Agricultural Systems 177 (2020) 102697
Contents lists available at ScienceDirect
Agricultural Systems
journal homepage: www.elsevier.com/locate/agsy
Maize yield in Mexico under climate change
a,b,
c
d
e
Carolina Ureta ⁎, Edgar J. González , Alejandro Espinosa , Alejandro Trueba ,
Alma Piñeyro-Nelsonf,g, Elena R. Álvarez-Buyllaa,g,⁎
T
a
Departamento de Ecología Funcional, Instituto de Ecología, Universidad Nacional Autónoma de México, Mexico City, Mexico
Departamento de Ciencias Atmosféricas, Centro de Ciencias de la Atmósfera, Universidad Nacional Autónoma de México, Mexico City, Mexico
Departamento de Ecología y Recursos Naturales, Facultad de Ciencias, Universidad Nacional Autónoma de México, Mexico City, Mexico
d
Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias, Campo Experimental Valle de México (INIFAP, CEVAMEX), Km 13.5 Carretera Los Reyes –
Texcoco, C.P. 56250, Coatlinchan, Texcoco, Estad de México, Mexico
e
Dirección General de Educación Tecnológica Agropecuaria-SEP, Mexico City, Mexico
f
Departamento de Producción Agrícola y Animal, Universidad Autónoma Metropolitana-Xochimilco, Mexico City, Mexico
g
Centro de Ciencias de la Complejidad, Universidad Nacional Autónoma de México, Mexico City, Mexico
b
c
A R T I C LE I N FO
A B S T R A C T
Keywords:
Climate change
Crop yields
Maize
Irrigation
Rainfed
Understanding the effects of climate change on maize yield in Mexico is important from both a national and
international perspective. Maize is Mexico's staple food crop, thus, decrements in national production would
strongly compromise food security in the country.
Internationally, maize is the most important grain crop in terms of human consumption and the conservation
in situ of its germplasm should be a global priority. Mexico harbors half of the known genetic diversity for this
crop in the American continent, which is instrumental for future genetic improvement efforts that could generate
new, environmentally resilient varieties. In this study, we analyze the link between maize yield and several
climate variables in rainfed and irrigated crop areas in Mexico to project yield variations under future climate
change scenarios. We used municipality-level data for seven states that account for ∼65% of the annual maize
production in Mexico and cover an important amount of national climatic variability. We used public data
published by the Mexican government on yield and climate from 2003 to 2015 and built linear models to assess
the impact of climate on maize yield. We considered the municipality to be a random effect and accounted for
potential autocorrelation in the 13-year time series. We also evaluated how many municipalities reached their
states' breakeven point in order to project the geographic areas that will earn higher profits due to increased
yields. Our results showed that the municipality had a significant effect on yield, and consequently, our results
could not be extrapolated to other geographic areas in the country. We found temperature to be the most
influential factor on yield under rainfed conditions, while precipitation was the most influential factor for irrigated crops. Like earlier studies at a global scale, we found higher yield stability for irrigated fields than for
rainfed fields when considering different climate change scenarios. Our projections indicate that yields from
rainfed fields will be reduced significantly under future scenarios. We argue that average yield data in rainfed
fields does not include data on the diversity of native maize varieties or their potentially different responses to
changes in the environment. Finally, under current conditions, there are by far more municipalities reaching
their breakeven point in rainfed fields than in irrigated fields, suggesting that higher yields do not necessarily
translate into greater profits for farmers because costs can also increase depending on the type of agriculture
practiced.
1. Introduction
Climate change has become increasingly recognized as an important
threat to agriculture (IPCC, 2014). Maize (Zea mays L.) is one of the
most important crops in terms of production volume for direct and
⁎
indirect human consumption around the world (Ranum et al., 2014),
and negative impacts on its yield due to temperature increase and
changes in precipitation patterns have already been detected (Cakir,
2004; Cicchino et al., 2010; Ray et al., 2015). In Mexico, maize's center
of origin and diversification (domesticated ∼8000 years ago; Matsuoka
Corresponding authors at: Ciudad Universitaria, 04510, Instituto de Ecología, Mexico City, Mexico.
E-mail addresses: carolinaus@atmosfera.unam.mx (C. Ureta), eabuylla@gmail.com (E.R. Álvarez-Buylla).
https://doi.org/10.1016/j.agsy.2019.102697
Received 15 December 2018; Received in revised form 12 September 2019; Accepted 16 September 2019
0308-521X/ © 2019 Elsevier Ltd. All rights reserved.
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
on yield.
Two approaches have been used to explicitly evaluate possible impacts of climate change on maize production in Mexico (Conde et al.,
1997; Ureta et al., 2016; Eakin et al., 2018): a CERES-maize approach
and an ecological niche modeling approach. Conde et al. (1997) applied
the former by using seven locations to create CERES-maize models
(Jones and Kiniry, 1986) in rainfed fields only. This kind of modeling is
site-specific and does not allow extrapolations, and therefore cannot be
used to project future yields beyond the field studied. They found that
six out of seven sites presented negative impacts on yield from climate
change, but they also recognized that their approach lacked some data
necessary for a thorough and accurate model (Conde et al., 1997).
The study performed by Ureta et al. (2016) used the ecological
niche-centroid theory (Maguire Jr, 1973; Martínez-Meyer et al., 2013)
to project yield changes under climate change for nine native races.
They considered it was important to recognize that the niche centroid
can be related to higher yield values when evaluating individual races
(Ureta et al., 2016). Their projections indicate that under climate
change, the evaluated races would have higher yields in geographic
regions different from where they are currently being sown.
Another recent study related seven climate variables with maize
yield in Mexico using data from 1980 to 2000 to calibrate their model
(Eakin et al., 2018). However, the model was created at a state scale,
which is quite a coarse measure considering the heterogeneity in maize
yield inside any one state. Furthermore, even when their study gives
some insight pertaining to fluctuations in yield under different climatic
conditions, they did not project potential impacts on yield under future
climate change scenarios.
Thus, to date, available studies evaluating climate change impacts
on maize yield in Mexico have been site or race-specific or focused only
on identifying whether a relationship between climate and yield had
existed during the last decades of the twentieth century. Consequently,
there is a need for a study focused on evaluating the impacts of climate
change on Mexican maize yield at a broader scale, additionally comparing irrigated and rainfed fields.
Yields from rainfed and irrigated fields are expected to be affected
by climate change; however, rainfed fields might be more susceptible
because of their dependence on precipitation (SAGARPA a, 2016). In
this study we analyzed the link between maize yield and climate variables for both rainfed and irrigated fields at a municipality level in
seven out of the 32 Mexican states. The states analyzed were selected
because they account for ∼65% of maize production in the country and
together encompass an important amount of climatic variability
(Trueba, 2012), representing five different Mexican regions proposed
by the Mexican ministry of agriculture (SADER, 2019). We also considered information about breakeven points (UNISEM, 2017) in irrigated and rainfed fields under current and future climate change conditions with the purpose of evaluating if higher yields directly translate
into greater profits (low input vs. high input agriculture; FIRA, 2007).
et al., 2002; Vigouroux et al., 2008; Kato et al., 2009), this grain is the
country's staple crop, thus, negative impacts on its yield could have
important consequences for local, regional and global food security.
Even when climate projections in Mexico are very heterogeneous in
terms of precipitation depending on the geographic area evaluated,
there is still a general tendency of temperature increase (Hijmans et al.,
2005); and most studies dealing with possible climate change effects on
Mexican maize have detected negative impacts (Conde et al., 1997,
Mercer and Perales, 2010, Monterroso et al., 2011, Ureta et al., 2012,
Ray et al., 2015, Ureta et al., 2016).
