International Journal of Biometeorology (2022) 66:2237–2249
https://doi.org/10.1007/s00484-022-02352-9
ORIGINAL PAPER
Variable climate suitability for wheat blast (Magnaporthe oryzae 
pathotype Triticum) in Asia: results from a continental‑scale modeling 
approach
Carlo Montes1 · Sk. Ghulam Hussain2 · Timothy J. Krupnik2
Received: 18 January 2022 / Revised: 9 August 2022 / Accepted: 16 August 2022 / Published online: 22 August 2022
© The Author(s) 2022
Abstract
Crop fungal diseases constitute a major cause of yield loss. The development of crop disease monitoring and forecasting tools 
is an important effort to aid farmers in adapting to climate variability and change. Recognizing weather as a main driver of 
fungal disease outbreaks, this work assesses the climate suitability for wheat blast (Magnaporthe oryzae pathotype Triticum, 
MoT) development in Asian wheat-producing countries. MoT was reported for the first time in Bangladesh in 2016 and 
could spread to other countries, provided that environmental conditions are suitable to spore development, distribution, and 
infection. With results from a generic infection model driven by air temperature and humidity, and motivated by the neces-
sity to assess the potential distribution of MoT based on the response to weather drivers only, we quantify potential MoT 
infection events across Asia for the period 1980–2019. The results show a potential higher incidence of MoT in Bangladesh, 
Myanmar, and some areas of India, where the number of potential infection (NPI) events averaged up to 15 during wheat 
heading. Interannual trends show an increase in NPI over those three countries, which in turns show their higher interannual 
variability. Cold/dry conditions in countries such as Afghanistan and Pakistan appear to render them unlikely candidates 
for MoT establishment. The relationship between seasonal climate anomalies and NPI suggests a greater association with 
relative humidity than with temperature. These results could help to focus future efforts to develop management strategies 
where weather conditions are conducive for the establishment of MoT.
Keywords Crop damage · Fungal disease · Infection model · Early warning · Climate services
Introduction new diseases that have been imported (Bebber et al. 2014). 
This is the case of wheat blast (MoT) disease caused by 
The occurrence of crop diseases caused by fungal patho- the fungus Magnaporthe oryzae pathotype Triticum (MoT), 
gens is among the main factors affecting crop yields glob- which evolved and has been present in South America since 
ally (Figueroa et al. 2018). Although advances in resistant 1985 (Igarashi et al. 1986). MoT was reported for the first 
varieties and efficient and environmentally friendly con- time in South Asia in Bangladesh in 2016 (Malaker et al. 
trol options are numerous, losses associated with fungal 2016; Ceresini et al. 2018), and more recently in Southern 
diseases remain very important and, in some cases, dev- Africa in Zambia (Tembo et al. 2020). MoT is considered a 
astating (Fisher et al. 2012). The risk of crop disease out- potentially devastating fungal disease in countries where it 
breaks is increasing given the global trade of agricultural has been historically present, such as Brazil (Igarashi et al. 
commodities, which can increase the exposure of crops to 1986), Bolivia (Barea and Toledo 1996), and Argentina 
(Perelló et al. 2015), causing periodic and significant yield 
losses (Cruz et al. 2016; Duveiller et al. 2016). MoT is also 
*  Carlo Montes an emerging threat to wheat production and food security 
 c.montes@cgiar.org in countries where wheat is a major staple, such as in Asia 
1 International Maize and Wheat Improvement Center (Islam et al. 2020a, b).
(CIMMYT), Texcoco, Mexico In Bangladesh, the first MoT outbreak affected about 
2 International Maize and Wheat Improvement Center 15,000 ha of wheat, with an estimated reduction of nearly 
(CIMMYT), Dhaka, Bangladesh 30% in production in 2016 (Islam et al. 2020a, b; Yesmin 
Vol.:(012 3456789)
2238 International Journal of Biometeorology (2022) 66:2237–2249
et al. 2020). Although subsequent outbreaks have not been climate variables (e.g., atmospheric humidity and tempera-
recorded, the disease remains present in Bangladesh with ture), provided that parameters are adequately set for a spe-
low to moderate severity when detected (Singh et al. 2021). cific disease (Bregaglio et al. 2012; Bregaglio and Donatelli 
The impacts of MoT on wheat yields and grain quality can 2015). In the case of MoT, Fernandes et al. (2017) devel-
be devastating for susceptible cultivars, but they can vary oped a wheat blast–specific model aiming at implementing 
greatly in response to other factors such as weather con- an early warning system for Brazil, which was applied and 
ditions, growth stage, or planting date (Cruz and Valent evaluated at the local level using a single-location approach 
2017). In this way, when weather conditions are suitable and then extended to Bangladesh (Fernandes et al. 2021). 
for MoT infection, grain yield losses can range from slight The need for decision-making tools for farmers from other 
to total (Duveiller et al. 2011; Singh et al. 2021), as it has wheat-growing regions in Asia has been emphasized later 
been reported in South American countries such as Brazil or (Singh et al. 2020), given the potential risk for the range of 
Bolivia, where yield losses have reached up to 100% (Gou- disease expansion (Islam et al. 2020a, b). In this context, the 
lart and Paiva 1992; Barea and Toledo 1996). aim of this work is to provide a large-scale and long-term 
Although there are still no official reports of the presence assessment of the climate suitability for MoT development 
of MoT in others countries than Bangladesh in Asia, studies over wheat-growing areas of Asia in terms of mean histori-
of climate suitability for MoT have suggested that it may cal (1980–2019) weather conditions and interannual vari-
spread to areas with humid and warm climates in neigh- ability, based on the analysis of the results obtained from 
boring countries such as India or Pakistan (Motaleb et al. high-resolution meteorological data and a generic infection 
2018a). Research has also suggested that Ethiopia (Duveiller model. The results from this work, which represent an esti-
et al. 2011) and the USA (Cruz et al. 2016) may be risk- mate of the potential pressure that can be exerted by MoT 
prone. In these regions, both seed-born and air-born spore driven by background meteorological conditions, can con-
propagation, suitable weather conditions, and disease sus- tribute to the understanding of the spatial patterns in suitable 
ceptible cultivars can act synergistically to increase the risk weather conditions for MoT and their main large-scale driv-
of disease outbreak, potentially threatening food security ers, and can potentially provide guidance for future efforts 
(Ceresini et al. 2018). These risks have motivated a number and regional prioritization in the development of early warn-
of efforts to monitor pathogen presence (Fernandes et al. ing systems based on weather monitoring and forecasting.
