Introduction

Bean is the second most important crop after maize in Mexicans food, it has a surface of 1 590 876 ha, of which 85.5% is grown under rainfed conditions in the spring- summer cycle (SIACON, 2014 cycle), which makes it highly dependent on weather conditions that occur in each crop cycle. In addition of vulnerability to weather conditions, there is the uncertainty of the impact of climate change on yield and production of this legume in the years to come.

Currently climate change is one of the main problems facing agriculture, due to disturbances caused in weather patterns and therefore in agro-climatology related with the management of agricultural activities. Temperature rise caused by the increase in atmospheric concentration of greenhouse gases (GHG) (^{IPCC, 2013}), promotes among other things the drying in many regions due to increased evaporation (^{Woodhouse et al., 2010}) and modification of rainfall patterns (^{Duran et al., 2014}).

Recent studies have shown that temperature of the agricultural areas in Mexico have been increasing noticeably since the early nineties (^{Ruiz et al., 2010}; ^{Zarazúa-Villaseñor et al., 2011}), this increase in temperature brings changes in agro-climatic variables such as the accumulation of cold in the winter period (^{Medina-Garcia et al., 2011}). As in other countries, in Mexico there is a concern about climate change and its possible impacts on the primary production sector, due to there are few quantitative and documented studies focused to agricultural areas (^{Tejeda and Rodriguez, 2007}).

Moreover, as a result of the increase in greenhouse gases in the atmosphere, there is an increase in temperature that could have both positive and negative effects on productive potential of crops. An increase in temperature accelerates crop maturity, reduces the duration of leaf area and thus total water requirement to crop to maturity (^{Ojeda-Bustamante et al., 2011}; ^{Hatfield et al., 2011}).

Several studies have been developed to identify areas where crop production could be performed more likely to succeed and more profitable, also called productive potential (^{Diaz et al., 2008}; ^{Medina et al., 2009}; ^{Aguilar et al., 2010}; ^{Liu et al., 2011}); however, in few crops has been studied the effect of climate change on areas with productive potential.

Changes in weather patterns have profound effects on plant growth and productivity in the short term (^{Attipalli et al., 2010}). In Mexico there have been studies on the subject of climate change and its impact on agriculture, but few have analyzed in detail the effects on product systems, which limit the design of strategies for adaptation to climate change in this sector (^{Ruiz et al., 2011}).

The set of changes in agro-climatic indices will cause a reduction in the area with optimal growing conditions for corn production in the tropics, subtropics and transitional areas and increase of the surface very high valleys and high valleys. These results could be used in designing strategies to address climate change in corn production areas in Mexico (^{Ruiz et al., 2011}).

^{Tinoco-Rueda et al. (2011)} found similar results in the state of Jalisco, where in an agro-ecological zoning in future scenarios, the potential area to grow corn shows a decrease between 64 and 91% for the period 2041-2060. Similarly ^{Gonzalez et al. (2011)}, established the areas that currently have optimal agro-climatic conditions to grow bean in the state of Jalisco, as well as areas that will have optimum conditions in the decades from 2021-2030 and from 2041-2050 and selected varieties showing different adaptation capacity with current climate and climate change scenarios.

^{Ruiz et al. (2011)} and ^{Medina et al. (2011)} applied the results of the assemble model from ^{Magaña et al. (2007}) to obtain climate scenarios and its effects on corn production and in the accumulation of cold hours for apple, respectively.

The aim of this study was to estimate the impact of climate change on future climate scenarios, on bean potential areas, under rainfed in Mexico.

Materials and methods

The first part of the study consisted in the estimation of future climate scenarios, to do so, the monthly database of anomalies for maximum temperature, minimum temperature and average monthly precipitation obtained from the assembly model made up of 11 general circulation models (GCM) reduced in scale and calibrated (^{Walton et al., 2013}) selected for Mexico (BCC-CSM1-1, CCSM4, GISS-E2-R, HadGEM2-AO, HadGEM2-ES, IPSLCM5A-LR, MIROC-ESM-CHEM, MIROC- ESM, MIROC5, MRI-CGCM3, NorESM1-M) for representative concentration pathways (RCP) of greenhouse gases (GHGs) 4.5 and 8.5. All this information was obtained from data portal WorldClim global change with a resolution of 30 "arc. As reference climatology the period 1961-2010 was used.

To apply the results of downscaled GCM, in studies related to agriculture, including the productive potential, it is convenient to use a good resolution, so the image resolution used for climate generated by INIFAP 90 x 90 m, so that the results of the productive potential have enough detail for the application in decision making to long-term plans.

Subsequently, reference climatology which was worked in raster format and adding algebraically future anomalies resulting from assembly model; the maximum temperature, minimum temperature and monthly precipitation values for the years 2021 to 2080 were estimated. With the monthly data the spring-summer cycle (PV) (May-October) values for those years were obtained, then temperatures and average rainfall of spring-summer for climate scenarios 2021-2040, 2041-2060 and 2061-2080 were obtained, hereinafter referred to as climatology or years 2030, 2050 and 2070 respectively.

The second part of the study was to estimate the productive potential, which is based on agro-ecological requirements of the species (^{Medina et al., 1997}). Potentially productive areas to grow beans under rainfed conditions in the spring- summer cycle were obtained. Potential areas were obtained for base climatology and for three climate scenarios. To obtain the potential productive areas, temperature and precipitation information were collected from INIFAP, the information on land use 5 series and edaphologic scale 1: 250 000 from INEGI. IDRISI Selva and ArcGIS View 10.1 were used.