Despite corn being the staple crop in Mexico (representing over 50%
of the caloric intake for the poorest sectors of the population; Bourges,
2013) and that Mexico is one of the top ten producers in the world
(ASERCA, 2018), the country imports 34.12% of its maize; the second
highest maize import rate in the world (FIRA, 2016). Over the past
10 years, average maize yield in Mexico has been 2.9 t/ha (SIAP, 2018),
significantly lower than the world's average of 5.1 t/ha (2003–2014;
FAO, 2017). If yield increases, Mexico could be self-sufficient in its
maize requirements, but climate change could be an impediment
(Conde et al., 1997; Mercer and Perales, 2010; Monterroso et al., 2011;
Ureta et al., 2012; Ray et al., 2015; Ureta et al., 2016). A deficit in
maize production in Mexico could have additional negative implications, as it is culturally and culinarily important for a large part of its
population (Florescano, 2003; Fernández Suárez et al., 2013).
From a global perspective, Mexico represents one of the greatest
reservoirs of maize genetic diversity, harboring approximately 50% of
known genetic diversity for the American continent (Vigouroux et al.,
2008). In Mexico, most maize is cultivated on rainfed fields by farmers
on plots smaller than 5 ha; some of these farmers still grow maize in a
traditional agroecosystem called milpa in which several species are
being grown simultaneously (Mercer et al., 2012). These farmers are
the ones preserving maize diversity in situ and it should be made possible for them to attain enough yield from native maize varieties.
However, during the last three decades changes in temperature and
precipitation have made maize yield in Mexico one of the most variable
in the world (Ray et al., 2015), consequently, it is important to better
understand the relationship between maize yield and changes in different climate variables so that more accurate projections of yield are
available to policy makers.
Previous studies that have investigated the impacts of climate
change on maize in Mexico have focused on shifts in the suitability of
geographic areas by projecting the plants' climate threshold (Conde
et al., 1997; Monterroso et al., 2011) or have focused on the impact of
changes in climatic variables on specific native races (Ureta et al.,
2012) or varieties (Mercer and Perales, 2010). The concept of “races”
was coined in the 1940s and corresponds to a unit of analysis that has
been very useful to organize and study maize diversity in Mexico
(Anderson and Cutler, 1942). Races have been defined as “a group of
related individuals with enough characteristics in common to permit
their recognition as a group” (Anderson and Cutler, 1942); they share
phenological, morphological and to an extent, agroecological similarities. Thus, a race comprises several varieties or landraces. Ureta et al.
(2012) projected the potential distribution area of all Mexican races of
maize and their wild relatives that had enough information to be projected under different climate change scenarios through ecological
niche modeling to evaluate the potential impacts in their geographic
distribution. They found that the potential distribution area of most
races was negatively impacted by climate change. Work by Mercer and
Perales (2010) discussed the role of phenotypic plasticity, evolution and
gene flow in maintaining productivity in specific landraces (native
varieties). They found that when grown at mid-elevations, the productivity of highland varieties strongly decreased. The main focus of
Mercer and Perales (2010) was to analyze the plant's ecological strategies to reduce the impacts of climate change impacts, and both studies
(Ureta et al., 2012; Mercer and Perales, 2010) focused more on analyzing climate thresholds, distribution and adaptation possibilities than
2. Materials and methods
2.1. Modeling data
Yield data were obtained from the Mexican government database
(SIAP, 2018). These data were recorded for each municipality between
2003 and 2015 and divided into rainfed (4442 data points) and irrigated (2240) fields. We used yield data from the following states, encompassing around 400 municipalities: Sinaloa, Tamaulipas, Guanajuato, Jalisco, Michoacán, México and Chiapas (Fig. 1). Yield data used
here were originally recorded without reporting maize variety or race
(Appendix A) due to a lack of detailed information in the SIAP database;
still, these data are very useful to gain insight of future yield fluctuations in different areas of the country. Observed data showed that
rainfed areas with greater yields were found in Jalisco and Michoacán,
followed by Estado de México. Other states evaluated presented
2
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
Fig. 1. Regionalization of the country. The ministry of agriculture of the Mexican government regionalized the territory in order to apply different agricultural
policies depending on several characteristics of the geographic area including climatic features. In the map is possible to identify every state's name and the region to
which it belongs. The states that were taken into account for this study have been highlighted in the states lisrt.
very broad between states. In rainfed areas, Sinaloa has a mean annual
temperature above 25 °C, while Estado de México is below the 16 °C.
Similarly, in irrigated areas, differences among states in terms of precipitation and temperature were also documented. Irrigated areas in
Chiapas have average annual precipitations around 1400 mm, while the
lowest annual precipitation was found in Guanajuato with 614 mm.
This information suggests that maize is being grown under very different environmental conditions and that higher yields do not necessarily coincide with higher precipitations or warmer temperatures as
it could be expected (Khan et al., 2001; Sánchez et al., 2014).
average yields < 2 t/ha (Fig. 2). In irrigated fields, greater average
yields were found in Sinaloa, Guanajuato and Jalisco (Fig. 3). Interestingly, Jalisco has greater yields under both agricultural systems:
rainfed and irrigated.
We requested the climate data recorded by local meteorological
stations from the Mexican Meteorological Service (SMN, 2017) for
2003–2015. With this information, we created the 19 bioclimatic
variables most commonly used for ecological niche modeling to reflect
annual trends, limiting conditions and seasonality (Nix, 1986; Hijmans
et al., 2005). We discarded nine of these variables due to strong correlations (r ≥ 0.85) with the geographic area from our projections. Less
collinearity between predicted variables increases extrapolation performance (Steen et al., 2017). The 10 climate variables considered to
build the models were: annual mean temperature, maximum temperature of the warmest month, minimum temperature of the coldest
month, temperature annual range, mean temperature of the driest
quarter, mean temperature of the warmest quarter, mean temperature
of the coldest quarter, annual precipitation, precipitation of the driest
month and precipitation of the driest quarter.
In the evaluated states, the precipitation and temperature observed
during the thirteen years studied was very different depending on the
geographic area. For example, Chiapas has an average annual precipitation of 2069 mm in rainfed areas, while Sinaloa only reaches
739 mm. In terms of mean annual temperature, differences are also
2.2. Predicting yield under current and future climate conditions
We related (log-transformed) yield data to climate through linearmixed models (LMMs). We considered the municipality as a random
effect and accounted for potential autocorrelation in the 13-year time
series through an autoregressive process of order 1. In comparison to
regular linear models that only have fixed effects, LMMs allow controlling possible sources of variability coming from unmeasured characteristics. In our study, LMMs helped estimate the variability between
and within municipalities and integrate it into the calculation of model
parameters. We used municipalities as a random effect because each
municipality presents characteristics that makes it different from
others.
3
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
Fig. 2. Current and future yield under rainfed growing conditions. This figure shows five maps of Mexico (A, B, C, D and F) that represent different climate scenarios.
Seven states were evaluated: Chiapas, Estado de México, Michoacán, Guanajuato, Jalisco, Sinaloa and Tamaulipas. Modeled yield data coming from rainfed fields are
represented with a gray scale; a darker hue of gray indicates a higher yield value.
reproducing climate in Mexico (Conde and Gay, 2008). Projections
were made under two out of four different climate pathways proposed
by the IPCC (IPCC, 2014): +2.6 and +8.5 W/m2. These pathways or
representative concentration pathways (RCPs) represent scenarios of
different greenhouse gas concentrations in the atmosphere that relate to
differences in the solar energy being absorbed by the planet and the
energy radiated back to space. The two chosen pathways represent the
best- and worst-case scenarios. These scenarios were projected under
two temporal horizons: 2045–2069 and 2075–2099. With the best yield
model for each agricultural method, we wanted to project yield for each
municipality under current and future conditions, and depending on R2
values, the projected yield in the entire country.