2021; Islam et al. 2016; Yesmin et al. 2020), and to develop 
management strategies including resistant varieties (Hos-
sain et al. 2019), chemical and nonchemical control methods Data and methods
(Singh et al. 2021), and early warning systems (Fernandes 
et al. 2021; Kim and Choi 2020). Study area
Multiple tools have been developed for the monitoring 
and forecasting of fungal disease outbreaks based on field Eight Asian countries were identified based on the extent of 
observations or empirical and deterministic models com- wheat cultivation and consumption and the recent emergence 
bining weather variables to generate early warnings of the of wheat blast disease in 2016 in Bangladesh, which are sum-
potential risk of disease outbreaks (Launay et al. 2014). Con- marized in Table 1 for 2019. In alphabetic order, these include 
sidering MoT in Asia, studies of its potential spread have Afghanistan, Bangladesh, Bhutan, China, India, Myanmar, 
been carried out using monthly climate statistics (e.g., Mot- Nepal, and Pakistan. In these countries, winter wheat is planted 
aleb et al. 2018a) or limited temporal and spatial domains in the autumn, with a long vegetative stage during the dry sea-
(e.g., Kim and Choi 2020). No large-scale, continental, and son in winter, and the reproductive stage generally occurring 
high-resolution assessments have conversely been carried with the onset of the spring. In addition, spring wheat is culti-
out in Asia. However, given increasing availability of envi- vated in areas with mild winters such as in India, and at eleva-
ronmental data and computing capacities, the use of simula- tion in the Himalayas, where wheat is sown in autumn and har-
tion models to diagnose and forecast favorable conditions for vested after the winter without vernalization, though land area 
the development of crop diseases has grown in importance devoted to spring wheat is limited in South Asia (Curtis 2002; 
(Donatelli et al. 2017). Krupnik et al. 2021). For this reason, this study focuses on 
Diagnosis and applications vary from regional assess- winter wheat as the predominant crop. Wheat is a major staple 
ments of climate suitability (Bebber et al. 2017), sensitiv- food in Afghanistan and Pakistan, with a total production of 
ity analysis to environmental drivers and parameterizations 4.9 and 24.3 Mt (million tonnes) over 2.3 and 8.7 Mha (million 
(Bregaglio et al. 2012), and future projections in risks asso- ha) in 2019, respectively (Fig. S1; FAOSTAT 2021). Wheat 
ciated with climate change (Bregaglio et al. 2013). The latter consumption has been increasing progressively in India, Bhu-
suggested that the suitable conditions for the establishment tan, Myanmar, and Bangladesh, becoming the second most 
of fungal diseases can be well captured by models forced by important staple food after rice (Motaleb et al. 2018b). China 
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2239
Table 1  Main wheat production Country Production (tonnes) Area harvested (ha) Aver-
statistics for the eight Asian age yield 
countries considered in this (tonnes/ha)
study for the year 2019. Values 
in brackets correspond to the Afghanistan 4,890,000 (39,767) 2,334,000 (− 372) 2.095 (0.02)
slope of the linear fit of the 
corresponding statistics for the Bangladesh 1,016,811 (24,511) 330,348 (8360) 3.078 (0.04)
period 1961 through 2019. Data Bhutan 1319 (− 18) 1004 (− 68) 1.314 (0.02)
from FAOSTAT 2021 China 133,596,300 (2,079,597) 23,730,000 (− 37,782) 5.630 (0.09)
India 103,596,230 (1,612,311) 29,318,790 (297,338) 3.533 (0.05)
Myanmar 110,663 (2125) 58,866 (289) 1.880 (0.02)
Nepal 2,005,665 (34,027) 703,992 (12,368) 2.849 (0.03)
Pakistan 24,348,983 (412,302) 8,677,730 (73,567) 2.806 (0.04)
is the world’s largest wheat producer, with 133.5 Mt grown Tmin, Tmax, and Topt are the minimum, maximum, and opti-
on 23.7 Mha in 2019; India is the second largest producer, mum temperatures for infection, respectively. These cardi-
growing 103.6 Mt in 2019 on 29.3 Mha (FAOSTAT 2021). nal temperatures were taken from Cruz et al. (2016), who 
Bangladesh conversely is a net importer of wheat, producing suggested the following values for MoT: Tmin = 10  °C, 
1 Mt over 0.33 Mha in 2019, cultivated exclusively during Tmax = 32 °C, and Topt = 27.5 °C. As an example, Fig. S2 
the winter after monsoon season rice fields are drained. In shows the resulting shape of f(T), where, following a slow 
Nepal, wheat is grown in the low-lying Terai (up to 500 m response, exponential increasing response to temperature is 
above sea level) and in the Himalayan mid-hills (Morris et al. observed between Tmin and around 20 °C, which turns from 
1994; Krupnik et al. 2021), with a total of 2 Mt produced from almost linear to a decreasing-rate increment until Topt, to 
0.7 Mha in 2019. In Myanmar, more than 90% of the wheat then drops rapidly until f(T) = 0 at Tmax. The air temperature 
is found in the hilly Sagaing and Shan states (USDA, 2019), response f(T) is subsequently scaled to the wetness dura-
with a production of 110,000 tonnes from 59,000 ha in 2019. tion requirement for infection according to the following 
In Bhutan, wheat is also produced at elevation, reaching 1319 relationship:
tonnes from 1004 ha in 2019 (Tshewang et al. 2017).
{
WDmin WD
, if min
< WD
W(T) = f (T) f (T) max ,
Modeling potential wheat blast infections (2)
0, elsewhere
Model description where W(t) (dimensionless, values from 0 to 1) corresponds 
to the wetness response function, and WDmin and WDmax 
The generic infection model developed by Magarey et al. (hours), taken as 12 and 24, respectively (Cruz et al. 2016), 
(2005) was used to assess the climate suitability of MoT infec- are the minimum and maximum leaf wetness duration 
tions. This model has been previously applied for large-scale requirement for infection, respectively. Therefore, when the 
studies of fungal disease infections given the biological sig- infection models use hourly forcing data, it is necessary to 
nificance of its parameterizations and simple implementation account for the number of hours that may interrupt a wet 
(Bregaglio et al. 2012, 2013). The model considers both the period without terminating the infection process, as Magarey 
effect of hourly air temperature and plant surface wetness (or et al. (2005) explained. For this, the model considers the 
relative humidity) duration on the development response of a impact of critical dry periods through the parameter D50 
generic fungal pathogen by using two functions describing its that is calculated as:
sensitivity to air temperature and humidity. The model uses the {
air temperature response function proposed by Yan and Hunt W
W 1 +W2, ifD ≤ D50
sum = , (3)
(1999), which combines a set of pathogen’s cardinal tempera- W1,W2, elsewhere
tures to estimate the shape of the response as: where Wsum is the sum of the surface wetting periods and W1 
( )( ) and W2 indicate two wet periods separated by a dry period 
T (Topt−Tmin)∕(Tmax−Topt )
max − T T − Tmin
f (T) = , (D, in hours). As in Magarey et al. (2005), D50 is defined as 
Tmax − Topt Topt − Tmin the duration of a dry period at relative humidity < 95% that 
(1) will result in a 50% reduction in disease compared with a 
where f(T) (dimensionless, values from 0 to 1) is the temper- continuous wetness period. Therefore, if D > D50, the model 
ature response function; T (°C) is the hourly air temperature; considers the two wet periods as separate wetting events. 
