington, E.J., Dowle, A., Gomez, L.D., Mcqueen- 
Mason, S.J., 2021a. Overcoming the trade-off between grain weight and number in 
wheat by the ectopic expression of expansin in developing seeds leads to increased 
yield potential. N. Phytol. 230, 629–640.

Calderini, D.F., Castillo, F.M., Arenas-M, A., Molero, G., Reynolds, M.P., Craze, M., 
Bowden, S., Milner, M.J., Wallington, E.J., Dowle, A., Gomez, L.D., Mcqueen- 
Mason, S.J., 2021b. Overcoming the trade-off between grain weight and number in 
wheat by the ectopic expression of expansin in developing seeds leads to increased 
yield potential (Online). N. Phytol. 230, 629–640. https://doi.org/10.1111/ 
nph.17048.

Calderini, D.F., Savin, R., Abeledo, L.G., Reynolds, M.P., Slafer, G.A., 2001. The 
importance of the period immediately preceding anthesis for grain weight 
determination in wheat. Euphytica 119, 199–204.

Ceglar, A., Zampieri, M., Gonzalez-Reviriego, N., Ciais, P., Schauberger, B., Van Der 
Velde, M., 2020. Time-varying impact of climate on maize and wheat yields in 
France since 1900. Environ. Res. Lett. 15, 094039.

Colmer, T.D., Voesenek, L.A.C.J., 2009. Flooding tolerance: suites of plant traits in 
variable environments. Funct. Plant Biol. 36, 665–681.

De San Celedonio, R.P., Abeledo, L.G., Miralles, D.J., 2014. Identifying the critical period 
for waterlogging on yield and its components in wheat and barley. Plant Soil 378, 
265–277.

Else, M.A., Coupland, D., Dutton, L., Jackson, M.B., 2001. Decreased root hydraulic 
conductivity reduces leaf water potential, initiates stomatal closure and slows leaf 
expansion in flooded plants of castor oil (Ricinus communis) despite diminished 
delivery of ABA from the roots to shoots in xylem sap. Physiol. Plant 111, 46–54.

Fischer, R.A., 1985. Number of kernels in wheat crops and the influence of solar 
radiation and temperature. J. Agric. Sci. 105, 447–461.

Garcia-Vila, M., Dos Santos Vianna, M., Tom Harrison, M., Liu, K., De Nóia-Júnior, R.S., 
D Weber, T.K., Zhao, J., Acutis, M., Archontoulis, S., Asseng, S., Aubry, P., 
Balkovic, J., Basso, B., Chen, X., Chen, Y., De Jong Van Lier, Q., Delandmeter, M., De 
Wit, A., Dumont, B., Ferrise, R., Folberth, C., Gabbrielli, M., Gaiser, T., Gorooei, A., 
Hoogenboom, G., Christian Kersebaum, K., Kim, Y.-U., Kraus, D., Liu, B., Martin, L., 
Metselaar, K., Nendel, C., Padovan, G., Perego, A., Maria Seserman, D., Scheer, C., 
Shelia, V., Stocca, V., Tao, F., Wang, E., Webber, H., Zhao, Z., Zhu, Y., Palosuo, T., 
2025. Gaps and strategies for accurate simulation of waterlogging impacts on crop 
productivity. Nat. Food Online. https://doi.org/10.1038/s43016-025-01179-y.

Githui, F., Beverly, C., Aiad, M., Mccaskill, M., Liu, K., Harrison, M.T., 2022. Modelling 
waterlogging impacts on crop growth: a review of aeration stress definition in crop 
models and sensitivity analysis of APSIM. Int. J. Plant Biol. 13, 180–200.

Herrera, J., Calderini, D.F., 2020. Pericarp growth dynamics associate with final grain 
weight in wheat under contrasting plant densities and increased night temperature. 
Ann. Bot. 126, 1063–1076.

Herzog, M., Striker, G.G., Colmer, T.D., Pedersen, O., 2016. Mechanisms of waterlogging 
tolerance in wheat – a review of root and shoot physiology. Plant Cell Environ. 39, 
1068–1086.