Finally, according to the results, some adaptation measures to climate scenarios studied are proposed.

Results and discussion

Table 1 shows the areas with potential production for bean under rainfed conditions in the spring-summer cycle, current climate conditions and for climatology 2030, 2050 and 2070, in two representative concentration pathways of greenhouse gases. The obtained surface potential is independent from the current agricultural land use, that is, not necessarily that surface is available to grow beans under rainfed.

Table 1 High and medium potential productive area of bean under rainfed PV cycle, under current climate and future climate scenarios 2030, 2050 and 2070 in RCP 4.5 and 8.5. 

In this table can be seen how the surface of high potential for bean under rainfed decreases as it moves forward in the years 2030, 2050 and 2070 in RCP 4.5, regarding to the average or current weather conditions, passing from 1 887 426 ha with current climate to 1 144 100 ha in 2070. Similarly the average productive potential decreases towards the future, from 6 319 789 ha in the current scenario to 4 738 986 ha in 2070 (Table 1).

Figure 2 Potential production of bean under rainfed PV, cycle under current average weather conditions and in climatologies 2030, 2050 and 2070 in the RCP 4.5. 

Similar behavior occurs in RCP 8.5, just that surface decrease is greater with respect to the current scenario, passing the high productive potential of 1 887 426 a 727 933 ha, a reduction of 61.4% of the surface and the average potential declines its surface from 6 319 789 to 2 048 765 ha, a percentage of 67.6% in 2070 (Figure 3). Similar to that found by ^{Tinoco-Rueda et al. (2011}) for maize in the state of Jalisco for the period 2041-2060.

Figure 3 Potential production of bean under rainfed PV, cycle under current average weather conditions and climatology 2030, 2050 and 2070 in RCP 8.5. 

Surface reduction with high and medium potential production was mainly due to increase in average temperature from May to October in different climate scenarios. This can be seen in Table 2, the average temperature of the current scenario on high potential areas is 18.9°C, while in the first two climates from RCP 4.5 is 20.0 and 21.0 °C; that is, there is an increase of 1.1 and 2.1 °C respectively; in the third climate the increase is 2.6 °C regarding reference climatology, this increase in temperature decreases the optimum conditions for beans growth. The other important factor in potential production of bean under rainfed is precipitation, but in the crop cycle spring-summer it decreases 19 mm in high and medium potential areas in 2070 in RCP 4.5, which does not have great influence on the reduction of potential; in RCP 8.5 there is a significant reduction in precipitation of up to 91 mm in 2070 in high potential areas and 84 mm at medium potential areas (Table 3). Together with the decrease in rainfall, the increase in temperatures will cause higher levels of evapotranspiration, due to which the bean crop will suffer most from lack of moisture in its water balance (^{Ruiz et al., 2011}).

Table 2 Average temperature from May to October on the areas with high and medium potential production of bean under rainfed PV cycle, obtained considering the current potential surface in the different scenarios and RCP. 

Dif. esc. actual= diferencia con respecto al escenario actual.

Table 3 Average rainfall from May to October of the areas with high and medium potential production of bean under rainfed PV cycle, obtained considering the current potential surface, in the different scenarios and RCP. 

Dif. esc. actual= diferencia con respecto al escenario actual.

Figure 4A shows the surface with high productive potential in some of the main producing states of bean in the country, for different climate scenarios, it can be seen that the effect of climate change is not the same in different regions of the country for bean production under rainfed. Overall the trend is downward on the surface with high potential, but in temperate regions the surface will increase in the future, such as the state of Mexico. In the medium potential there is a differentiated trend among states, in the main producing state (Zacatecas) the trend is slightly downward and in the state of Mexico where the surface with high potential increases, it also decreases the medium potential, as this will pass to high potential surface. In the states of Durango and Guanajuato, the medium potential increase first and then decreases (Figure 4B). Similar results were found by ^{Ruiz et al. (2011)} in corn producing areas in Mexico.

Figure 4 High (A) and medium (B) potential production surface of bean under rainfed PV cycle, under average weather conditions and climates from RCP 4.5, in different states. 

In RCP 8.5 (Figure 5A and 5B) show a similar behavior in high potential, but the downward trend is more pronounced, and in four of the five states with high potential decreases to almost zero hectares by year 2070. With medium potential happens something similar, decreasing in all cases reduced to about 200 thousand hectares. These areas with high and medium potential will pass to low potential.

Figure 5 High (A) and medium (B) potential production of bean under rainfed PV cycle, under current average weather conditions and climates from RCP 8.5, in different states. 

These results could be used in planning or designing strategies to address climate change in bean producing areas, as might be seeking new bean varieties that adapt to higher temperature and tolerant to low moisture availability conditions.

Conclusions

The high potential areas of bean under rainfed in Mexico will decrease as climate scenarios moves to the future, passing from 1 million 887 thousand hectares in the reference climatology to 1 million 121 thousand hectares in 2070 from RCP 4.5; surface with medium potential will increase in climatology 2030 and then decrease in the following years.

The results from this study can serve as a basis to design strategies to address climate change in bean producing areas under rainfed in Mexico, as could be the generation of new varieties that adapt to higher temperature and evapotranspiration.