1 3
2240 International Journal of Biometeorology (2022) 66:2237–2249
When the plant surfaces are wet and f(T) > 0, the model are obtained from dewpoint (Td) temperature and actual air 
assumes that inoculant is present in the environment and temperature (T), respectively (Allen et al. 1998):
adds a cohort of spores. Infection events are triggered when 
( )
17.27×Td
the value of Wsum ranges between WDmin and WDmax (Brega- e = 0.611 × exp 237.3+Td (4)
a
glio et al. 2012). Although the values of the D50 parameter 
were gathered by Magarey et al. (2005) for a number of ( )
17.27×T
species of fungal diseases, D50 has not yet been calibrated e = 0.611 × exp 237.3+T (5)
s
for MoT. We however included a value of D50 of 4, which 
was used by Bregaglio et al. (2013) for the assessment of ea
potential infections of Pyricularia oryzae, a MoT anamorph, RH = 100 ×
e (6)
s
in Europe (Martínez et al. 2019). The above set of equations 
were solved for the wheat heading period, which was esti- with ea and es are expressed in kPa, and temperatures in 
mated using a phenological model based on thermal time °C. Maps of the seasonal climatology of these variables are 
accumulation, as presented below. provided in Fig. S3.
Representing wheat distribution and phenology
Infection model forcing
The spatial distribution of wheat area was represented using 
Multiple global gridded climate products are currently avail- the Spatial Production Allocation Model SPAM 2010 v1.0 
able, which can be potentially used to model and diagnose global crop production data product developed by the Inter-
crop diseases. However, meteorological information must be national Food Policy Research Institute (IFPRI) (Wood-
provided at appropriate temporal and spatial scales, given Sichra et  al. 2016; International Food Policy Research 
the behavior of crop pathogens. Among the meteorological Institute (IFPRI) 2019). This product provides statistics on 
variables most used for crop disease modeling are air tem- crop production by merging sub-national statistics, satellite-
perature, precipitation, relative humidity, and leaf wetness derived land cover, environmental crop suitability, popula-
(Donatelli et al. 2017). More complex and highly demanding tion, cropping systems, and markets, among other variables. 
in computer resources, transport-based Lagrangian models The operational product is generated after the crop produc-
require wind speed and direction to calculate fungal spores’ tion data derived from the above-mentioned information 
trajectories and deposition (Meyer et al. 2017). is aggregated into a regular grid of spatial resolution of 
Most global gridded climate products are provided at around 10 km × 10 km using a cross-entropy method (You 
daily time-steps as the higher temporal resolution, which and Wood, 2006). In this work, the original data grid was 
may be limiting for the simulation of crop diseases. bilinearly interpolated to the 0.25° × 0.25° climate forcing 
Although there are methods to statistically disaggregate resolution and then converted into a binary mask (Fig. S4).
daily time series to hourly values via empirical models or MoT infections were estimated for the phenological 
weather generators (Bregaglio et al. 2010), their accuracy period from heading to the end of the reproductive phase 
can be limited by the available historical data and their (maturity). The starting and ending dates of this suscepti-
implementation can be difficult when it comes to large ble period were calculated using crop growth modeling and 
datasets. This study used hourly data from the last genera- global climate products. Thus, the spatially explicit critical 
tion European Centre for Medium-Range Weather Forecasts dates necessary for bounding the modeling time window are 
(ECMWF) ERA5 global atmospheric reanalysis as mete- sowing date, emergence, beginning of the heading stage, 
orological observations to force the infection model. This and beginning of physiological maturity. After represent-
product is provided at an hourly time scale with a horizontal ing the spatial distribution of wheat, the key phenological 
resolution of 0.25° × 0.25° (~ 31 km), covering the period dates were stated. First, winter wheat sowing dates were 
1979 to present for single (surface) and multiple vertical obtained from the interpolated Crop Calendar Dataset of 
levels (Hersbach et al. 2020). ERA5 is generated using a Sacks et al.’s (2010) product, which provides 5′ × 5′ spatial 
4D-Var data assimilation scheme to optimally combine out- resolution global dates of crop sowing and harvest dates rep-
puts from the ECMWF Integrated Forecasting System with resentative of the year 2000. Here, the original resolution 
satellite and ground observations. We utilized the hourly dataset was bilinearly aggregated to match the 0.25° × 0.25° 
ERA5 air and dewpoint temperature at surface level (2 m ERA5 resolution (Fig. S4).
height), and rainfall data for the period from January 1980 The wheat heading period was estimated using point-
through December 2019. Relative humidity (RH) for the based simulations with the CSM-CROPSIM-CERES-wheat 
infection model was calculated using the widely used equa- model, embedded in the Decision Support System for Agro-
tion involving actual and saturated vapor pressure, which technology Transfer (DSSAT) v.4.6 (Jones et al. 2003). 
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2241
CROPSIM-CERES simulates wheat phenology according 
to the Zadoks stages (Zadoks et al. 1974) as a function of 
growing degree day accumulation and accounting for envi-
ronmental stresses, vernalization, and photoperiod effects. 
Simulations were performed over a set of 163 wheat-grow-
ing locations belonging to the International Wheat Improve-
ment Network (IWIN; Reynolds et al. 2017) for the period 
1979 through 2019 (Fig. S5). The meteorological forcing 
(air temperature, solar radiation, rainfall, relative humidity, 
wind speed) was performed using ECMWF’s AgERA5 prod-
uct (Copernicus Climate Change Service (C3S), 2019), a 
statistically downscaled (0.1° × 0.1°) daily version of ERA5. 
Global soil profiles from the HC27 product (Koo and Dimes 
2010) were used to provide soil physical and chemical prop-
erties to CROPSIM-CERES. Genetic coefficients necessary 
for wheat simulations were set based on the cultivar distribu-
tion over the International Maize and Wheat Improvement 
Center’s (CIMMYT’s) wheat mega-environments, which 
correspond to homogeneous agroecological zones for wheat 
cultivation (Pequeno et al. 2021). A comparison between Fig. 1  Flow diagram of the modeling approach for number of poten-
CROPSIM-CERES simulated number of days from sow- tial infections (NPI) of MoT. See text for acronyms and product 
ing to anthesis and IWIN observations showed a normalized names
root mean square error of 7.6% (data not shown). Finally, 
using sowing dates from Sacks et al. (2010), simulated dates series. Anomalies were obtained by removing the corre-
of anthesis and maturity were obtained for every location, sponding long-term average.
and the heading date was assumed to occur 10 days before 
anthesis. Both heading and maturity dates were bilinearly 
interpolated to 0.25° × 0.25° working spatial resolution of Results
the SPAM product (Fig. S5). A flow diagram schematically 
describing the main steps of the modeling approach is shown Mean patterns and interannual variability 
in Fig. 1. of number of potential MoT infections
Figure 2a shows the map of interannual mean total seasonal 
Analysis NPI in Asia. The mean seasonal NPI is 7.5 (median of 6) 
and interquartile range from 4 to 9 (Fig. 3a). In Fig. 2a, while 
Data analysis involved three analytical steps. The first step 57.6% of SPAM wheat grid cells present suitable conditions 
focused on the quantification of the average and interan- for MoT, 6.7% of them present favorable conditions during 
nual variability (1980–2019) in the weather-driven number all 39 years studied. The map shows a spatial distribution of 
of potential infections (NPI) of MoT summarized for all NPI indicating higher climate suitability for MoT develop-
selected countries, except for Bhutan, where model results ment over areas near the ocean over the southern fraction of 
showed non-suitable climate conditions for wheat blast the domain, such as in Bangladesh, some areas of West Ben-
development. The interannual trends (slope of the linear fit) gal and Bihar India, and in Myanmar. The model suggests 
in NPI were also evaluated, and their statistical significance that MoT can also establish over large areas of central India, 
was assessed using the non-parametrical Mann–Kendall test Myanmar, and China, though at lower NPI levels. Con-
(Kendall, 1955) at a confidence level of 0.05. In the second versely, other wheat-producing regions have air temperature 
step, the covariability between NPI and climate variables and humidity ranges that would not represent favorable con-
was assessed. This was performed by computing the Pear- ditions for MoT outbreaks, including most areas in Afghani-
son correlation coefficient between pairs of detrended time stan, Pakistan, and central China. Figure 2b shows the inter-
series of NPI and air temperature, relative humidity, and annual variability (standard deviation) of potential infections 
rainfall anomalies. Lastly, a composite analysis of anomalies in Asia, where a strong interannual variability is observed in 
of the above-presented climate variables was performed for areas of higher incidence (Bangladesh, Myanmar), but also 
the years of highest MoT incidence predicted by the model, a southward increase in potential infection risks in India. 