Future predicted yields for each general circulation model were
averaged to produce ensemble results. Modeling ensembles have been
used as an alternative to deal with uncertainty coming from different
climate change models (Araújo and New, 2007). Alternatives to mitigate uncertainty coming from general circulation models are very important because it is one of the main sources of uncertainty in climate
change studies (Steen et al., 2017). Finally, we also decided to use the
We excluded zero yield values because we could not determine if
such values were related to the total loss of harvest, a decision not to
cultivate in that municipality, or a lack of recorded data. We generated
all possible linear models combining up to 10 bioclimatic variables,
building a total of 1023 models for each agricultural condition (rainfed
and irrigated). Model selection was performed with the sample corrected Akaike information criterion (AICc, see Appendix B; Burnham
and Anderson, 2002). When compared with the best model, a model
having a ∆AICc value < 2 (Burnham and Anderson, 2002) was considered to have a descriptive power like the best one. We then calculated the conditional and marginal R2 (R2c and R2m; Nakagawa and
Schielzeth, 2013) for each model to evaluate its goodness of fit with and
without including the municipality random effect, respectively.
To project yield in the studied municipalities under future climatic
conditions, we used the best-supported model and a 5-km2 climatology
from Worldclim (Hijmans et al., 2005) with three different general
circulation models: HADGEM2-ES, GFDL-CM3 and MPI-ESM-LR (IPCC,
2014). The performance of these general circulation models has been
evaluated, and they are considered to be the best models for
4
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
Fig. 3. Current and future yield distribution with irrigation. This figure shows five maps of Mexico (A, B, C, D and F) that represent different climate scenarios. Seven
states were evaluated: Chiapas, Estado de México, Michoacán, Guanajuato, Jalisco, Sinaloa and Tamaulipas. Modeled yield data coming from irrigation fields are
represented with a gray scale; a darker hue of gray indicates a higher yield value.
is the average breakeven points of all states evaluated (6.2 5 t/ha). In
irrigated fields, breakeven points varied significantly among states because of different variables such as: water costs, field lease (a fraction of
farmers are not landowners) and farm-worker salaries.
For rainfed fields we used a 2-t/ha yield as the breakeven point.
Rainfed fields are commonly grown by small-hold farmers (< 5 ha)
who own their land and have lower investment inputs. A previous study
estimated 3 t/ha as the breakeven point for rainfed growing conditions,
considering the cost of using hybrid seeds (UNISEM, 2017); nevertheless, most rainfed fields are cultivated using native races that reproduce their own seeds (Trueba, 2012). Given that no single specific
breakeven point has been suggested for rainfed fields, we carried out a
sensibility analysis beginning with 1 t/ha and finishing with 4 t/ha,
getting an idea of the percentage of municipalities that reached a yield
category in a particular year. The model was created with observed
data at the municipality level. Afterwards, we used observed data and
projected yields under future climatic conditions to calculate the
number of municipalities that reach their breakeven point under
rainfed and irrigated conditions.
ExDet tool (Mesgaran et al., 2014) to evaluate range change under the
future scenarios considered for the bioclimatic variables we used. ExDet
facilitates the visualization of geographic areas where future changes
are expected to be greater and where extrapolation might be troublesome because climate ranges change or variables are combined in different ways (Mesgaran et al., 2014).
2.3. Analyzing the breakeven point under current and future conditions
Once we had the modeled yields projected onto the geography
under current and future climate conditions, we tried to localize geographic regions reaching their breakeven point (when costs equal
profits). With this analysis, it was possible to have an idea of which
geographic areas are economically viable in terms of production under
the two agricultural conditions investigated here: irrigated and rainfed.
For irrigation we used two breakeven points: The first is a specific yield
value for each state based on studies carried out by FIRA (2007)
(Chiapas ∼ 5 t/ha, Guanajuato ∼ 7 t/ha, Jalisco ∼ 6 t/ha, Michoacán
∼ 6 t/ha, Sinaloa ∼8.5 t/ha and Tamaulipas ∼ 5 t/ha), and the second
5
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
have on crop yields is in accordance with a recently published work by
Tigchelaar et al. (2018), where decreased corn yields were reported
worldwide due to higher temperatures derived from climate change. At
a local scale, there are also studies showing how landraces from higher
altitudes (colder temperatures) have more difficulties being grown in
lowlands than the other way around (Mercer and Perales, 2010, 2018);
meaning that it might be more difficult to adapt to warmer conditions.
Our study contradicts what was found by Eakin et al. (2018), who
found that temperature did not play an important role in rainfed yields.
However, in their study, they failed to consider that during the period
they analyzed (the 1990s), there was an important reduction in the rate
of temperature increase, thought to be a result of the implementation of
the Montreal Protocol (diminishing the use of ozone-depleting substances) and due to changes in agricultural practices in Asia (translated
into a reduction in methane emissions) (Estrada et al., 2013); obscuring
temperature's role on yield fluctuations.
Higher temperatures at specific plant stages can be very detrimental
for their development and consequently negatively impact yield (Lobell
et al., 2011; Sánchez et al., 2014). For example, flowering and grainfilling stages are very heat sensitive in maize and sterility increases with
higher temperatures (Sánchez et al., 2014). There is a temperature
optimum beyond which yield decreases dramatically. In maize, the
optimum temperatures for most of its developmental phases range between 26 and 32 °C (Sánchez et al., 2014). However, Mexican native
maize races show an important amount of optimum climate conditions
(Ureta et al., 2012, 2016) and thresholds (Ureta et al., 2012). For example, the races Ancho, Blando de Sonora, Bofo, Chapalote, Conejo,
Elotero de Sinaloa, Olotillo, Pepitilla, Reventador, Tabloncillo, Tabloncillo Perla, Tuxpeño and Vandeño have maximum temperature
thresholds that go up to 40 °C (Ureta et al., 2012). These climatic
thresholds were obtained by overlapping geographic records with
temperature layers. Although this is an indirect approach (in comparison to direct physiological studies) of the plant's heat tolerance, these
thresholds can still provide some insight because they help document
the maximum temperature of a site where a given race has been grown.
As an adaptation strategy to climate change, it would be important to
focus attention on the native races listed above because they have been
grown under thermal stress and seem to survive it relatively well, thus,
they could be an additional basis for genetic improvement focused on
heat tolerance (Bellon et al., 2003; Smale et al., 2003). In this line of
thought, the maintenance of genetic diversity of maize and its local
adaptation through a continuous evolutionary process in the field is of
paramount importance for its adaptation to future climate conditions
(Mercer et al., 2012; Bellon et al., 2018). Consequently, the dynamic in
situ preservation of the diversity of maize by small-hold farms is an
evolutionary service that should be rewarded (Bellon et al., 2018).
Alternatively, the incorporation of irrigation in specific areas could be
considered because it has been observed that irrigation reduces the
negative impacts of temperature changes (Butler and Huybers, 2013).
This last strategy could nevertheless be unfeasible for most Mexican
fields because of future threats to our water supply (Kotschi, 2007).
Another explanation for rainfed areas being more influenced by
temperature than by precipitation, particularly in Mexico, is that
rainfed fields are commonly located in places that are suitable for maize
growth (Montesillo-Cedillo, 2016). Consequently, although some native
maize races are able to give good yield in areas with low precipitation
(Bellon et al., 2003; Smale et al., 2003), in most of these fields, precipitation is not necessarily scarce if we consider that the mean annual
precipitation average is of 1160 mm (SMN, 2017). In these rainfed
areas, the average precipitation is 23 mm and 118 mm for the driest
month (depending on the geographic area) and the dry quarter of the
year (three months), respectively. Consequently, most of the rain falls
during the rainy season, when maize is being grown. Of course, this
situation does not necessarily mean that dry spells do not occur
(Mendoza et al., 2006; Méndez and Magaña, 2010), and droughts are
expected to become more intense under climate change conditions
3. Results and discussion
3.1. Yield best models
After model selection, we obtained one best model for irrigated and
one best model for rainfed conditions. In both cases, all other competing models had ∆AICc values > 2. We calculated two R2 values: the
conditional (R2c), which takes into account the random effect of the
municipality factor, and the marginal (R2m), which does not. For the best
irrigation model, large differences between these two determination
coefficients existed, with R2c = 0.775 and R2m = 0.001. Similarly, for
the rainfed maize model, we obtained an R2c = 0.729 and an
R2m = 0.023. Even when the states under study were chosen in order to
encompass enough climate variability to potentially allow us to extrapolate our results to the entire Mexican territory, the municipality
identity had such a large effect on yield that the extrapolation could not
be performed. It is likely that other environmental and social factors
associated with the municipality's location have important effects on
yield. As it has been documented in other studies (Mercer and Perales,
2010; Ureta et al., 2013), maize cultivation and distribution in Mexico
is also related to environmental factors, such as soil type (fertility, organic matter, slope, texture and deepness) and altitude (Turrent
Fernández et al., 1997), which we did not integrate into our model
because we wanted to evaluate the effect of climate change in particular. Additionally, given the average nature of our yield data, recorded
at the municipality level, incorporating other factors, such as soil type
and altitude into our model would require adding assumptions because
several types of soil and different altitudes can be found in one single
municipality. The municipality is also related to social factors, such as
ethnic group or socioeconomic level that can affect maize distribution
and yield (Mercer and Perales, 2010; Ureta et al., 2013). Still, there is
evidence that supports that climate variables have an influence on yield
(Wilson, 1987; Grant et al., 1989; Ray et al., 2015; Ureta et al., 2016).
