Introduction

The availability of soil moisture is an agroclimatic component highly relevant to the practice of agriculture. Under rainfed conditions, the availability of soil moisture is a function of the water balance, which is determined by the amount of precipitation, water storage in soil and water losses through processes such as evapotranspiration, runoff and deep percolation (^{Ines et al, 2001}). In tropical and subtropical areas, the growing season for crops is basically determined by the length of available soil moisture, which depends on the texture and it goes from the beginning of the rainy season until the end of it, plus the number of days required for depletion (through crop evapotranspiration) the residual moisture stored in the soil (^{Eitzinger et al., 2004}).

Climate change alters patterns of temperature, cloudiness and precipitation and, consequently, also the patterns of evapotranspiration and available soil moisture (^{Hatfield et al., 2011}; ^{Ojeda et al., 2011}; ^{Ruiz et al., 2011}).

The change in precipitation has less effect on wet soil moisture basins, but in dry basins percent change in soil moisture levels may be higher than the percentage change in rain; this is ofgreat importance in clay, thin soils. Compared to precipitation, temperature increases alone have a negligible impact on runoff and soil moisture (^{Chiew et al., 1995}). When kept constant all the variables that influence the ETP and only temperature increases; the ETP reference increases 3.4% per Celsius degree (^{Kimball, 2007}).

Climate change may also increase rainfall in some regions, but this is usually accompanied by an increase in rainfall variability, which by interacting with higher temperatures and drying can lead to regional droughts (^{Izaurralde et al., 2011}).

In Mexico, one of the most important States for rainfed agriculture is Jalisco. Recent studies have shown that the temperature of the agricultural areas of Mexico has been increasing noticeably since the early 1990's of the last century (^{Zarazúa et al., 2011}; ^{Ruiz et al., 2014}.). There is evidence of the presence of climate change in Jalisco, Mexico, which show changes in regional rainfall patterns in recent decades (^{Ruiz and Regalado, 2014}); which in turn are causing a gradual decrease in the growing season (^{Ruiz et al., 2000a}; ^{Zarazúa et al., 2011}) and a reduction in surface potential crop production (^{Ruiz et al, 2000b}; ^{Zarazúa et al., 2011}).

The predicted for the twenty-first century, climate scenarios set out in general terms, an increase in temperature and a slight decrease in annual rainfall in growing areas (^{IPCC, 2007}; ^{Ruiz et al., 2011}). The predictions for the State of Jalisco also establish a sustained increase in temperature during the first five decades of this century, reaching an increase of1.5 to 2.5 °C, over the period 2051-2060 (^{Ruiz et al, 2011}; ^{Ruiz and Regalado, 2014}). The consequences ofthese changes are directly related to a less positive water balance in soils, and a reduction in the growing season in tropical and subtropical areas (^{Ojeda et al, 2011}; ^{Ruiz et al., 2011}).

The present and future changes in soil moisture in agricultural areas are already demanding the development of adaptation measures for agricultural systems, but one of the first details that should be known is the extent of these changes. Accordingly, the objective of this study was to estimate the effect of climate change on the seasonality and duration ofsoil moisture in rainfed farmland ofJalisco, using the model Newhall.

Materials and methods

Results and discussion

The results of the statistical analysis of temperature and annual precipitation from weather stations that correspond to the three textural groups are shown in Table 1. As can be seen, all climate data sets resulted in a normal distribution, even when it was analysed statistical different according to the number ofobservations ofthe three textural groups. The ANOVA reported no significant difference between groups for textural precipitation and temperature (Table 2).

Table 1 Basic statistics, normality and statistical value of Kolmogorov-Smirnov for normal annual precipitation data and temperature 1961-1990 at sites with coarse textured, medium and fine soil. 

Table 2 Analysis of variance for temperature and precipitation data from three sites textural 

The results of the F tests comparing variances and t test for comparison of means for the periods 1961-1990 versus 2040-2069 in relation to humidity parameters resulting from the simulation model Newhall 1.6, are shown in Table 3. As can be seen, six cases were identified in which the means of the parameters are statistically different, most with a significance p< 0.01. Three of these cases corresponded to sites with soil of medium texture and the other three sites with fine-textured soil. For sites with coarse textured cases not statistically significant for either test variance (F) and for the mean test (t) differences were identified.

Table 3 Results of test F and t test compared in the means from 1961 to 1990 versus 2040 to 2069 of various climatic parameters and soil moisture at sites of coarse texture, fine texture and medium texture. 

The values of the parameters analysed comparatively for the climatology 1961-1990 and 2040-2069, as can be seen in the Tables 4 (coarse textured sites), 5 (fine texture sites) and Table 6 (medium texture sites).

Table 4 Parameters of simulated soil moisture with the model Newhall 1.6 at sites with coarse texture, two climatologies: A: 1961-1990; B: 2040-2069. 

Pa= Annual average accumulated precipitation; Ta= Average anual temperature; etpa= anual cumulative potential evapotranspiration ; H= wet days in the year; MS= days measured dry in the year ; S= dry days in the year; EC= growing season.