taking the upper quartile (75th percentile) of the NPI time Similarly, interannual variability of NPI shows a wider 
1 3
2242 International Journal of Biometeorology (2022) 66:2237–2249
Fig. 2  Maps of a mean and b 
interannual variability repre-
sented by the standard deviation 
(1980–2019) of the number of 
Magnaporthe oryzae patho-
type Triticum (MoT) potential 
infections (NPI) in Asia. Black 
dots represent grid cells with 
presence of wheat but where the 
climate appears to not be suit-
able for MoT outbreaks. P99th 
is the 99% percentile
range over the areas of higher MoT potential incidence (e.g., shown in Fig. 2a, although only a small fraction is statisti-
Bangladesh). The southward increasing pattern in India and cally significant according to the Mann–Kendall test. Posi-
Myanmar suggests that the pressure of the disease could tive trends are dominant in Bangladesh, central Myanmar, 
be much higher than the average conditions during more and over portions of the Indo-Gangetic Plains (IGP) of India. 
favorable years for its development, so long as sufficiently Decreasing trends are observed further south over warmer 
susceptible wheat cultivars are grown and alternative hosts areas of India and in Myanmar’s delta, where recent tem-
maintain inoculum outside the wheat-growing season. The perature trends may be above the maximum MoT develop-
distributions of total multi-year NPI cases and NPI normal- ment temperature in the model (IPCC 2021).
ized by the corresponding infected area and aggregated by 
country are presented in Fig. 3a and b, respectively. There is Seasonal climate anomalies and number 
a considerable variation in the mean and spread of the dis- of potential MoT infections
tribution of normalized NPI across countries. However, it is 
clear that Bangladesh is the country with the highest relative The relationship between NPI and seasonal anomalies 
potential MoT pressure associated with climate, followed by (from wheat heading to maturity) of air temperature (T), 
India and Myanmar, which present similar disease risk sce- relative humidity (RH), and total rainfall (R) is described in 
narios, and then in Nepal, China, Pakistan, and Afghanistan. Fig. 5a–c, which show the correlation coefficient calculated 
The long-term interannual trends in seasonal total NPI are between NPI and these variables. The correlation map of 
displayed in Fig. 4, including their statistical significance. NPI and T shows that most of the grid points with suitable 
Our model outputs suggest generally increasing trends in climate conditions for MoT described in Fig. 2 do not present 
NPI that concentrate over areas of higher MoT pressure statistically significant correlations. This is likely due to the 
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2243
suitability considered in the model, indicative of reduced 
infection risk potential. On the other hand, the correlation 
between NPI and RH (Fig. 5b) is much more apparent than 
with temperature; this is indicative of the importance of 
RH conditions for the potential development of the disease. 
In this case, strong positive correlations are observed over 
Bangladesh, Myanmar, and some areas of India, which cor-
respond to those of higher MoT pressure (Figs. 2 and 3). 
A very weak correlation between NPI and total rainfall is 
observed in Fig. 5c for the whole geographical domain. 
Since the calculation period falls in general within the dry 
season (Fig. S6), our models suggest that precipitation may 
not be a significant determining factor of the incidence of 
MoT at the scales of the present work, as other factors asso-
ciated with atmospheric water vapor transport might be more 
relevant (Ahmed et al. 2020). However, the 2016 outbreak of 
MoT in Bangladesh has been associated with strong storm 
events during the dry season (Singh et al. 2021), which is 
not captured by a correlation-based analysis.
The maps of mean composite anomalies of T, RH, and R 
calculated for the years of the highest infection events, repre-
sented by the upper quartile of NPI, are displayed in Fig. 5d–f. 
Figure 5d shows an area of negative temperature anomalies 
in northern Bangladesh that was already observed with high 
interannual correlations in Fig. 5a. In general, both India and 
Myanmar present a pattern of anomalies that are not very 
clear, although they in general follow what is observed in 
terms of correlations (Fig. 5a). Conversely, RH shows a spa-
tial distribution (Fig. 5e) that appears to be consistent with 
the interannual correlations (Fig. 5b), where anomalously 
high seasonal NPI is associated with positive anomalies in 
RH, appearing again as a variable with high discriminatory 
power for modeled NPI. Figure 5f shows that high incidence 
of MoT is likely to be associated with negative precipitation 
anomalies. The latter suggests that, despite the null correlation 
between both variables, drier-than-normal winters may tend 
to be more favorable for MoT outbreaks.
Fig. 3  a Histogram of multi-year number of potential MoT infections Discussion
(NPI) in Asia; P25% and P75% are the corresponding percentiles. b 
Boxplots of interannual distribution of NPI. In b, the red central mark Climate suitability for MoT in Asia
shows the median and the box edges are the 25th and 75th percen-
tiles; dashed lines extend to the most extreme values not considered 
outliers, and outliers are plotted individually (× signs) Assuming the presence of inoculum and susceptible culti-
vars, the primary goal of this study was to provide a general 
and objective overview of the background climate conditions 
scaling used in the temperature response function (Eq. 1), for the development of MoT over a region where wheat cul-
which implies a non-linear relationship between temperature tivation is important for food security (Yonar et al. 2021), 
and NPI. However, a small area in northern Bangladesh with and whose agricultural landscape is described as being 
relatively high MoT pressure (Fig. 2) has significant negative highly exposed to the shocks associated with weather and 
correlations. This area exhibits high temperatures during the climate variability (Amarnath et al. 2017). At the time of 
wheat heading period (Fig. S6), which may imply a higher writing, MoT has only been officially reported in Bangladesh 
frequency of hours with the temperature above the range of (Malaker et al. 2016), though unofficial reports of the disease 
1 3
2244 International Journal of Biometeorology (2022) 66:2237–2249
Fig. 4  Map of interannual 
trends (1980–2019) in NPI over 
Asia. Black dots represent areas 
where linear trends are statisti-
cally significant (α = 0.05) 
according to the Mann–Kendall 
test
have also been published in the popular media in eastern On the other hand, we also observed increasing NPI risks in 
India, prompting the temporary banning of wheat cultivation northwestern India. This result could potentially be associ-
in some areas (cf. Islam et al. 2020a, b); our modeling efforts ated with irrigation, which is intensively applied to wheat 
also suggest high disease pressure risks, though not always on over 80% of the land area devoted to rice–wheat rotations 
with significant and positive interannual trends, potentially in northwestern India (Hussain et al. 2003; Jain et al. 2017; 
backing the consistent but spatially variable incidence and Ram et al. 2013;), which could contribute to land surface 
severity of MoT observed in this country between 2017 and cooling during the pre- and post-monsoon period (Mishra 
present (CSISA 2021). The reasons for the variable nature et al. 2020). Intensified use of irrigation on the other hand 
of infections observed in Bangladesh and South Asia remain has also increased evaporation, increasing actual water vapor 
unclear; though recognizing weather conditions as a major pressure (Tuinenburg et al. 2014), which can determine an 
driver of fungal disease outbreaks (Bregaglio et al. 2012; increase in relative humidity creating conditions that are 
Juroszek et al. 2019), improved knowledge regarding the more suitable for MoT development (Bregaglio et al. 2012). 