Even without spatial extrapolation, the seven states under study included hundreds of municipalities that in conjunction are very important for Mexico in terms of overall production and conservation of
the diversity of maize (Trueba, 2012; SIAP, 2018). Given that the climatology used for future climatic conditions is a 5-km cell map, we
averaged the values for each bioclimatic variable, taking into account
the number of cells that are included in every municipality.
Results evaluating whether the range of climate variables changed
over time show that the states studied will not present important range
changes (Appendix C). From the climate variables included in our
models, only two showed changes in their range under future conditions: maximum temperature of warmest month and mean temperature
of warmest quarter. The areas where these changes are expected to
occur are not widely distributed in the states evaluated (Appendix C)
consequently, uncertainties derived from time extrapolation in our results were not important.
3.2. Yield changes in rainfed fields under current and future conditions
In the following model we show that in rainfed fields, temperature,
not precipitation, is by far the most important variable affecting yield:
y = 1.20538 + 0.0001·x1 –0.0005·x2 –0.02301·x3 + 0.00411·x 4 + 0.00139·x5
(1)
where y is the log-transformed yield, x1 is the precipitation of the driest
month, x2 is the mean temperature of the coldest quarter, x3 is the mean
temperature of the warmest quarter, x4 is the minimum temperature of
the coldest month and x5 is the maximum temperature of the warmest
month. As shown by our model, the warmer it gets during the year, the
lower the yields are in the rainfed fields. Conversely, yield increased
with more precipitation and less extreme colds, but the more influential
variable was mean temperature increases in the warmest and coldest
quarters of the year. The important impact that warmer temperatures
6
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
(Mendoza et al., 2006). Another possible explanation to the more deleterious effects on yield of temperature rather than precipitation could
be that higher temperatures increase rates of evapotranspiration both
from the crop and from the substrate, increasing photosynthetic stress
and removing moisture from the soil (Taiz and Zeiger, 2006), as well as
potentially affecting the most sensitive stages in maize: flowering
(Bellon, 1991).
The fact that reductions in yield in rainfed fields were strongly related to temperature increases reveals their vulnerability to climate
change (there is a strong consensus about temperature increases among
general circulation models; IPCC, 2014). As shown in Fig. 2, under
current climate conditions, the highest yield category for rainfed agriculture goes from 6 to 8 t/ha, while the highest yield category in the
future decreases substantially to 0.75–1 t/ha. Under current climate
conditions, the highest yields are visible in rainfed areas in Jalisco, but
under future climatic conditions most of the rainfed territory is predicted to have an average yield of 0.25 to 0.5 t/ha.
It is also important to bear in mind that our dramatic results under
different climate change scenarios do not take into account the important role that diversity in general and particularly maize agrobiodiversity has on adapting to a changing environment, while also not
considering the dynamic management of native maize races by farmers
who exert an ongoing selection pressure for plants that will produce
seed under even the most stressful conditions (Bellon et al., 2018). For
example, Smale et al. (2003) found that the race Bolita was being
grown in Oaxaca because of its resistance to the canicula (a period of
extreme warm weather in the middle of the rainy season; Smale et al.,
2003). There are also other native races adapted to hydric stress, such
as Chapalote, Dulcillo del Noroeste, Gordo, Tablilla de Ocho, Cónico
Norteño and Tuxpeño Norteño (Ruiz Corral et al., 2008 and Ruiz Corral
et al., 2013). The length of life cycle is another characteristic to consider because varieties with shorter life cycles fulfill their development
in a shorter period of time and growing them once the rainy season has
started represents a strategy that has been applied by peasants in
Mexico to attain a harvest even when the rainy season starts late
(Munguía-Aldama et al., 2015). Also, short life cycles increase the
probability of finishing the setting of seed before or shortly after the
canicular period and avoiding heat stress during flowering (Bellon,
1991). This strategy has already been selected upon and is common in
native races, such as Apachito and Nal-Tel, while other races have very
long-life cycles, such as Tehua, Jala, Tuxpeño and Chalqueño (Ortega
Paczka, 2007). Additionally, it is important to stress that native races
are also being grown because of their better grain quality and lower
susceptibility to pests in the field and under storage (Trueba, 2012), as
well as their capacity to grow well in places where improved modern
varieties do not fare well.
Most rainfed fields are cultivated by small-hold farmers (with < 5
ha, producing 1.6–2 t/ha) who primarily use native races (Cruz Delgado
et al., 2012; Perales and Golicher, 2014); these varieties contain the
majority of the genetic diversity recorded for maize (Vigouroux et al.,
2008). The great adaptive capacity of these races is obscured by a lack
of detail on the municipality data, where only average yield but not
maize type or race used is reported (SIAP, 2018). Furthermore, during
the last three decades of anthropogenic-induced climate change, smallhold, rainfed areas have shown to be resilient in terms of maize cultivation and production (Eakin et al., 2018). Consequently, the human
factor plays a very important role in maize production's resilience, as it
is being shown not only by small farmers facing climate change, but
also by large-scale producers who report high yields despite cultivating
land which is not highly suitable from a climatic standpoint (Eakin
et al., 2018). It should be noted, however, that large-scale producers are
at a great economic and technical advantage over small-hold farms
because the large-scale producers get the majority of direct and indirect
subsidies for maize production in Mexico (Eakin et al., 2018).
3.3. Yields in irrigated fields under current and future conditions
Irrigated fields show greater yield stability than rainfed fields under
different climate conditions, but they can be affected by specific climatic variables (their best model depended only on two climate variables instead of the five variables considered for rainfed fields).
Surprisingly, precipitation had greater influence on their yield than
temperature:
y = 1.5218 + 0.00134·x1 –0.00018·x2
(2)
where y is the log-transformed yield, x1 is the precipitation of the driest
month and x2 is the minimum temperature of the coldest month. Thus,
both precipitation and temperature influence maize yield even when
there is access to irrigation. The more it rains during the driest month of
the year, the more they can yield. We also found a negative correlation
between yield and the minimum temperature of the coldest month. A
possible explanation for this result is that water coming from irrigation
helps diminish negative impacts from extreme temperatures (cold and
hot) (personal observations).
The dependence of irrigated fields on precipitation could be explained by the fact that irrigation was first implemented in Mexico as a
way to expand the agricultural frontiers in arid and semiarid areas
(Montesillo-Cedillo, 2016). That is why even when this activity is
consuming around 75% of the country's water (CONAGUA, 2014), these
fields still require precipitation. Furthermore, with the exception of a
few municipalities, irrigation in the country is inefficient (Vélez and
Saez, 2012). Fields irrigated with more water produce higher yields
across the country. As a result, federal resources, such as subsidies for
water and its transport, are targeted toward farmers with the capacity
to irrigate (Huacuja, 2015).
