Crop requirements matrices

The matrices of environmental requirements of the three crops of the study, corn, sorghum, wheat, were elaborated, based on a bibliographic review in technical data sheets, FAO guides and articles related to the crops, emphasizing the climatic variables such as quantity and distribution of the annual average precipitation, water balances, monthly and annual average temperature, in addition to the soil and landscape conditions in which they have the best development, considering the areas where these crops are currently established. Four aptitude grade classes were used: suitable, moderately suitable, marginally suitable, and unsuitable.

Preparation of maps of degree of aptitude of crops

From the matrices of environmental requirements, the current aptitude maps were elaborated for each one of the crops of the study in the areas with agricultural use of the country, which were delimited from the map of series VI of the use of the soil and vegetation (^{INEGI, 2017}), this with ArcGis 10.3 Software. For all the agricultural areas of the country, files were created in shapefile format with databases of climatic variables such as mean annual precipitation and mean annual temperature, duration of the growth period by available humidity, average temperature of the warmest month, average temperature of the coldest month, soil texture, terrain slope, drainage condition, pH, salinity and sodicity as well as soil depth.

Agricultural areas with similar soil and climatic attributes were generated with the superimposition of the values of the indicated parameters. For each polygon with values of the climatic and soil variables, the ranges of the climatic and soil variables for each crop that define each aptitude class were included, and for each, their current aptitude label was generated.

For this work in the estimation of the degree of aptitude under climate change scenarios, it was decided to include only mean annual precipitation (PMA) and mean annual temperature (TMA), these considering the monthly mean rainfall and temperatures. The aptitude class was defined by the most limiting climatic parameters under the following combination described in Table 1.

Once the aptitude classes were defined, the ArcGis 10.3 Software generated the dynamic tables detailing the surface for each combination of the parameters that define the aptitude degree and the regional scale maps were prepared (1: 250 000) for each crop.

Climate change scenarios

The climate change scenarios obtained from ^{Fernández et al. (2015)} for three models: GFDL-CM3, HadGEM2-ES and REA; for the distant time horizon (2070-2099), in addition to two RCPs: 4.5 and 8.5 W m^-2. The models were taken from the page of the Center for Atmospheric Sciences of UNAM (^{UNIATMOS, 2018}). The rates of change for the monthly mean rainfall and the mean monthly temperatures for each agricultural area of the country were obtained and the mean annual precipitation and mean annual temperature were calculated from the monthly values for each of the climate change scenarios.

The procedure described under current conditions was applied to define the aptitude degree classes of agricultural land for each crop, obtaining the databases and their respective maps for each crop, that is, with each model, two dynamic tables were generated with the databases and their maps, one for RCP 4.5 and the other for RCP 8.5. With this information, the current aptitude changes in each climate change scenario for each crop were analyzed.

Grain corn (Zea mays L.) current conditions

Of the 32 429 966 ha delimited as for agricultural use (^{INEGI, 2017}), 81.59% have some degree of aptitude for corn under temporary conditions (Table 2), of which 10.91% have no limitations (suitable), 35.67% as moderately suitable and 35% as marginally suitable. The climatic parameters that represent restrictions for this crop and that define the agricultural areas as unsuitable, are in the first place the temperature.

As the only limitation or in combination with other variables with 10.82% of the total agricultural land, most of these areas have annual average temperatures below 12 °C and are located in the high areas of the Eje Transversal Mexicano and high areas of the Sierra Madre Occidental. Precipitation as a limitation for agricultural land to be suitable for this cultivation under temporary conditions, either alone or in combination with other variables represents 8.85% of total agricultural land.

These areas are associated with agricultural areas with rainfall less than 350 mm annually and some areas with average annual rainfall greater than 4 000 mm. Within agricultural areas with average annual rainfall of less than 350 mm, they correspond mostly to irrigated areas, although there are some seasonal agricultural areas that have this condition of low rainfall where, in addition to other crops, corn is also planted but with high accident rate.

Climate change scenarios

Table 3 shows the results of the degree of aptitude of the country's agricultural land for seasonal maize under climate change scenarios for a RCP of 4.5 and 8.5 for the distant time horizon (2 070 to 2 099) and applying the rates of change for the three general models of atmospheric circulation (MGCA).