environmental suitability for the establishment of MoT can Additionally, our scenarios suggest that high incidence of 
aid in anticipating the development of disease management MoT is likely to be associated with negative precipitation 
strategies. These include but are not limited to the deploy- anomalies. The latter could be associated with the regulatory 
ment of new resistant varieties, cultural control methods, and effect of rainfall on air temperature, which affects relative 
the use of early warning systems in wheat regions where the humidity during a period of the year where precipitation 
disease is a potential threat for food security. events are sporadic. However, further analysis is necessary 
The results highlight the importance of spatial variabil- to validate this hypothesis.
ity in the climate suitability for the establishment of MoT On the other hand, results suggest that wheat-producing 
in Asia, with a higher potential observed in Bangladesh, regions with low temperature and humidity in Afghani-
Myanmar, and some areas of India, where low elevation, sea stan, Pakistan, or some areas of India are unlikely to be 
proximity, or regional low-level circulation can favor factors at significant risk for MoT outbreaks, as climatic regime 
such as atmospheric water transport (Ahmed et al. 2020). appears to be out of the range for the disease development 
Using a methodologically similar approach, a similar spa- (Magarey et al. 2005; Cruz et al. 2016). According to the 
tial pattern in the suitability of rice leaf blast (Magnaporthe observed relationship between interannual variability in 
oryzae pathotype Oryzae) driven by summer weather over NPI and the selected climate variables, a clear association 
North India was described by Viswanath et al. (2017) for between anomalies of RH and NPI was observed, which 
India. At the same time, regions that appear to have higher is explained by the structure of the infection model. This 
potential risks for infection in our model are also associated observation confirms those of Kim and Choi (2020) and 
with higher interannual variability. This appears to reflect Fernandes et al. 2017) that suggested that this variable 
in literature on MoT from South America (Fernandes et al. could be potentially used for the development of seasonal 
2017) and observations in Bangladesh (CSISA 2021) that the early warning systems. Indeed, recent efforts to develop 
disease is irregularly periodic, increasing in incidence and weather-based early warning systems in Bangladesh and 
severity only during years of higher favorable conditions. Brazil (e.g., Fernandes et al. 2021; http://b eatth eblas tews. 
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2245
Fig. 5  a–c Maps of local Pearson correlation coefficient between a mean air temperature, b relative humidity, and c total rainfall asso-
number of potential infections (NPI) and a mean temperature (T), b ciated with the upper quartile of NPI. Only grid cells exceeding the 
relative humidity (RH), and c total rainfall (R). Only significant cor- 95% confidence interval are displayed in d–f 
relations at the 10% level are displayed. d–f Composites of seasonal 
net/) rely largely on humidity and temperature as driv- Limitations and uncertainties
ing variables. Nevertheless, the association between NPI 
and climate anomalies seems to be clearer when using The approach used in this study considers the combination 
a composite approach, which could open the possibility of multiple sources of secondary information and modeling 
of generating probabilistic seasonal forecasts of favorable (cropping calendars, phenology, potential infections, etc.). 
conditions for MoT outbreaks. The latter could be further This in turn implies multiple sources of uncertainty and 
explored using indices from large-scale drivers (El Niño/ limitations that should be considered in order to improve 
La Niña) and suitable lead times. In addition, although the understanding of the conditions conducive to the devel-
most of the area of the geographical domain studied did opment of crop diseases, including future projections in cli-
not exhibit statistically significant trends, areas that exhibit mate. For instance, the generic infection model considers the 
show positive trends in NPI that could increase in response moment when favorable weather conditions for the devel-
to projected climate change scenarios, which should be opment of MoT are fulfilled to declare an outbreak. How-
addressed in future studies. ever, other complex disease-host interaction processes could 
1 3
2246 International Journal of Biometeorology (2022) 66:2237–2249
determine the successful establishment of the disease, which products of key phenological stages. Similarly, although 
are not considered in the model, which also assumes that the SPAM dataset (Wood-Sichra et al. 2016; International 
inoculant is present in the environments being studied (e.g., Food Policy Research Institute (IFPRI) 2019) is widely 
Bregaglio et al. 2012, 2021). Additionally, other relevant used (e.g., Joglekar et al. 2019; Yu et al. 2020), it is par-
variables that have been considered in other similar works tially based on administrative report data for 2005 and has 
could be included. For instance, the “wash-off” effect of not been thoroughly ground-truthed and as such there may 
spores from spikes by rainfall above a specific intensity has be spatial over- and under-estimation of wheat cultivation 
been considered by Fernandes et al. (2017), which can be area. For example, Myanmar has a declining trend and less 
relevant over more where significant rains occur during the than 100,000 ha of wheat (FAOSTAT 2021; USDA PS&D 
heading wheat period. Another source of uncertainty are the 2022), while the SPAM product suggests a cultivation area 
model parameters, and specifically the D50 parameter used of around 85,500 ha, and a more southern distribution 
in Eq. 3, which was extracted from similar fungal species of cultivation than other sources suggest (USDA PS&D 
(Puccinia sp. and Bipolaris sp.) in rice and wheat. Although 2022). Future researchers may therefore consider making 
this can lead to errors, Bregaglio et al. (2012) found that use of satellite-derived estimates for wheat phenology to 
results are not very sensitive to the values of D50 for other complement this data source, for example, using methods 
fungal diseases. We however conducted a simple sensitiv- described by Jain et al. (2017). Yet despite these potential 
ity analysis, S7, that was performed using a set of values of inconsistencies, our model outputs still provide a useful 
D50, from “sensitive” to “insensitive” to dry interruptions indication of the potential for MoT infection risks, and can 
according to Magarey et al. (2005), presented in Fig. S7, therefore be used to help in crop planning and zoning, in 
which suggests that the calculated NPI are not very sensitive addition to integrated pest management efforts, although 
to variations in D50 values. In any case, a global sensitiv- care should be exercised when interpreting our results.