Our geographic projections show that the highest yield category
(9–13 t/ha) under irrigation was widely distributed in Sinaloa and it
would actually increase under predicted climate conditions (Fig. 1).
However, yields do not increase in the same way in all scenarios or at
all time-scales. Yield in 2050 scenarios show increases in Sinaloa, Jalisco, Guanajuato and even Tamaulipas and Chiapas. These increments
are reduced by the year 2070 under the RCP 8.5 W/m2 scenario. The
variable precipitation of the driest month, which was the most influential for yield in irrigated fields, is expected to increase in the area
evaluated under future climate conditions (Hijmans et al., 2005). The
lowest value in the future is expected for the year 2070 RCP 8.5 W/m2
scenario, as are the lower yield values.
Our results in irrigated fields are in line with what has been found
globally: more intensively managed fields are less susceptible to climate
change (Tigchelaar et al., 2018). This apparent stability could be related to the fact that climate scenarios represent average climate conditions in the future, akin to a photograph of the future (IPCC, 2014)
that does not account for climate variability. Nevertheless, variability is
expected to increase and in a more variable world, genetic diversity will
be indispensable (Kotschi, 2007; Visser, 2008) not only for adaptation
to changes in climate but also to withstand other factors derived from
climate change, like new potential pests (Rosenzweig et al., 2001;
Lamichhane et al., 2015). Under these circumstances, monocultures of
genetically homogenous plants, even with irrigation, become vulnerable to any change in their environment. We agree that greater and
improved technology, such as the use of irrigation, machinery and
fertilizers increases adaptive capacity (Chhetri et al., 2012); this is
evident from our results showing a lesser impact on yield from climate
change in irrigated fields. However, we argue that under a variable
world undergoing rapid change, agrobiodiversity will be as important,
if not more important, than technology (Conde et al., 1997; Kotschi,
2007; Visser, 2008). In irrigated areas, the genetic diversity of hybrid
and improved varieties could also be incremented. Nowadays, there
are > 300 hybrid maize varieties in the Mexican market, but very few
are being grown everywhere (Trueba, 2012). The great importance of
biodiversity in a changing world has not been sufficiently stressed in
7
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
The estimated breakeven points in irrigated fields of the states
evaluated here are: Chiapas ∼ 5 t/ha, Guanajuato ∼ 7 t/ha, Jalisco ∼
6 t/ha, Michoacán ∼ 6 t/ha, Sinaloa ∼8.5 t/ha and Tamaulipas ∼ 5 t/
ha (FIRA, 2007). Estado de México was the only state that did not have
a specific breakeven point reported, thus we used the average of the
other states' breakeven points (6 t/ha) as a proxy. The second threshold
we used for all evaluated states was the average yield for the seven
states: 6.2 t/ha. In irrigated fields, farmers are barely producing enough
maize under current conditions to cover production costs (Trueba,
2012). Under current climate conditions, only 30% of the municipalities evaluated that have irrigation systems reached their specific
states' breakeven point (Fig. 4) or the averaged breakeven point (6.2 t/
ha). States such as Sinaloa—the one with the highest yields in the
country—did not reach its own breakeven point (8.5 t/ha) in a large
number of its fields (Fig. 4). Meanwhile, in other high yield states, like
Jalisco and Guanajuato, the breakeven point is only attained in less
than half of the municipalities surveyed. If we consider the averaged
breakeven point (6.2 t/ha) as reference for comparison, a larger geographic area would achieve breakeven status. Nevertheless, we found
that in some states, such as Chiapas, Estado de México and Tamaulipas,
it is not profitable to grow maize under irrigation in either current or
future climatic conditions (Table 2).
Our results show that irrigated fields have much higher yields than
rainfed fields, but irrigated maize production is not economically viable
in 70% of the studied municipalities (Fig. 4). This is because despite the
higher yield of irrigated plots, production costs consume most of the
profit, making subsidies indispensable for agriculture under irrigation
(Huacuja, 2015). These results suggest that both subsidizing and giving
technical support to improve agricultural practices in rainfed maize
fields could be a more strategic way of attaining food security at a lower
economic cost, while simultaneously promoting the active conservation
of maize agrobiodiversity. In line with our results, a study carried out
by Guzmán Soria et al. (2014) showed that maize production in Guanajuato under irrigation is not competitive because even when yields
increase in 50% in comparison to rainfed fields, costs increase by 65%.
Thus, it is currently too expensive to cultivate maize in the way
Mexican farmers are doing it under intensified, irrigated monocultures
(Guzmán Soria et al., 2014). Under future climatic conditions, the costbenefit relationship is projected to worsen. A way to start reducing costs
would be to avoid buying seed every year. In the short term, farmers
would probably have reduced yield due to segregation of characters in
hybrids (Reyes, 1990), but it would still be economically convenient
due to cost reductions. Currently, 92% of commercial seed used in high
production states in Mexico is commercialized by large corporations
whose prices are unregulated (Trueba, 2012), translating into having
one of the highest prices for maize hybrid seeds worldwide (Espinosa
et al., 2003). Under climate change conditions, there will be unprecedented volatility in maize prices (Tigchelaar et al., 2018) and
consequently on maize seeds. This situation strongly suggests that
Mexico should not depend on imports for staple crops, which are now
above 30% of what is being consumed in the country (SAGARPAb,
2016). It is time for farmers to start saving their own seeds and for the
Mexican government to enhance the development of national hybrids
and other improved varieties, stimulate national hybrid production,
and provide farmers with seed supply options. There should also be
guidance for proper management in terms of pesticides and fertilizers
(Lechenet et al., 2014), which could also reduce costs. It would be very
helpful for farmers to have specific production guidance for each geographic area in order to be more efficient (Trueba, 2012), including
knowing which kind of improved seeds can be grown at each location.
With this kind of support, hybrid diversity and local selection and improvement efforts could also be increased. Additionally, other strategies
of genetic improvement have focused on generating open pollinated
varieties rather than hybrids, which are inherently unstable in subsequent generations. This approach should be actively promoted and
financed.
Table 1
Sensibility analysis for rainfed municipalities reaching different possible
breakeven points.
Yield (t/ha)
Evaluated municipalities reaching the breakeven point (%)
>1
> 1.5
>2
> 2.5
>3
> 3.5
>4
97
92
82
72
39
49
36
Evaluated area reaching the breakeven point: The percentage of area (5 km)
that could reach the breakeven point under current climatic conditions in
rainfed areas.
global studies addressing possible climate change impacts on crops (Ray
et al., 2015; Tigchelaar et al., 2018). Unfortunately, given the nature of
our yield data, the importance of agrobiodiversity has been obscured
and could not be integrated, but it remains a central subject in a
changing world.
3.4. Reaching the breakeven point under current and future conditions
Although rainfed maize fields yield less, a higher percentage of the
municipalities growing maize this way is able to reach their breakeven
point compared with those using irrigated fields (see Tables 1 and 2).
Given that there is an estimate of 3 t/ha counting hybrid seeds' cost for
rainfed fields (UNISEM, 2017), a 2 t/ha could be expected because most
rainfed farmers use native races and produce their own seeds. At a
breakeven point of 2 t/ha, 82% of the municipalities evaluated reached
it. If cost reduction is further considered due to a minor use of fertilizers
and pesticides, a 1 t/ha could be a breakeven point that 97% of municipalities analyzed could reach (Table 1). Practices such as seed selection after each cultivation cycle, seed exchange and local selection
efforts, not only help maintain the genetic diversity of maize in situ
(Mercer and Perales, 2010), but they also make it more profitable to
produce in this way (Fig. 4, Table 1). In addition, rainfed fields do not
have the cost demands associated with the electricity and fuel costs
required to pump water for irrigation. Thus, policies and management
need to take into account the considerable contribution made by
farmers who use rainfed fields (Mercer et al., 2012).
Table 2
Geographic area reaching the breakeven point under current and future climatic conditions.