In cases of significant variation in the average value of fine-textured sites, projecting average annual temperature, ETPA and length of the growing season, increasing the temperature 2.5 °C and ETPA21%, while the LEC dropped 10 days changing from 1961-1990 to 2040-2069 (Table 5). For significant change in variance reports the dry half days, going from a value of 173.5 to 784.7; ie, it would turn into a changing parameter in the future (2040-2069), which is consistent with previous reports on the increase in randomness of the parameters of precipitation and soil moisture in the presence of climate change (^{IPCC, 2007}).

Table 5 Parameters of modelled soil moisture at sites with fine-textured, two climatologies: A: 1961-1990; B: 2040-2069. 

With regard to cases of significant variation in the average value of medium-textured sites, highlight the Ta, ETPA, MS and LEC, increased 26.6% to the first, the second 2.7 °C, 15% MS decrease and decrease 27% of the LEC. Regarding the variance, significant changes are identified in the LEC with 240% decrease in variance to spend 1961 1990 to 2040-2069. This was due to the decrease of the highest values of LEC (Table 6).

Table 6 Parameters of modelled soil moisture at sites with medium texture, two climatologies: A: 1961-1990; B: 2040-2069. 

When analysing comprehensively the information in Tables 4, 5 and 6 can also be noted that in general the effect of climate change is manifested by the following trends: a) an increase of precipitation in areas of fine texture and texture mean; b) an increase in temperature in all textural type sites; c) increased evapotranspiration potential sites all textural classes; d) a decrease in the number of wet days during the year at all sites textural groups, this trend being more marked in places coarse texture; e) a decrease in the dry half days in the year and a half sites thick texture; f) an increase in the number of dry days per year in sites of all textural groups, this trend being more pronounced in sandy soils of medium texture; g) a decrease in the length of the growing season all sites textural classes, being more evident in the case of medium texture, where the average is equivalent to 54 days, highlighting the cases of Pihuamo, Saint Maria de los Angeles, Tapalpa,Atemajac, Tacotán, San Juan de Potreros, Corrinchis II, Presa Achimec, El Nogal, Bocas, Mazamitla, San Gaspar de los Reyes, Jilotlán, Colotlán and Ajojúcar, with a decrease in the LEC 129, 116, 116, 115, 114, 95,93,93, 86, 83, 76, 75, 69, 66 and 60 days respectively.

Decreasing of the LEC in the State of Jalisco based on the effects of climate change had already been reported in advance by ^{Ruiz et al. (2000}) and was attributed by the authors to the combination of a decrease in the precipitation of potential increased temperature due to the increase of evapotranspiration. In the context ofthis study, the decrease in the LEC explained directly by the decrease in the number of wet days and dry environment and by increasing the number ofdry days, and indirectly by increasing the potential evapotranspiration.

For coarse textured sites precipitation varies almost did not responded to the average value ofthe five sites (Table 4); for fine-textured sites (Table 5) and of medium texture (Table 6) increased precipitation on average, but this increase did not offset the increase in the ETP so that the resulting decrease was mentioned in the LEC at the sites of all textural classes. In this regard, ^{Chiew et al. (1995)} mentioned that, the change in precipitation has less effect on wet soil moisture basins, but in dry basins percent change in soil moisture levels may be higher than the percentage change in the rain, being of great importance in clay, thin soils.

^{Izaurralde et al. (2011}) mentioned that the increase in precipitation that can cause climate change is usually accompanied by an increase in rainfall variability, which by interacting with higher temperatures and drying can lead to regional drought. This is what seems to happen in fine-textured sites where precipitation would increase 8.7% to climate change from 2040 to 2069, but the variance of rainfall would increase by 33.6%.

The implications of the changes discussed above could bring crops currently produced under rainfed conditions in the State of Jalisco, this have to do with an increase in growth rate due to increased temperature (^{Ojeda et al., 2011}), an increase in water demand ofthe crop by increased evapotranspiration, but in turn may be diminishing culture water requirement due to shortening of the production cycle the effect of increased rate of plant development (^{Ruiz et al., 2011}).

Changes in the length ofthe growing season a measure will necessitate adaptation consisting in changing seed varieties or model change crops, in places where the decrease in LEC is quite drastic (^{Zarazúa et al., 2011}).

The increase in the rate of development of the crop and thereby reducing the production cycle shortening directly impact the reproductive phase of crops and therefore in performance (^{Conde et al., 2006}).

Conclusions

Climate change of the 2040-2069 period will impact the soil moisture conditions at sites of all kinds of soil texture. However, the most significant changes were observed in areas of medium texture, followed by fine-textured sites.

Climatic and soil moisture parameters that will be affected with statistical significance of climate change are temperature, potential evapotranspiration and length of the growing season. For temperature and potential evapotranspiration an increasing trend is identified, and the length ofthe growing season a declining trend, inferring that climate change will be negative for the availability of soil moisture for the practice of agriculture.

The decrease in the length of the growing season keeps correspondence with the decrease in the number of wet days and dry environment, and increasing the number ofdry days during the year.