ity and uncertainty quantification analysis would provide a 
better understanding of the model structure and sensitivity 
to parameters and threshold values (Bregaglio et al. 2012), 
which is, however, out of the scope of this work. Conclusions
In addition, and in spite of using phenology dates that 
are comparable to other works (e.g., Liu et al. 2020), the The sudden, unexpected arrival of wheat blast disease in 
use of fixed planting and heading/maturity dates may rep- Bangladesh in 2016 underscores the risk associated with 
resent a source of error during, for instance, anomalously this disease. Although formally reported in Bangladesh at 
warm/cold years, in which the phenological stages can the time of writing, there is a lack of clarity on the potential 
be accelerated/delayed. The latter may represent a limi- distribution of the MoT species and its effect on wheat cul-
tation when developing early warning systems based on tivation throughout the Asian continent. Our results suggest 
seasonal climatic forecasts, which currently provide infor- a differential suitability for the development of MoT—and 
mation on a monthly or longer scale. Moreover, we relied a large interannual variation in some key wheat-producing 
on a single, though comprehensive data source for plant- areas—across Asia. The contrasting potential risk of MoT 
ing dates from the interpolated Crop Calendar Dataset of between Bangladesh, Myanmar, and some states within 
Sacks et al. (2010). Although widely used, this dataset India, with infection events averaging up to 15 during the 
may have inaccuracies with observed planting dates, which wheat spike, and limited risks in Afghanistan and Pakistan, 
can in turn affect phenological development. For exam- and in central China, could allow focusing efforts to increase 
ple, this dataset shows quite late wheat sowing dates into the resilience and preparation of farmers for potential 
December in the north western IGP, and specifically in future biotic shocks. Importantly, our results also highlight 
the Indian states of Haryana and Punjab (Fig. S4). These a stronger association between relative humidity and MoT 
locations however tend to be associated with earlier plant- infection than with temperature regime. Accordingly, future 
ing than in the eastern IGP (Lobell et al. 2013; Jain et al. improvements should further investigate if and how rela-
2017). The reasons for the lack of congruence between tive humidity can be used to simplify data requirements and 
the Sacks et al. (2010) dataset and observations are not modeling efforts. New research should also focus on includ-
clear, although future studies should query the relationship ing more complex pathogen-plant interaction processes, 
between crop establishment dates and disease incidence, dynamic wheat phenology, source-sink relationships, and 
in an effort to identify if and how MoT risks could be wind dispersal patterns, higher resolution climate forcing for 
mitigated through manipulation of sowing dates. Remote historical and future assessments. Although still preliminary 
sensing–based regional sowing date estimations based in nature, our results nonetheless may aid in the develop-
on the seasonality of satellite time series such TIMESAT ment or refinement of early warning systems and agricultural 
(Jönsson and Eklundh, 2004) could help to generate global climate services associated with MoT and similar diseases.
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2247
Supplementary Information The online version contains supplemen- Bregaglio S, Donatelli M, Confalonieri R, Orlandini S (2010) An inte-
tary material available at https://d oi.o rg/1 0.1 007/s 00484-0 22-0 2352-9. grated evaluation of thirteen modelling solutions for the genera-
tion of hourly values of air relative humidity. Theor Appl Climatol 
Acknowledgements The authors gratefully acknowledge the ECMWF 102:329–438
for providing the ERA5 data. CM acknowledges Ernesto Girón for Bregaglio S, Cappelli G, Donatelli M (2012) Evaluating the suitability 
his support on CROPSIM-CERES simulations. The valuable com- of a generic fungal infection model for pest risk assessment stud-
ments and suggestions from two anonymous reviewers are deeply ies. Ecol Model 247:58–63
acknowledged. Bregaglio S, Donatelli M, Confalonieri R (2013) Fungal infections of 
rice, wheat, and grape in Europe in 2030–2050. Agron Sustain 
Dev 33:767–776
Funding Funding for this research was supplied by USAID and the Bill 
and Melinda Gates Foundation (BMGF) through the Cereal Systems Bregaglio S, Donatelli M (2015) A set of software components for 
Initiative for South Asia (CSISA), and the Climate Services for Resil- the simulation of plant airborne diseases. Environ Modell Soft 
ient Development (CSRD) in South Asia project, supported by USAID. 72:426–444
Additional support was provided by the CGIAR Research Program on Bregaglio S, Willocquet L, Kersebaum KC, Ferrise R, Stella T, Ferreira 
Climate Change, Agriculture and Food Security (CCAFS; https://c cafs. TB, Pavan W, Asseng S, Savary S (2021) Comparing process-
cgiar. org), by the CGIAR Regional Integrated Initiative Transforming based wheat growth models in their simulation of yield losses 
Agrifood Systems in South Asia, or TAFSSA (https:// www. cgiar. org/ caused by plant diseases. Field Crop Res 265:108108
initia tive/2 0-t ransf ormin g-a grifo od-s ystem s-i n-s outh-a sia-t afssa/), and Ceresini PC, Castroagudín VL, Rodrigues FA, Rios JA, Eduardo 
by the CGIAR Foresight and Metrics to Accelerate Inclusive and Sus- Aucique-Pérez C, Moreira SI, Alves E, Croll D, Nunes Macie JL 
tainable Agrifood System Transformation initiative (https://w ww.c giar. (2018) Wheat blast: past, present, and future. Annu Rev Phyto-
org/ initi ative/ 24- fores ight- and- metri cs- to- accel erate- inclu sive- and- pathol 56(427):456
sustai nable-a grifo od-s ystem-t ransf ormat ion/). The contents expressed Copernicus Climate Change Service (C3S) (2019) Data Stream 2: 
herein are those of the author(s) and do not necessarily reflect the AgERA5 historic and near real time forcing data. Product User 
views of USAID, BMGF, CGIAR, or the US government, and shall not Guide and Specification. Copernicus Climate Change Service Cli-
be used for advertising or product endorsement purposes. mate Data Store (CDS). Online: http://d atast ore.c opern icus-c lima 
te.e u/d ocume nts/s is-g lobal-a gricu lture/C 3S422 Lot1.W EnR.D S2_ 
Produ ctUse rGuid eSpec ifica tion_ v2.2. pdf. Accessed 10 Jan 2022
Open Access This article is licensed under a Creative Commons Attri- Cruz CD, Magarey RD, Christie DN, Fowler GA, Fernandes JM, 
bution 4.0 International License, which permits use, sharing, adapta- Bockus WW, Valent V, Stack J (2016) Climate suitability for 
tion, distribution and reproduction in any medium or format, as long Magnaporthe oryzae Triticum pathotype in the United States. 
as you give appropriate credit to the original author(s) and the source, Plant Dis 100:1979–1987
provide a link to the Creative Commons licence, and indicate if changes Cruz CD, Valent B (2017) Wheat blast disease: danger on the move. 
were made. The images or other third party material in this article are Trop Plant Pathol 42:210–222
included in the article’s Creative Commons licence, unless indicated CSISA (2021) Cereal Systems Initiative for South Asia (CSISA) 
otherwise in a credit line to the material. If material is not included in Annual Report. October 2020 – September 2021. International 
the article’s Creative Commons licence and your intended use is not Maize and Wheat Improvement Center (CIMMYT). Dhaka, 
permitted by statutory regulation or exceeds the permitted use, you will Bangladesh. Available online: https://c sisa.o rg/w p-c onten t/u ploa 
need to obtain permission directly from the copyright holder. To view a ds/ sites/2/ 2021/ 11/ 211031- CSISA- PIII- APR- Oct20- to- Sept21_ 
copy of this licence, visit http://c reati vecom mons.o rg/l icens es/b y/4.0 /. web. pdf. Accessed: 2 January 2022.