% Area above break-even point
Current
RCP 2.6
RCP 8.5
2050
2070
2050
2070
Irrigation
30
37
37
37
38
Irrigation average
30
34
36
36
37
Rainfed
82
0
0
0
0
RCP 2.6: best case climate change scenario, RCP 8.5: worst case climate change
scenario, 2050: 2050 scenario (2045–2069), 2070: 2070 scenario (2075–2099).
Breakeven point: yield at which costs and profit are in equilibrium. Irrigation:
maize grown under irrigation. Each evaluated state had its own breakeven
point, as suggested by several studies carried out by FIRA in different regions of
the country (2017) (Chiapas = 5 t/ha, Guanajuato = 7 t/ha, Jalisco = 6 t/ha,
Michoacán = 6 t/ha, Sinaloa = 8.5 t/ha, Tamaulipas = 5 t/ha and Estado de
México = 6 t/ha. Irrigation Average: Percentage of area reaching the average
yield from studied cities that was 6.2 t/ha.
8
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
Fig. 4. Areas reaching the Breakeven point. A)
Geographic area with irrigation that reaches the
specific breakeven point of each state under current
climatic conditions, B) Geographic area with irrigation that reaches the average breakeven point of each
state under future climate conditions, C) Geographic
area under rainfed conditions that reaches 2 t/ha
under current climate conditions.
seven Mexican states that represent the five regions of the country
encompassing 65% of national maize production. It is also the first
study that evaluates the impacts of climate change on rainfed and irrigated maize fields. We conclude that climate change will have
strongly negative impacts on yield for rainfed fields, while irrigated
fields will remain stable. In general, temperature increase will negatively affect yield in rainfed fields, while precipitation will have a larger
impact on irrigated fields. In rainfed areas the variable that influenced
yield the most was the temperature of the warmest quarter (mostly
related to the spring-summer season). From all the variables evaluated,
five influenced yield in rainfed areas and only one was associated with
precipitation (precipitation of the driest month) and its contribution to
decrements in yield was small. On the other hand, in irrigated areas,
only two variables had an influence over yield value: one precipitation
variable (precipitation of driest month) and one temperature variable
(minimum temperature of coldest month). However, the influence of
precipitation is greater than the influence of temperature. The fact that
irrigated fields even showed yield increases under three evaluated
scenarios is possibly related to a future increase in the precipitation of
the driest month in three scenarios, except for 2070 RCP 8.5 W/m2
scenario in which decreases in precipitation and yield are expected. Our
climatic mixed model helped us visualize that when the municipality
was taken into account: the climatic variables used can explain a great
amount of yield variability (R2c = 0.729 and R2c = 0.775, for rainfed
and irrigated conditions, respectively). Consequently, yield extrapolation to future scenarios in the municipalities evaluated could be projected but this extrapolation to other geographic areas of the country
was not possible. We also demonstrated that higher yield does not necessarily result in higher profits for maize producers, particularly in
irrigated conditions because production costs are high.
To abate negative impacts on yield such as those projected by climate change scenarios, we should increase the ability of small-hold
farmer to adapt through more equitable water distribution. Costs in
irrigated field can be offset by increasing the use of locally adapted,
domestic hybrid seeds. While we could not evaluate it explicitly, we
want to highlight that we could take advantage of the standing Mexican
maize agrobiodiversity and its known capability of reacting differently
under different circumstances. In irrigated fields where such diversity is
scarce, it should be increased by generating different hybrid varieties,
3.5. Limitations
The use of yield data collected at the municipality level presents two
main problems for modeling its relationship with climate change. On
the one hand, climate variables were taken from meteorological stations, and in many cases, there were more than one station per municipality. In these cases, we had to average this information.
Additionally, there were several stations with incomplete information.
Furthermore, while extreme climate events, such as droughts and
storms, which play a very important role in yield (Deryng et al., 2014),
are expected to increase in frequency and intensity (IPCC, 2013); as far
as we know, there are still no geographic layers representing these
extreme climatic events.
Additionally, we acknowledge that our study lacks a detailed description of the plant's physiology through time under different climatic
conditions and along its entire lifecycle. However, this is a trade-off
between being able to project the plants ecological response into spatiotemporal macro scales and having a better understanding of its ecology
at a microscale. Both approaches answer different questions and are of
relevance to gain understanding about the plants' biology.
Finally, given the lack of information about which native varieties
are being sown where and a lack of distinction among these kinds of
materials and improved or hybrid varieties when collecting yield estimates, the role of agrobiodiversity under a changing environment is
being obscured. Proper yield information about specific races and/or
hybrid varieties does not exist or is not publicly available. The
CONABIO database has some yield data for native races, but it is still
incomplete (CONABIO, 2017). If yield data from specific hybrid varieties or native races could be collected in a broad enough scale to allow
for extrapolation, specific adaptive strategies could be analyzed per
geographic region. However; guidelines are still needed to undertake
this kind of survey. Nevertheless, we think this investigation gives some
insights on the limitations of maize production in Mexico under current
and future scenarios.
4. Conclusions
This study is the first attempt to evaluate the relationship between
yield and 10 different climate variables at the municipality scale over
9
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
as well as improved native races. We should also explore new germoplasm to breed for higher resistance under heat and hydric stress.
C.F., Davenport, F., Bausch, J.C., 2018. Agricultural change and resilience: agricultural policy, climate trends and market integration in the Mexican maize system.
Anthropocene. 23, 43–52.
Espinosa, A., Sierra, M., Gómez, N., 2003. Producción y tecnología de semillas mejoradas
de maíz por el INIFAP en el escenario sin la PRONASE. Agronomía mesoamericana.
pp. 14.
Estrada, F., Perron, P., Martínez-López, B., 2013. Statistically derived contributions of
diverse human influences to twentieth-century temperature changes. Nat. Geosci. 6,
1050.
FAO, 2017. Data. Organización de las Naciones Unidas para la Alimentación y la
Agricultura. (Accessed September 2019). http://www.fao.org/faostat/en/#
data/QC.
Fernández Suárez, R., Morales Chávez, L.A., Gálvez Mariscal, A., 2013. Importancia de los
maíces nativos de México en la dieta nacional: Una revisión indispensable. Rev.
Fitotec. Mex. 36, 275–283.
FIRA, 2007. Rentabilidad y Costos. Fideicomisos Instituidos en Relación con la
Agricultura. (Accessed September 2019). http://www.fira.gob.mx/Nd/
Rentabilidad_y_Costos.jsp.
FIRA, 2016. Dirección de Investigación y Evaluación Económica y Sectorial. Panorama
Agroalimentario. Fideicomisos Instituidos en Relación con la Agricultura, México
(Accessed September 2019). https://www.gob.mx/fira/documentos/panoramaagroalimentario.
Florescano, E., 2003. Imágenes y significados del dios del maíz. In: Marielle, Esteva G.Y.C.
(Ed.), Sin maíz no hay país. CONACULTA, Mexico City, pp. 36–55.
Grant, R.F., Jackson, B.S., Kiniry, J.R., Arkin, G.F., 1989. Water deficit timing effects on
yield components in maize. Agron. J. 81, 61–65.
Guzmán Soria, E., de la Garza Carranza, M.T., González Farías, J.P., Hernández Martínez,
J., 2014. Análisis de los costos de producción de maíz en la Región Bajío de
Guanajuato. Análisis Económico. 29.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A., 2005. Very high-resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25,
1965–1978.
Huacuja, F.E., 2015. Government Subsidies for Price Risks: The Case of Maize Producers
in Mexico. (Papeles de Geografía:1).
IPCC, 2013. Summary for policymakers. In: Stocker, T.F., Qin, D., Plattner, G.-K., Tignor,
M., Allen, S.K., Boschung, J. ... Midgley, P.M. (Eds.), Climate Change 2013: The
Physical Science Basis. Contribution of Working Group I to the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge University
Press, Cambridge, United Kingdom and New York, NY, USA.