Jitsuyama, Y., 2017. Hypoxia-responsive root hydraulic conductivity influences soybean 
cultivar-specific waterlogging tolerance. Am. J. Plant Sci. 08, 770–790.

Kassie, B.T., Asseng, S., Porter, C.H., Royce, F.S., 2016. Performance of DSSAT-Nwheat 
across a wide range of current and future growing conditions. Eur. J. Agron. 81, 
27–36.

Kaur, G., Singh, G., Motavalli, P.P., Nelson, K.A., Orlowski, J.M., Golden, B.R., 2020. 
Impacts and management strategies for crop production in waterlogged or flooded 
soils: a review. Agron. J. 112, 1475–1501.

Kim, Y.-U., Webber, H., Adiku, S.G.K., Nóia Júnior, R., De, S., Deswarte, J.-C., Asseng, S., 
Ewert, F., 2024. Mechanisms and modelling approaches for excessive rainfall stress 
on cereals: waterlogging, submergence, lodging, pests and diseases. Agric. Meteor. 
344, 109819.

Kulkarni, A.V., Shirsat, T.S., Kulkarni, A., Negi, H.S., Bahuguna, I.M., Thamban, M., 
2021. State of Himalayan cryosphere and implications for water security. Water 
Secur 14, 100101.

Kuppel, S., Houspanossian, J., Nosetto, M.D., Jobbágy, E.G., 2015. What does it take to 
flood the Pampas?: Lessons from a decade of strong hydrological fluctuations 
(Online). Water Resour. Res 51, 2937–2950. https://doi.org/10.1002/ 
2015WR016966.

León, J., Castillo, M.C., Gayubas, B., 2021. The hypoxia–reoxygenation stress in plants 
(Online). J. Exp. Bot. 72, 5841–5856. https://doi.org/10.1093/jxb/eraa591.

Li, H., Cai, J., Liu, F., Jiang, D., Dai, T., Cao, W., 2012. Generation and scavenging of 
reactive oxygen species in wheat flag leaves under combined shading and 
waterlogging stress. Funct. Plant Biol. 39, 71–81.

Liu, B., Asseng, S., Müller, C., Ewert, F., Elliott, J., Lobell, D.B., Martre, P., Ruane, A.C., 
Wallach, D., Jones, J.W., Rosenzweig, C., Aggarwal, P.K., Alderman, P.D., 
Anothai, J., Basso, B., Biernath, C., Cammarano, D., Challinor, A., Deryng, D., 
Sanctis, G.D., Doltra, J., Fereres, E., Folberth, C., Garcia-Vila, M., Gayler, S., 
Hoogenboom, G., Hunt, L.A., Izaurralde, R.C., Jabloun, M., Jones, C.D., 
Kersebaum, K.C., Kimball, B.A., Koehler, A.-K., Kumar, S.N., Nendel, C., O’leary, G. 
J., Olesen, J.E., Ottman, M.J., Palosuo, T., Prasad, P.V.V., Priesack, E., Pugh, T.A.M., 
Reynolds, M., Rezaei, E.E., Rötter, R.P., Schmid, E., Semenov, M.A., Shcherbak, I., 
Stehfest, E., Stöckle, C.O., Stratonovitch, P., Streck, T., Supit, I., Tao, F., 
Thorburn, P., Waha, K., Wall, G.W., Wang, E., White, J.W., Wolf, J., Zhao, Z., 
Zhu, Y., 2016. Similar estimates of temperature impacts on global wheat yield by 
three independent methods. Nat. Clim. Change 6, 1130–1136.

Liu, K., Harrison, M.T., Archontoulis, S.V., Huth, N., Yang, R., Liu, D.L., Yan, H., 
Meinke, H., Huber, I., Feng, P., Ibrahim, A., Zhang, Y., Tian, X., Zhou, M., 2021a. 
Climate change shifts forward flowering and reduces crop waterlogging stress. 
Environ. Res. Lett. 16.

Liu, K., Harrison, M.T., Archontoulis, S.V., Huth, N., Yang, R., Liu, D.L., Yan, H., 
Meinke, H., Huber, I., Feng, P., Ibrahim, A., Zhang, Y., Tian, X., Zhou, M., 2021b. 
Climate change shifts forward flowering and reduces crop waterlogging stress. 
Environ. Res. Lett. 16, 94017.