In both RCP, the decrease in the surface with some degree of aptitude for this crop is considerable, going from 81.59% of the total agricultural land to 61.22%, 53.81% and 60.50% in the RCP 4.5 and to 63.32%, 60.66% and 61.23% in the RCP 8.5, for the GFDL, HADGEM and REA models, respectively. As can be seen, there are no major changes in the two RCP, even the surface is slightly larger with some degree of aptitude in the RCP 8.5 associated with the fact that a smaller decrease in precipitation is forecast than in the RCP 4.5.

The increase in the temperature estimated in the three models is the dominant factor in the exclusion of agricultural areas for this crop, where it is estimated that the average annual temperature in areas with some degree of aptitude under current conditions will be higher than 26 °C which is the limit to consider some degree of aptitude, the proportion of agricultural areas without aptitude for this crop associated with temperature alone or in combination.

With other variables such as precipitation less than 350 mm per year, it changes from 10.82% to 31.62%, 29.96% and 23.6% for the RCP 4.5 and 30.6%, 30.24% and 24.07% for the RCP 8.5 in the GFDL, HADGEM and REA, respectively. It is important to point out that, although some agricultural areas of the country that in current conditions have annual average temperatures below 12 °C and that in these scenarios their temperature will be higher than this threshold, so they will present some degree of aptitude for this crop.

However, the area that increases its annual average temperature to values above 26 °C is considerably larger. As can be seen, the proportion of the surface that becomes unsuitable is similar in the two RCP in each model. The proportion of the country’s agricultural areas that will not be suitable for this crop under temporary conditions derived mainly from the decrease in precipitation to values less than 350 mm per year, goes from 8.85% to 14.02%, 25.39% and 21.54% in the RCP 4.5 and 12.05%, 14.96% and 18.69% in the RCP 8.5, for the GFDL, HADGEM and REA models, respectively. In the three models in RCP 8.5 there is less affectation on the surface with some degree of aptitude, this associated with the fact that it is estimated that the decrease in precipitation in different regions of Mexico will be less severe than for RCP 4.5.

Grain sorghum (Sorghum bicolor L.) current conditions

Of the total agricultural land, 61.54% have some degree of aptitude for grain sorghum under seasonal conditions (Table 4). The climatic parameters that represent restrictions for this crop and that define the agricultural areas as unsuitable, are firstly precipitation as the only limitation or in combination with other variables, 28.16% of the total agricultural land, these areas present average annual precipitation less than 400 mm or greater than 1 800 mm.

The temperature as a limitation for agricultural land to be suitable for this crop under temporary conditions, either alone or in combination with precipitation, represents 23.2% of total agricultural land, these areas are associated with agricultural areas with lower annual average temperatures at 15 °C or above 28 °C. In some regions of the country where this irrigated crop is established, the average annual rainfall condition is not necessarily met.

Climate change scenarios

Table 5 shows the results of the degree of aptitude of the country’s agricultural land for grain sorghum in storm conditions for climate change scenarios for RCPs 4.5 and 8.5 for the distant time horizon (2070 to 2099) and applying the reasons for change for the three General Atmospheric Circulation Models (MGCA) of this study.

In RCP 4.5, it is highlighted that in the GFDL model, the surface with some degree of aptitude for this crop will increase, going from 61.54% to 65.58%, while for HADGEM and REA models, the surface with some degree of aptitude for the development of this crop decreases to 53.06% and 53.98%, respectively.

For RCP 8.5 the proportion of the surface of agricultural land with some degree of aptitude increases slightly with respect to the current conditions in the GFDL and HADGEM models with 63.23% and 62.14%, respectively, and decreases to 55.16% in the REA model. With respect to the results of RCP 4.5, the greatest change is estimated with the HADGEM model, which increases by almost 10% of the total agricultural area, while in the other two models the changes are less.

The proportion of the country’s agricultural areas that will not be suitable for this crop under temporary conditions derived mainly from the decrease in precipitation to values less than 400 mm per year, goes from 28.16% under current conditions to 23.9%, 38.31% and 38.39% in RCP 4.5 and 27.18%, 29.38% and 35.87% in RCP 8.5 for the GFDL, HADGEM and REA models, respectively.

In the GFDL model in the RCP 4.5 an increase in the surface that is in the precipitation range with some degree of aptitude is estimated, mainly due to the increase in the surface with annual average rainfall less than 1800 mm, while in the RCP 8.5 changing the surface within the precipitation range with some degree of aptitude is similar to current conditions.