IPCC, 2014. Summary for policymakers. In: Field, C.B., Barros, V.R., Dokken, D.J., Mach,
K.J., Mastrandrea, M.D., Bilir, T.E. ... White, L.L. (Eds.), Climate Change 2014:
Impacts,Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects.
Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.
Jones, C.A., Kiniry, J.R., 1986. CERES-Maize: A Simulation of Maize Growth and
Development. Texas A&M University Press, College Station.
Kato, T.A., Mapes, C., Mera, L.M., Serratos, J.A., Bye, R.A., 2009. Origen y diversificación
del maíz: una revisión analística. Universidad Nacional Autónoma de México,
Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, México, DF.
Khan, M.B., Hussain, N., Iqbal, M., 2001. Effect of water stress on growth and yield
components of maize variety YHS 202. Journal of Research Science. 12, 15–18.
Kotschi, J., 2007. Agricultural biodiversity is essential for adapting to climate change.
GAIA-Ecological Perspectives for Science and Society. 16, 98–101.
Lamichhane, J.R., Barzman, M., Booij, K., Boonekamp, P., Desneux, N., Huber, L., Kudsk,
P., Langrell, S.R., Ratnadass, A., Ricci, P., 2015. Robust cropping systems to tackle
pests under climate change. A review. Agron. Sustain. Dev. 35, 443–459.
Lechenet, M., Bretagnolle, V., Bockstaller, C., Boissinot, F., Petit, M.-S., Petit, S., MunierJolain, N.M., 2014. Reconciling pesticide reduction with economic and environmental sustainability in arable farming. PLoS One 9, e97922.
Lobell, D.B., Bänziger, M., Magorokosho, C., Vivek, B., 2011. Nonlinear heat effects on
African maize as evidenced by historical yield trials. Nat. Clim. Chang. 1, 42.
Maguire Jr., B., 1973. Niche response structure and the analytical potential of its relationships to the habitat. Am. Nat. 107, 213–246.
Martínez-Meyer, E., Díaz-Porras, D., Peterson, A.T., Yáñez-Arenas, C., 2013. Ecological
niche structure and rangewide abundance patterns of species. Biol. Lett. 9, 20120637.
Matsuoka, Y., Vigouroux, Y., Goodman, M., Sánchez Gonzáles, J.J., Buckler, E.D.,
Doebley, J., 2002. A single domestication for maize shown by multilocus microsatellite genotype. Proceedings of the National Academy of Science of the United
States of America. 99, 6080–6084.
Méndez, M., Magaña, V., 2010. Regional Aspects of Prolonged Meteorological Droughts
over Mexico and Central America. pp. 1175–1188.
Mendoza, B., Velasco, V., Jáuregui, E., 2006. A study of historical droughts in southeastern Mexico. J. Clim. 19, 2916–2934.
Mercer, K.L., Perales, H.R., 2010. Evolutionary response of landraces to climate change in
centers of crop diversity. Evol. Appl. 3, 480–493.
Mercer, K.L., Perales, H., 2018. Structure of local adaptation across the landscape:
flowering time and fitness in Mexican maize (Zea mays L. subsp. mays) landraces.
Genet. Resour. Crop. Evol. 1–19.
Mercer, K.L., Perales, H.R., Wainwright, J.D., 2012. Climate change and the transgenic
adaptation strategy: smallholder livelihoods, climate justice, and maize landraces in
Mexico. Glob. Environ. Chang. 22, 495–504.
Mesgaran, M.B., Cousens, R.D., Webber, B.L., 2014. Here be dragons: a tool for quantifying novelty due to covariate range and correlation change when projecting species
distribution models. Divers. Distrib. 20, 1147–1159.
Monterroso, R.A., Conde, A.C., Gómez Díaz, J.D., Gay, C., 2011. Assessing current and
Declaration of Competing Interest
None.
Acknowledgements
We would like to thank DGAPA, Instituto de Ecología (UNAM) and
CONACyT (Cátedras CONACyT) for the economic support to Carolina
Ureta. We would also like to thank Cecilio Mota and the group of
Análisis de Riesgo y Bioseguridad (CONABIO) for their help. Finally, we
would like to acknowledge financial support from the following grants:
CONACYT240180, 2015-01-687; UNAM-DGAPA-PAPIIT: IN211516,
IN208517, IN205517, IN204217, IE-485.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.agsy.2019.102697.
References
Anderson, E.F., Cutler, H.C., 1942. Races of Zea mays: I. their recognition and classification. Ann. Mo. Bot. Gard. 29, 69–88.
Araújo, M.B., New, M., 2007. Ensemble forecasting of species distributions. Trends in
Ecology and Evolution. 22 (42–27).
ASERCA, 2018. Maíz grano cultivo representativo de México. Agencia de Servicios a la
Comercialización y Desarrollo de Mercados Agropecuarios (Accessed September
2019). https://www.gob.mx/aserca/articulos/maiz-grano-cultivo-representativo-demexico?idiom=es.
Bellon, M.R., 1991. The ethnoecology of maize variety management: a case study from
Mexico. Hum. Ecol. 19, 389–418.
Bellon, M.R., Berthaud, J., Smale, M., Aguirre, J.A., Taba, S., Aragón, F., Díaz, J., Castro,
H., 2003. Participatory landrace selection for on-farm conservation: an example from
the central valleys of Oaxaca, Mexico. Genet. Resour. Crop. Evol. 5, 401–416.
Bellon, M.R., Mastretta-Yanes, A., Ponce-Mendoza, A., Ortiz-Santamaría, D., OliverosGalindo, O., Perales, H., Acevedo, F., Sarukhán, J., 2018. Evolutionary and food
supply implications of ongoing maize domestication by Mexican campesinos.
Proceedings of the Royal Society in Biological Science. 285, 20181049.
Bourges, H., 2013. El maíz: su importancia en la alimentación de la población mexicana.
In: Álvarez-Buylla, E.R., Piñeyro Nelson, A. (Eds.), El maíz en peligro ante los
transgénicos: un análisis integral del caso de México. UNAM, UCCS, UV, pp. 231–248.
Burnham, K.P., Anderson, D.R., 2002. Model Selection and Inference: A Practical
Information-Theoretic Approach, 2nd edition. Springer-Verlag, New York. https://
doi.org/10.1007/b97636.
Butler, E.E., Huybers, P., 2013. Adaptation of US maize to temperature variations. Nat.
Clim. Chang. 3, 68.
Cakir, R., 2004. Effect of water stress at different development stages on vegetative and
reproductive growth of corn. Field Crop Res. 89, 1–16.
Chhetri, N., Chaudhary, P., Tiwari, P.R., Yadaw, R.B., 2012. Institutional and technological innovation: understanding agricultural adaptation to climate change in Nepal.
Appl. Geogr. 33, 142–150.
Cicchino, M., Edreira, J., Uribelarrea, M., Otegui, M., 2010. Heat stress in field-grown
maize: response of physiological determinants of grain yield. Crop Sci. 50,
1438–1448.
CONABIO, 2017. Proyecto Global de Maíces. Comisión Nacional para el Conocimiento y
Uso de la Biodiversidad. (Accessed September 2019). https://www.biodiversidad.
gob.mx/genes/proyectoMaices.html.
CONAGUA, 2014. Diagnóstico del programa presupuestario K141. Rehabilitación y
modernización de infraestructura de riego y rainfed tecnificado. Comisión Nacional
del Agua (Accessed September 2019). https://www.transparenciapresupuestaria.
gob.mx/work/models/PTP/Reingenieria_Gasto/imagenes/Ventanas/Ramo_16/
16K141.pdf.
Conde, A.C., Gay, C., 2008. Guía para la Generación de Escenarios de Cambio Climático a
Escala Regional. Centro de Ciencias de la Atmósfera, UNAM, México, D.F.
Conde, C., Liverman, D., Flores, M., Ferrer, R., Araújo, R., Betancourt, E., Villareal, G.,
Gay, C., 1997. Vulnerability of rainfed maize crops in Mexico to climate change.