Liu, K., Harrison, M.T., Ibrahim, A., Manik, S.M.N., Johnson, P., Tian, X., Meinke, H., 
Zhou, M., 2020a. Genetic factors increasing barley grain yields under soil 
waterlogging. Food Energy Secur 9, e238.

Liu, K., Harrison, M.T., Shabala, S., Meinke, H., Ahmed, I., Zhang, Y., Tian, X., Zhou, M., 
2020b. The state of the art in modeling waterlogging impacts on plants: what do we 
know and what do we need to know. Earths Future 8.

Liu, K., Harrison, M.T., Shabala, S., Meinke, H., Ahmed, I., Zhang, Y., Tian, X., Zhou, M., 
2020c. The state of the art in modeling waterlogging impacts on plants: what do we 
know and what do we need to know. Earths Future 8, e2020EF001801.

Liu, K., Harrison, M.T., Yan, H., Liu, D.L., Meinke, H., Hoogenboom, G., Wang, B., 
Peng, B., Guan, K., Jaegermeyr, J., Wang, E., Zhang, F., Yin, X., Archontoulis, S., 
Nie, L., Badea, A., Man, J., Wallach, D., Zhao, J., Benjumea, A.B., Fahad, S., Tian, X., 
Wang, W., Tao, F., Zhang, Z., Rötter, R., Yuan, Y., Zhu, M., Dai, P., Nie, J., Yang, Y., 
Zhang, Y., Zhou, M., 2023. Silver lining to a climate crisis in multiple prospects for 
alleviating crop waterlogging under future climates. Nat. Commun. 14, 765.

Liu B., Martre P., Ewert F., Porter J.R., Challinor A.J., Müller C., Ruane A.C., Waha K., 
Thorburn P.J., Aggarwal P.K., Ahmed M., Balkovič J., Basso B., Biernath C., Bindi M., 
Cammarano D., De Sanctis G., Dumont B., Espadafor M., Eyshi Rezaei E., Ferrise R., 
Garcia-Vila M., Gayler S., Gao Y., Horan H., Hoogenboom G., Izaurralde R.C., Jones 
C.D., Kassie B.T., Kersebaum K.C., Klein C., Koehler A.-K., Maiorano A., Minoli S., 
Montesino San Martin M., Naresh Kumar S., Nendel C., O’leary G.J., Palosuo T., 
Priesack E., Ripoche D., Rötter R.P., Semenov M.A., Stöckle C., Streck T., Supit I., 
Tao F., Van Der Velde M., Wallach D., Wang E., Webber H., Wolf J., Xiao L., Zhang 
Z., Zhao Z., Zhu Y. and Asseng S. 2019 Global wheat production with 1.5 and 2.0◦C 
above pre-industrial warming Glob Chang Biol 25 1428–44.

Lizaso J.I. 1993 Flooding and field grown maize: above- and below-ground responses and 
a simulation model (Michigan State University).

Malik, A.I., Colmer, T.D., Lambers, H., Schortemeyer, M., 2001. Changes in physiological 
and morphological traits of roots and shoots of wheat in response to different depths 
of waterlogging. Funct. Plant Biol. 28, 1121–1131.

Marti, J., Savin, R., Slafer, G.A., 2015. Wheat yield as affected by length of exposure to 
waterlogging during stem elongation. J. Agron. Crop Sci. 201, 473–486.

Nóia Júnior, R., De, S., Asseng, S., García-Vila, M., Liu, K., Stocca, V., Dos Santos 
Vianna, M., Weber, T.K.D., Zhao, J., Palosuo, T., Harrison, M.T., 2023a. A call to 
action for global research on the implications of waterlogging for wheat growth and 
yield. Agric. Water Manag 284.

Nóia Júnior, R., De, S., Asseng, S., García-Vila, M., Liu, K., Stocca, V., Dos Santos 
Vianna, M., Weber, T.K.D., Zhao, J., Palosuo, T., Harrison, M.T., 2023b. A call to 
action for global research on the implications of waterlogging for wheat growth and 
yield. Agric. Water Manag 284, 108334.