Curtis BC (2002) Wheat in the world. In Bread wheat: improvement 
and production. Curtis, B. C.; Rajaram, S.; Gómez Macpherson, 
H. (eds) 2002 xi + 554 Food and Agriculture Organization, Vialle 
delle Terme di Caracalla, 00100 Rome, Italy
References Donatelli M, Magarey RD, Bregaglio S, Willocquet L, Whish JPM, 
Savary S (2017) Modelling the impacts of pests and diseases on 
Ahmed T, Jin HG, Baik JJ (2020) Spatiotemporal variations of pre- agricultural systems. Agric Syst 155:213–224
cipitation in Bangladesh revealed by nationwide rain gauge data. Duveiller E, Hodson D, Sonder K, Tiedermann A (2011) An interna-
Asia-Pacific J Atmospheric Sci 56:593–602 tional perspective on wheat blast. Phytopathol 101:S220
Allen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotran- Duveiller E, He X, Singh PK (2016) Wheat blast: an emerging disease 
spiration - guidelines for computing crop water requirements. in South America potentially threatening wheat production. World 
FAO Irrigation and drainage paper 56. Food and Agriculture Wheat Book, Volume 3. A History of Wheat. Bonjean A. and 
Organization, Rome van Ginkel M. (Eds.) Pages 1107–1122. Lavoisier, Paris, France.
Amarnath G, Alahacoon N, Smakhtin V, Aggarwal P (2017). Mapping FAOSTAT (2021) FAOSTAT - Food and Agriculture Organization of 
multiple climate-related hazards in South Asia. In IWMI Research the United Nations. Available online: http://w ww.f ao.o rg/f aosta t/ 
Report 170, 1–46 Colombo Sri Lanka: International Water Man- en/ (Accessed 2 January 2022)
agement Institute (IWMI) Fernandes JM, Del Ponte EM, Ascari JP, Krupnik TJ, Pavan W, Vargas 
Barea G, Toledo J (1996) Identificación y zonificación de Pyricularia F, Berton T (2021) Towards an early warning system for wheat 
o brusone (Pyricularia oryzae) en el cutivo de trigo en el departa- blast: epidemiological basis and model development. In: Oli-
mento de Santa Cruz. Centro de Investigación Agrícola Tropical. ver R (ed) Advances in understanding biology/epidemiology of 
Informe Tecnico. Proyecto de Investigacion Trigo. Santa Cruz de blast infection of cereals. Burleigh Dodds Science Publishing, 
la Sierra, Bolivia, pp 76–86 Cambridge
Bebber DP, Holmes T, Gurr SJ (2014) The global spread of crop pests Fernandes JMC, Nicolau M, Pavan W, Hölbig CA, Karrei M, de Vargas 
and pathogens. Global Ecol Biogeogr 23:1398–1407 F, Bo Boeira Bavaresco JL, Lazzaretti AT, Tsukahara RY (2017) 
Bebber DP, Castillo AD, Gurr SJ (2017) Modelling coffee leaf rust A weather-based model for predicting early season inoculum 
risk in Colombia with climate reanalysis data. Phil Trans R Soc build-up and spike infection by the wheat blast pathogen. Trop 
b 371:20150458 Plant Pathol 42:230–237
1 3
2248 International Journal of Biometeorology (2022) 66:2237–2249
Figueroa M, Hammond-Kosack KE, Solomon PS (2018) A review socio-economic, and environmental challenges and opportu-
of wheat diseases - a field perspective. Mol Plant Pathol nities in Nepal’s cereal-based farming systems. Adv Agron 
19:1523–1536 170:155–287
Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw Launay M, Caubel J, Bourgeois G, Huard F, Garcia de Cortazar-
SL, Gurr SJ (2012) Emerging fungal threats to animal, plant and Atauri I, Bancal MO, Brisson N (2014) Climatic indicators for 
ecosystem health. Nature 484:186–194 crop infection risk: application to climate change impacts on 
Goulart ACP, Paiva FA (1992) Incidence of (Pyricularia oryzae) in five major foliar fungal diseases in Northern France. Agric Eco-
different wheat cultivars under field conditions. Fitopatol Bras sys Environ 197:147–158
17:321–325 Liu H, Pequeno DNL, Hernández-Ochoa IM, Krupnik TJ, Sonder K, 
Hersbach H, Bell B, Berrisford P et al (2020) The ERA5 global rea- Xiong W, Xua Y (2020) A consistent calibration across three 
nalysis. Q J R Meteorol Soc. 146:1999–2049 wheat models to simulate wheat yield and phenology in China. 
Hossain A, Motaleb KA, Farhad M, Barma NC (2019) Mitigating the Ecol Model 430:109132
twin problems of malnutrition and wheat blast by one wheat vari- Lobell DB, Ortiz-Monasterio JI, Sibley AM, Sohu VS (2013) Satel-
ety, “BARI Gom 33”, in Bangladesh. Acta Agrobot 72:1775 lite detection of earlier wheat sowing in India and implications 
Hussain I, Sakthivadivel R, Amarasinghe U, Mudasser Md, Molden for yield trends. Agric Syst 115:137–143
D (2003) Land and water productivity of wheat in the west- Magarey RD, Sutton TB, Thayer CL (2005) A simple generic infection 
ern Indo-Gangetic plains of India and Pakistan: a comparative model for foliar fungal plant pathogens. Phytopathol 95:92–100
Analysis Research Report 65. International water Management Malaker PK, Barma NCD, Tiwari TP, Collis WJ, Duveiller E, Singh 
Institute IWMI, Columbo Sri Lanka PK, Braun H-J (2016) First report of wheat blast caused by Mag-
Intergovernmental Panel on Climate Change (IPCC) (2021) Climate naporthe oryzae pathotype Triticum in Bangladesh. Plant Dis 
change 2021: the physical science basis. Contribution of work- 100:2330
ing group I to the sixth assessment report. Cambridge Univer- Martínez SI, Sanabria A, Fleitas MC, Consolo VF, Perelló A (2019) 
sity Press, Cambridge Wheat blast: aggressiveness of isolates of Pyricularia oryzae and 
International Food Policy Research Institute (IFPRI) (2019) Global effect on grain quality. J King Saud Univ Sci 31:150–157
spatially-disaggregated crop production statistics data for 2010 Meyer M, Cox JA, Hitchings MDT, Burgin L, Hort MC, Hodson DP, 
version 1.0. Harvard Dataverse, V2. https:// doi. org/ 10. 7910/ Gilligan CA (2017) Quantifying airborne dispersal routes of path-
DVN/ PRFF8V ogens over continents to safeguard global wheat supply. Nature 
Islam MT, Gupta DR, Hossain A, Roy KK, He X, Kabir MR, Singh Plants 3:780–786
PK, Khan MAR, Rahman M, Wang GL (2020) Wheat blast: a Mishra V, Ambika AK, Asoka A, Aadhar S, Buzan J, Kumar R, Huber 
new threat to food security. Phytopath Res 2:28 M (2020) Moist heat stress extremes in India enhanced by irriga-
Igarashi S, Utiamada CM, Igarashi LC, Kazuma AH, Lopes RS tion. Nat Geosci 13:722–728
(1986) Pyricularia in wheat.1. Occurrence of Pyricularia sp. in Morris ML, Dubin HJ, Pokhrel T (1994) Returns to wheat breeding 
Paraná State. Fitopatol Bras 11:351–352 research in Nepal. Agric Econ 10:269–282
Islam MT, Croll D, Gladieux P, Soanes DM, Persoons A, Bhattacha- Motaleb KA, Singh PK, Sonder K, Kruseman G, Tiwari TP, Barma 
rjee P et al (2016) Emergence of wheat blast in Bangladesh was NCD, Malaker PK, Braun HJ, Erenstein O (2018a) Threat of 
caused by a South American lineage of Magnaporthe oryzae. wheat blast to South Asia’s food security: an ex-ante analysis. 