Clim. Res. 9, 17–23.
Cruz Delgado, S., Gómez Valdez, M., Ortiz Pulido, M.E., Entzana Tadeo, A.M., Suárez
Hernández, C., Santillán Moctezuma, V., 2012. Situación actual y perspectivas del
maíz en México 1996–2012. Servicio de información agroalimentaria y pesquera,
México, D.F.
Deryng, D., Conway, D., Ramankutty, N., Price, J., Warren, R., 2014. Global crop yield
response to extreme heat stress under multiple climate change futures. Environ. Res.
Lett. 9, 034011.
Eakin, H., Sweeney, S., Lerner, A.M., Appendini, K., Perales, H., Steigerwald, D.G., Dewes,
10
Agricultural Systems 177 (2020) 102697
C. Ureta, et al.
Agroalimentaria y Pesquera (Accessed September 2019). http://infosiap.siap.gob.
mx/aagricola_siap_gb/icultivo/index.jsp.
Smale, M., Bellon, M.R., Aguirre, J.A., Rosas, I.M., Mendoza, J., Solano, A.M., Martínez,
R., Ramırez, A., Berthaud, J., 2003. The economic costs and benefits of a participatory project to conserve maize landraces on farms in Oaxaca, Mexico. Agric. Econ. 29,
265–275.
SMN, 2017. Información climatológica. Servicio Meteorológico Nacional. (Accessed
September 2019). https://smn.conagua.gob.mx/es/informacion-climatologica-porestado.
Steen, V., Sofaer, H.R., Skagen, S.K., Ray, A.J., Noon, B.R., 2017. Projecting species'
vulnerability to climate change: which uncertainty sources matter most and extrapolate best? Ecology and Evolution. 7, 8841–8851.
Taiz, L., Zeiger, E., 2006. Plant physiology. Sinauer. Sunderland, U.S.A.
Tigchelaar, M., Battisti, D.S., Naylor, R.L., Ray, D.K., 2018. Future warming increases
probability of globally synchronized maize production shocks. Proceedings of the
National Academy of Sciences 201718031.
Trueba, C., 2012. Semillas mexicanas mejoradas de maíz: su potencial productivo.
Colegio de Posgraduados, México, Estado de México.
Turrent Fernández, A., Montiel, N.G., Díaz, J.R., Andrade, H.M., Corona, A.O., Flores,
M.L., 1997. Plan de investigación del sistema maíz-tortilla en los Estados Unidos
Mexicanos. Document for internal agency circulation. Instituto Nacional de
Investigaciones Forestales, Agrícolas y Pecuarias, México, DF.
UNISEM, 2017. Unidos Sembrados. (Accessed September 2019). http://www.
semillastodoterreno.com.
Ureta, C., Martínez-Meyer, E., Perales, H.R., Álvarez-Buylla, E., 2012. Projecting the effects of climate change on the distribution of maize races and their wild relatives in
Mexico. Glob. Chang. Biol. 18, 1073–1082.
Ureta, C., González-Salazar, C., González, E.J., Álvarez-Buylla, E.R., Martínez-Meyer, E.,
2013. Environmental and social factors account for Mexican maize richness and
distribution: a data mining approach. Agric. Ecosyst. Environ. 179, 25–34.
Ureta, C., Martínez-Meyer, E., González, E.J., Álvarez-Buylla, E.R., 2016. Finding potential high-yield areas for Mexican maize under current and climate change conditions. J. Agric. Sci. 154, 782–794.
Vélez, E.P., Saez, E.M., 2012. Water use for agriculture in Mexico. Water Resources in
Mexico. Springer, Germany, Berlin.
Vigouroux, Y., Glaubitz, J., Matsuoka, Y., Goodman, M., Sánchez González, J.J., Doebley,
J., 2008. Population structure and genetic diversity of New World maize races assessed by DNA microsatellites. Am. J. Bot. 95, 1240–1253.
Visser, M.E., 2008. Keeping up with a warming world; assessing the rate of adaptation to
climate change. Proc. R. Soc. Lond. B Biol. Sci. 275, 649–659.
Wilson, E.O., 1987. The little things that run the world (the importance and conservation
of invertebrates). Conserv. Biol. 1, 344–346.
potential rainfed maize suitability under climate change scenarios in Mexico.
Atmósfera. 24, 53–67.
Montesillo-Cedillo, J.L., 2016. Rendimiento por hectárea del maíz grano en México:
distritos de riego vs temporal. Economía Informa. 398, 60–74.
Munguía-Aldama, J., Sánchez-Plata, F., Vizcarra-Bordi, I., Rivas-Guevara, M., 2015.
Estrategias para la producción de maíz frente a los impactos del cambio climático.
Revista de Ciencias Sociales (Ve) 21.
Nakagawa, S., Schielzeth, H., 2013. A general and simple method for obtaining R2 from
generalized linear mixed-effects models. Methods Ecol. Evol. 4, 133–142.
Nix, H.A., 1986. A Biographic Analysis of Australian Elapid Snakes. Australian
Government Publishing Service. Australia, Canberra.
Ortega Paczka, R., 2007. La diversidad del maíz en México. Sin maíz no hay país. Consejo
Nacional Para la Cultura y las Artes, México, D.F, pp. 123–154.
Perales, H., Golicher, D., 2014. Mapping the diversity of maize races in Mexico. PLoS One
9 (12), e114657.
Ranum, P., Peña-Rosas, J.P., Garcia-Casal, M.N., 2014. Global maize production, utilization, and consumption. Ann. N. Y. Acad. Sci. 1312, 105–112.
Ray, D.K., Gerber, J.S., MacDonald, G.K., West, P.C., 2015. Climate variation explains a
third of global crop yield variability. Nat. Commun. 6, 5989.
Reyes, C., 1990. El maíz y su cultivo. AGT-EDITOR S.A, México D.F.
Rosenzweig, C., Iglesias, A., Yang, X.B., Epstein, P.R., Chivian, E., 2001. Climate change
and extreme weather events; implications for food production, plant diseases, and
pests. Global Change and Human Health. 2, 90–104.
Ruiz Corral, J.A., Durán Puga, N., Sánchez Gonzáles, J.J., Ron Parra, J., González Euiarte,
R., Holland, J.N., Medina García, G., 2008. Climatic adaptation and ecological descriptors of 42 Mexican races. Crop Sci. 48, 1502–1512.
Ruiz Corral, J.A., Sánchez González, J.J., Hernández Casillas, J.M., Willcox, M.C.,
Ramírez Ojeda, G., Ramírez Díaz, J.L., González Eguiarte, D.R., 2013. Identificación
de razas mexicanas de maíz adaptadas a condiciones deficientes de humedad mediante datos biogeográficos. Revista Mexicana de Ciencias Agrícolas. 4, 829–842.
SADER, 2019. Oficinas de Operación Territorial. Secretaría de Agricultura y Desarrollo
Rural. (Accessed September 2019). https://www.gob.mx/agricultura.
SAGARPA a, 2016. Sube a 12 mil hectáreas siniestradas totalmente por las heladas.
Secretaría de Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación
(Accessed September 2019). https://www.gob.mx/agricultura%7Csinaloa/
articulos/sube-a-12-mil-hectareas-siniestradas-totalmente-por-las-heladas.
SAGARPA b, 2016. Planeación Agrícola Nacional 2017-2030. Maíz grano blanco y
amarillo mexicano. Secretaría de Agricultura, Ganadería, Desarrollo Rural, Pesca y
Alimentación (Accessed September 2019). https://www.gob.mx/cms/uploads/
attachment/file/256429/B_sico-Maiz_Grano_Blanco_y_Amarillo.pdf.
Sánchez, B., Rasmussen, A., Porter, J.R., 2014. Temperatures and the growth and development of maize and rice: a review. Glob. Chang. Biol. 20, 408–417.
SIAP, 2018. Anuario Estadístico de la Producción Agrícola. Servicio de Información
11