Nóia Júnior, R., De, S., Deswarte, J.-C., Cohan, J.-P., Martre, P., Van Der Velde, M., 
Lecerf, R., Webber, H., Ewert, F., Ruane, A.C., Slafer, G.A., Asseng, S., 2023c. The 
extreme 2016 wheat yield failure in France. Glob. Chang Biol. 29, 3130–3146.

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

12 

http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref2
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref3
https://www.sciencedirect.com/science/article/pii/S0378429016301277
https://www.sciencedirect.com/science/article/pii/S0378429016301277
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref5
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref6
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref6
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref6
https://doi.org/10.1111/pce.12119
https://doi.org/10.1111/pce.12119
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref8
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref8
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref8
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref8
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref8
https://doi.org/10.1111/nph.17048
https://doi.org/10.1111/nph.17048
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref10
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref10
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref10
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref11
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref11
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref11
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref12
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref12
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref13
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref13
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref13
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref14
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref14
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref14
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref14
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref15
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref15
https://doi.org/10.1038/s43016-025-01179-y
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref17
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref17
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref17
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref18
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref18
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref18
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref19
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref19
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref19
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref20
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref20
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref21
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref21
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref21
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref22
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref22
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref22
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref23
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref23
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref23
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref23
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref24
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref24
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref24
https://doi.org/10.1002/2015WR016966
https://doi.org/10.1002/2015WR016966
https://doi.org/10.1093/jxb/eraa591
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref27
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref27
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref27
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref28
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref29
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref29
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref29
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref29
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref30
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref30
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref30
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref30
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref31
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref31
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref31
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref32
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref32
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref32
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref33
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref33
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref33
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref34
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref35
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref35
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref35
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref36
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref36
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref37
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref37
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref37
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref37
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref38
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref38
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref38
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref38
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref39
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref39
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref39


Nóia Júnior, R., De, S., Martre, P., Finger, R., Van Der Velde, M., Ben-Ari, T., Ewert, F., 
Webber, H., Ruane, A.C., Asseng, S., 2021. Extreme lows of wheat production in 
Brazil. Environ. Res. Lett. 16, 104025.

Nóia-Júnior, R., De, S., Asseng, S., Müller, C., Deswarte, J.-C., Cohan, J.-P., Martre, P., 
2025a. Negative impacts of climate change on crop yields are underestimated. 
Trends Plant Sci. Online. 〈https://www.sciencedirect.com/science/article/pii/ 
S136013852500130X〉.

Nóia-Júnior, R., De, S., Martre, P., Deswarte, J.-C., Cohan, J.-P., Van Der Velde, M., 
Webber, H., Ewert, F., Ruane, A.C., Ben-Ari, T., Asseng, S., 2025b. Past and future 
wheat yield losses in France’s breadbasket. Field Crops Res 322, 109703. 〈htt 
ps://www.sciencedirect.com/science/article/pii/S0378429024004568〉 (Online). 

Pan, J., Sharif, R., Xu, X., Chen, X., 2021. Mechanisms of waterlogging tolerance in 
plants: research progress and prospects. Front. Plant Sci. 11.

Pedersen, O., Sauter, M., Colmer, T.D., Nakazono, M., 2021. Regulation of root adaptive 
anatomical and morphological traits during low soil oxygen. N. Phytol. 229, 42–49.

Porter C.H., Braga R. and Jones J.W. 1999 An Approach for Modular Crop Model 
Development.

R Core Team 2017 R: A Language and Environment for Statistical Computing R 
Foundation for Statistical Computing, Vienna, Austria {ISBN} 3-900051-07-0.

Reynolds, J.F., Acock, B., 1985. Predicting the response of plants to increasing carbon 
dioxide: a critique of plant growth models. Ecol. Model. 29, 107–129. 〈https://www. 
sciencedirect.com/science/article/pii/0304380085900493〉 (Online). 