BMC Biol 14:84 PLoS ONE 13:e0197555
Islam MT, Gupta DR, Hossain A, Roy KK, He X, Kabir MR, Singh Motaleb KA, Rahut DB, Kruseman G, Erenstein O (2018b) Wheat 
PK, Khan MAR, Rahman M, Wang GL (2020b) Wheat blast: a production and consumption dynamics in an Asian rice economy 
new threat to food security. Phytopathol Res 2:28 - the Bangladesh case. Eur J Dev Res 30:252–275
Jain M, Singh Balwinder, Srivastava AAK, Malik RK, McDonald Pequeno DNL, Hernández-Ochoa IM, Reynolds M, Sonder K, 
A, Lobell DB (2017) Using satellite data to identify the causes MoleroMilan A, Robertson RD, Lopes MS, Xiong W, Kropff 
of and potential solutions for yield gaps in India’s Wheat Belt. M, Asseng S (2021) Climate impact and adaptation to heat and 
Environ Res Lettr 12:094011 drought stress of regional and global wheat production. Environ 
Joglekar AKB, Wood-Sichra U, Pardey PG (2019) Pixelating crop Res Lett 16:054070
production: consequences of methodological choices. PLoS Perelló A, Martinez I, Molina M (2015) First report of virulence and 
ONE 14(2):e0212281 effects of Magnaporthe oryzae isolates causing wheat blast in 
Jones JW, Hoogenboom G, Porter C, Boote K, Batchelor W, Hunt L, Argentina. Plant Dis 99:1177
Wilkens P, Singh U, Gijsman A, Ritchie J (2003) The DSSAT Ram H, Dadhwal V, Vashist KK, Kaur H (2013) Grain yield and water 
cropping system model. Eur J Agron 18:235–265 use efficiency of wheat (Triticum aestivum L.) in relation to irri-
Jönsson P, Eklundh L (2004) TIMESAT - a program for analyzing gation levels and rice straw mulching in North West India. Agric 
time-series of satellite sensor data. Comput Geosci 30:833–845 Water Manage 128:92–101
Juroszek P, Racca P, Link S, Farhumand J, Kleinhenz B (2019) Reynolds MP, coauthors (2017) Improving global integration of crop 
Overview on the review articles published during the past 30 research. Science 357:359–360
years relating to the potential climate change effects on plant Sacks WJ, Deryng D, Foley JA, Ramankutty N (2010) Crop plant-
pathogens and crop disease risks. Plant Pathol 69(2):179–193 ing dates: an analysis of global patterns. Global Ecol Biogeogr 
Kendall MG (1955) Rank correlation methods. Griffin, London 19:607–620
Kim K-H, Choi ED (2020) Retrospective study on the seasonal fore- Singh PK, Gahtyari NC, Roy C, Roy KK, He X, Tembo B, Xu K, 
cast-based disease intervention of the wheat blast outbreaks in Juliana P, Sonder K, Kabir MR, Chawade A (2021) Wheat blast: 
Bangladesh. Front Plant Sci 11:570381 a disease spreading by intercontinental jumps and its management 
Koo J, Dimes J (2010) HC27 soil profile database. International Food strategies. Front Plant Sci 12:710707
Policy Research Institute, Washington D.C. http:// hdl. handle. Tembo B, Mulenga RM, Sichilima S, M’siska KK, Mwale M, Chikoti 
net/ 1902.1/ 20299 PC, Singh PK, He X, Pedley KF, Peterson GL, Singh RP, Braun 
Krupnik T, Timsina J, Devkota KP, Tripathi B, Karki TB, Urfels HJ (2020) Detection and characterization of fungus (Magna-
A, Gaihre YK, Choudhary D, Beshir AR, Pandey VP, Gar- porthe oryzae pathotype Triticum) causing wheat blast disease 
taula H, Brown B, Shahirn S, Ghimire YN (2021) Agronomic, 
1 3
International Journal of Biometeorology (2022) 66:2237–2249 2249
on rain-fed grown wheat (Triticum aestivum L) in Zambia. PLoS HarvestChoice Working Paper. HarvestChoice, International Food 
ONE 15(9):e0238724 Policy Research Institute (IFPRI), Washington, D.C.
Tshewang S, Park RF, Chauhan BS, Joshi AK (2017) Challenges and Yan W, Hun LA (1999) An equation modeling the temperature 
prospects of wheat production in Bhutan: a review. Exp Agr response of plant growth and development using only the cardinal 
54:428–442 temperatures. Ann Bot 84:607–614
Tuinenburg OA, Hutjes RWA, Stacke T, Wiltshire A, LucasPicher P Yesmin N, Jenny F, Abdullah HM, Hossain MM, Kader MA, Solo-
(2014) Effects of irrigation in India on the atmospheric water mon PS, Bhuiyan MAHB (2020) A review on South Asian wheat 
budget. J Hydrometeorol 15:1028–5100 blast: the present status and future perspective. Plant Pathol 
United States Department of Agriculture (USDA) (2019) Burma - 69:1618–1629
union of grain and feed annual report. U.S. Department of Agri- You L, Wood S (2006) An entropy approach to spatial disaggregation 
culture, Foreign Agriculture Service, Washington, D.C. of agricultural production. Agric Syst 90:329–347
United States Department of Agriculture, Production, Supply, and Dis- Yonar A, Yonar H, Mishra P, Kumari B, Abotaleb M, Badr A (2021) 
tribution (USDA PS&D) (2022). https:// ipad. fas. usda. gov/ count Modeling and forecasting of wheat of South Asian region coun-
rysumm ary/d efaul t.a spx?i d=B M&c rop=W heat. Accessed 1 Aug tries and role in food security. Adv Comp Int 1:11
2022 Yu Q, You L, Wood-Sichra U, Ru Y, Joglekar AKB, Fritz S, Xiong W, 
Viswanath K, Sinha P, Naresh Kumar S, Sharma T, Saxena S, Panjwani Lu M, Wu W, Yang P (2020) A cultivated planet in 2010 – part 2: 
S, Pathak H, Shukla SM (2017) Simulation of leaf blast infection the global gridded agricultural-production maps. Earth Syst Sci 
in tropical rice agro-ecology under climate change scenario. Clim Data 12(4):3545–3572
Change 142:155–167 Zadoks JC, Chang TT, Konzak CF (1974) A decimal code for the 
Wood-Sichra U, Joglekar AB, You L (2016) Spatial production growth stages of cereals. Weed Res 14:415–421
allocation model (SPAM) 2005: technical documentation. 
1 3