In the HADGEM model for the RCP 4.5 there is a considerable decrease in the surface with the precipitation range with some degree of aptitude, while in the RCP 8.5 the precipitation increases with respect to the previous RCP, mainly in the areas that had lower values at 400 mm per year, to have a surface similar to current conditions. In the REA model in both RCPs there is a decrease in the surface with some degree of aptitude in the precipitation range defined for this crop.

The increase in temperature estimated in the three models reports an increase in areas with an average annual temperature greater than 15 °C, as well as an increase in areas with temperatures greater than 28 °C, so the proportion of agricultural areas without aptitude for this crop associated with temperature alone or in combination with other variables such as precipitation less than 400 mm per year or greater than 1 800 mm per year, it changes slightly from 23.2% in current conditions to 20.94%, 24.68% and 24.31% for the RCP 4.5 and 22.14%, 21.58% and 23.75% for the RCP 8.5, in the GFDL, HADGEM and REA models, respectively.

In the GFDL model there is an increase in the surface with some degree of aptitude for this parameter in the two RCPs, being higher in the RCP 4.5. For the HADGEM model, a slight reduction of the surface with some degree of aptitude is estimated in the RCP 4.5, while a slight increase is reported for the RCP 8.5. For the REA model the values for both RCPs are similar in the proportion of surface with some degree of aptitude.

Grain wheat (Triticum spp. L.) current conditions

Of the total agricultural land in the country, only 26.95% have some degree of aptitude for growing wheat under temporary conditions (Table 6). The climatic parameters that represent restrictions for this crop and that define the agricultural areas as unsuitable, are in the first place the annual average temperature greater than 18 °C, which alone or in combination with precipitation represents 71.79% of the total agricultural land.

Precipitation as a limiting factor for agricultural land to be unsuitable for this crop due to having values greater than 2 000 mm per year or less than 300 mm per year, alone or in combination with other factors, represents 32.34% of the country’s agricultural area. It should be noted that some of the wheat-producing regions, such as the areas of the coastal plain of the northern Pacific in Sonora, cultivation takes place in winter and under irrigation.

Climate change scenarios

Table 7 shows the results of the degree of aptitude of the country’s agricultural land for wheat in temporary conditions for the climate change scenarios for the RCPs of 4.5 and 8.5 for the distant time horizon (2070 to 2099) and applying the rates of change for the three general models of atmospheric circulation (MGCA) in this study.

In both RCP, the decrease in area with some degree of aptitude for this crop is considerable, going from current conditions of 26.95% of the total agricultural land to 15.59%, 17.25% and 16.62% in RCP 4.5 and 17.11%, 17.21% and 16.62% in the RCP 8.5, for the GFDL, HADGEM and REA models, respectively. As can be seen only in the GFDL model, there is a change between the two RCPs, with a smaller decrease in area with some degree of aptitude in RCP 8.5 compared to that of 4.5. In the other two models the results are similar for the two RCPs.

The increase in the temperature estimated in the three models is the dominant factor in the exclusion of agricultural areas for this crop, where it is estimated that the annual average temperature in areas with some degree of aptitude under current conditions will be higher than 18 °C, which is the limit to consider some degree of aptitude, the proportion of agricultural areas without aptitude for this crop associated with temperature alone or in combination with other variables such as precipitation less than 300 mm annually, changes from 71.79%.

In current conditions at 84.1%, 81.31% and 81.15% for RCP 4.5 and 81.53%, 82.31% and 81.22% for RCP 8.5, in the GFDL, HADGEM and REA models, respectively. As can be seen, the proportion of the surface that becomes unsuitable is similar in the two RCPs for the HADGEN and REA models, in the GFDL model it is greater in the RCP 4.5. The proportion of the country's agricultural areas that will not be suitable for this crop under storm conditions derived mainly from the decrease in precipitation to values less than 300 mm per year or increase.

In precipitation at more than 2 000 mm per year they go from 32.34% in current conditions to 39.50%, 46.77% and 46.15% in RCP 4.5 and to 41.93%, 41.64% and 44.98% in RCP 8.5, for the GFDL, HADGEM and REA, respectively. In the three models in the RCP 4.5 there is less affectation on the surface with some degree of aptitude.