Reynolds, M.P., Slafer, G.A., Foulkes, J.M., Griffiths, S., Murchie, E.H., Carmo-Silva, E., 
Asseng, S., Chapman, S.C., Sawkins, M., Gwyn, J., Flavell, R.B., 2022. A wiring 
diagram to integrate physiological traits of wheat yield potential (Online). Nat. Food 
3, 318–324. https://doi.org/10.1038/s43016-022-00512-z.

Röll, G., Memic, E., Graeff-Hönninger, S., 2020. Implementation of an automatic time- 
series calibration method for the DSSAT wheat models to enhance multi-model 
approaches. Agron. J. 112, 3891–3912.

Rötter, R.P., Carter, T.R., Olesen, J.E., Porter, J.R., 2011. Crop–climate models need an 
overhaul (Online). Nat. Clim. Chang 1, 175–177. https://doi.org/10.1038/ 
nclimate1152.

Rötter, R.P., Hoffmann, M.P., Koch, M., Müller, C., 2018. Progress in modelling 
agricultural impacts of and adaptations to climate change. Curr. Opin. Plant Biol. 45, 
255–261. 〈https://www.sciencedirect.com/science/article/pii/S136952661 
7302339〉 (Online). 

Salazar, O., Vargas, J., Nájera, F., Seguel, O., Casanova, M., 2014. Monitoring of nitrate 
leaching during flush flooding events in a coarse-textured floodplain soil. Agric. 
Water Manag 146, 218–227.

Sayre, K.D., Van Ginkel, M., Rajaram, S., Ortiz-Monasterio, I., 1994. Tolerance to 
waterlogging losses in spring bread wheat: effect of time of onset on expression. 
Annu. Wheat Newsl. 40, 165–171.

Shelia V., Asseng S., Porter C. and Hoogenboom G. 2019 SIMULATION OF A PERCHED 
WATER TABLE WITH IMPACT ON WHEAT CROP GROWTH Agricultural and 
Biological Engineering University of Florida (Gainesville).

Slafer, G.A., Elia, M., Savin, R., García, G.A., Terrile, I.I., Ferrante, A., Miralles, D.J., 
González, F.G., 2015. Fruiting efficiency: an alternative trait to further rise wheat 
yield. Food Energy Secur 4, 92–109.

Slafer, G.A., Foulkes, M.J., Reynolds, M.P., Murchie, E.H., Carmo-Silva, E., Flavell, R., 
Gwyn, J., Sawkins, M., Griffiths, S., 2023. A ‘wiring diagram’ for sink strength traits 
impacting wheat yield potential. J. Exp. Bot. 74, 40–71.

Trought, M.C.T., Drew, M.C., 1982. Effects of waterlogging on young wheat plants 
(Triticum aestivum L.) and on soil solutes at different soil temperatures. Plant Soil 
69, 311–326.

Van Der Velde M., Lecerf R., D’andrimont R. and Ben-Ari T. 2020 Chapter 8 - Assessing 
the France 2016 extreme wheat production loss—Evaluating our operational 
capacity to predict complex compound events ed J Sillmann, S Sippel and S B T-C E 
and T I for I and R A Russo (Elsevier) pp 139–158.

Van Veen, H., Akman, M., Jamar, D.C.L., Vreugdenhil, D., Kooiker, M., Van 
Tienderen, P., Voesenek, L.A.C.J., Schranz, M.E., Sasidharan, R., 2014. Group VII 
Ethylene Response Factor diversification and regulation in four species from flood- 
prone environments. Plant Cell Environ. 37, 2421–2432.

Webber, H., Lischeid, G., Sommer, M., Finger, R., Nendel, C., Gaiser, T., Ewert, F., 2020. 
No perfect storm for crop yield failure in Germany. Environ. Res. Lett. 15, 104012.

Willmott, C.J., Ackleson, S.G., Davis, R.E., Feddema, J.J., Klink, K.M., Legates, D.R., 
O’donnell, J., Rowe, C.M., 1985. Statistics for the evaluation and comparison of 
models. J. Geophys Res 90, 8995.

Yan, H., Harrison, M.T., Liu, K., Wang, B., Feng, P., Fahad, S., Meinke, H., Yang, R., 
Liu, D.L., Archontoulis, S., Huber, I., Tian, X., Man, J., Zhang, Y., Zhou, M., 2022. 
Crop traits enabling yield gains under more frequent extreme climatic events. Sci. 
Total Environ. 808, 152170.

Yemelyanov, V.V., Puzanskiy, R.K., Shishova, M.F., 2023. Plant life with and without 
oxygen: a metabolomics approach. Int J. Mol. Sci. 24, 16222.

Zampieri, M., Ceglar, A., Dentener, F., Toreti, A., 2017. Wheat yield loss attributable to 
heat waves, drought and water excess at the global, national and subnational scales. 
Environ. Res. Lett. 12, 64008.

Zhao, C., Liu, B., Piao, S., Wang, X., Lobell, D.B., Huang, Y., Huang, M., Yao, Y., Bassu, S., 
Ciais, P., Durand, J.-L., Elliott, J., Ewert, F., Janssens, I.A., Li, T., Lin, E., Liu, Q., 
Martre, P., Müller, C., Peng, S., Peñuelas, J., Ruane, A.C., Wallach, D., Wang, T., 
Wu, D., Liu, Z., Zhu, Y., Zhu, Z., Asseng, S., 2017. Temperature increase reduces 
global yields of major crops in four independent estimates. Proc. Natl. Acad. Sci. 
114, 9326. LP – 9331. 

R.S. Nóia-Júnior et al.                                                                                                                                                                                                                         Field Crops Research 333 (2025) 110090 

13 

http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref40
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref40
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref40
https://www.sciencedirect.com/science/article/pii/S136013852500130X
https://www.sciencedirect.com/science/article/pii/S136013852500130X
https://www.sciencedirect.com/science/article/pii/S0378429024004568
https://www.sciencedirect.com/science/article/pii/S0378429024004568
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref43
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref43
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref44
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref44
https://www.sciencedirect.com/science/article/pii/0304380085900493
https://www.sciencedirect.com/science/article/pii/0304380085900493
https://doi.org/10.1038/s43016-022-00512-z
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref47
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref47
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref47
https://doi.org/10.1038/nclimate1152
https://doi.org/10.1038/nclimate1152
https://www.sciencedirect.com/science/article/pii/S1369526617302339
https://www.sciencedirect.com/science/article/pii/S1369526617302339
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref50
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref50
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref50
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref51
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref51
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref51
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref52
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref52
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref52
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref53
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref53
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref53
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref54
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref54
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref54
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref55
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref55
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref55
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref55
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref56
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref56
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref57
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref57
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref57
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref58
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref58
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref58
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref58
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref59
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref59
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref60
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref60
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref60
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61
http://refhub.elsevier.com/S0378-4290(25)00355-7/sbref61

	Enabling modeling of waterlogging impact on wheat
	1 Introduction
	2 Material and methods
	2.1 Review of waterlogging impacts on wheat grain number, average grain size and grain yield
	2.2 Modeling waterlogging on wheat with DSSAT-NWheat
	2.2.1 Modeling soil waterlogging
	2.2.2 Waterlogging impacts on roots
	2.2.3 Waterlogging impact on carbohydrate accumulation
	2.2.4 Waterlogging impact on potential grain size

	2.3 Evaluation of the adapted waterlogging module
	2.3.1 Outdoor controlled waterlogging experiment in Lleida, Spain
	2.3.2 Field experiment from Loirauxence, La Jaillière in France
	2.3.3 Sensitivity analysis

	2.4 Data analysis

	3 Results
	3.1 Review of waterlogging impacts on wheat yield components
	3.2 DSSAT-NWheat performance to simulate the impacts of pre-anthesis waterlogging on wheat yield components
	3.3 Sensitivity analysis

	4 Discussion
	4.1 Modeling performance of the new waterlogging module in DSSAT-NWheat
	4.2 Simulated impacts of waterlogging on wheat yield components
	4.3 Factors that affect the intensity of waterlogging impacts on wheat growth

	Author contributions
	CRediT authorship contribution statement
	Declaration of Competing Interest
	Acknowledgements
	Appendix A Supporting information
	Data availability
	References