Introduction

According to the ^{IPCC (2013)} climate change scenarios are coherent and consistent descriptions of how the Earth’s climate system may change in the future. The A1F1 scenario is grouped into the scenario family and evolutionary line A1. Describes a future world characterized by its technological emphasis, with the use of intensive energy sources of fossil origin. Global population growth and the consequent increase in food demand, as well as the current model(s) of food production, are indicators of the need to improve crop yields, better manage each time more scarce water resource, and become aware of the exhaustion of the productive capacity of the soil.

The territorial extension of Mexico and the influence that the oceanic masses and the georelief have on the climatic expression and the use of the ground, base the necessity to realize studies on how they influence the processes of change of land use. According to ^{Pielke et al. (2006)} changes in the landscape alter the spatial pattern of convective rains. A similar effect has been attributed to some physiographic factors such as elevation, slope, exposure, and terrain roughness, associated mainly with radiant flow, soil moisture retention capacity, seasonality of production cycles, accumulation of units heat and cold units and leaf area index.

The slope of the soil favors the runoff velocity, as well as the loss of soil by water erosion, and the upwelling of salts. Southern exposures receive a higher rate of radiant energy which generates warmer, generally drier, and higher evapotranspiration rates. According to ^{NOAA (2017)} land cover refers to how much area of a region is protected by forests, wetlands, grasslands, agriculture, and other types of land and water. Land use shows how the population makes use of geographic space-either for development, conservation, or both. The prospect of changes in land use is largely attributed to the demand for food and changes in the population’s diet.

Considered as the largest reservoir of biological diversity, a healthy soil with sufficient organic carbon content is essential to sustain food production, to ensure continuity in the recharge of water to the aquifer, and to regulate gas exchange processes towards the atmosphere. The stability of the soil is compromised by the salt content as it has been found strongly associated with its rate of physical, chemical and biological degradation. From a sustainable point of view, this dynamic process restricts the ecosystem’s ability to provide goods and services and ensure their functionality over a period of time (^{FAO, 2015}).

According to ^{CONAFOR (2010)}, forest degradation processes are faster than deforestation processes. The degradation of the soil in any of its forms is a global problem and whose immediate consequence is evident in the decrease in the productive capacity of the agricultural soil. The rate of soil acidification, although it is a natural process, may be encouraged as a result of inadequate production practices, e.g. removal of crop residues, increased use of chemical fertilizers, or inadequate water use irrigation.

These practices as a whole are the main causes of the decrease in the organic carbon content in the soil, modify its structure making it more vulnerable to wind and water erosion, increase in the content of salts and the reduction or even disappearance of the vegetation cover. According to the Special Climate Change Program (^{PECC, 2014}) future climatic conditions are associated with variations in soil degradation rate, increase in salinity of irrigated areas, increase in vegetation cover by fire, drought and landslides, change in the production systems and dynamics of pests and diseases, and water availability.

Soil salinity is a limiting factor to the production of biomass in open ecosystems and an excess in the soil of agricultural use causes a decrease in the yield of the crops. According to ^{Richards et al. (1954)} a saline soil is one that has an electrical conductivity value (EC) greater than 4 dS m^-1 at 25 ºC. The most common salts are: sodium chloride and sulfate, calcium and magnesium, with sodium and calcium being the dominant ions.

Production techniques applied in agricultural, forestry and livestock production systems are intrinsically linked to the effects of climate change. The future perspective of water use where the current scenario is that aquifers have a deficit between recharge/extraction, more frequent and intense occurrence of periods of drought, as well as extreme events of rain, wind and temperature, cause severe damages to crops and even completely disinherit them.

In this study, it is proposed to jointly use databases of soil salinity and compaction of satellite origin, to simulate their perspective of change according to the expected climatology for the year 2050 of the most catastrophic climate change scenario where the use of fossil fuels is favored. Through the

interpretative analysis of the response surfaces, it is intended, on the one hand, to differentiate if the management practices in the production systems can accelerate the deterioration of the soil, and on the other to evidence the influence of the natural or induced changes in land use in large-scale ecosystems such as the arid and semi-arid.

Materials and methods

Three data sources were used in grid format (grid), soil compaction index (IC) and soil salinity index (IS), both databases resulting from the project land degradation assessment in drylands (LADA), ^{Blancalani et al. (2013)}, and the A1F1 climate change scenario according to HadCM3 (Hadley Center Coupled Model, version 3) by 2050. The LADA database is 5 × 5 minutes arc (9 km × 9 km); the coverage of Mexico is reached with 24 958 centroids.

Using the kriging technique and the elevation model with resolution at 90 m, the soil index data were interpolated and the respective response surfaces were obtained (Figure 1).

Figure 1 LADA indices of soil compaction (left) and salinity (right). 

According to the IC value, four classes were proposed: class 1 (very low - IC < 5); the porous space allows the passage of water and the horizontal and vertical movement of the gases. Class 2 (Light - >5 IC <10), the upper horizon is firm - wet - and the pore space is fine in size with a few large vacuoles. Class 3 (moderate - >10 IC < 15); soil of firm consistency and moderate presence of pores, but very few. Class 4 (severe - IC > 15), massive structure and consistency from firm to extreme and very little or no pore space. The IS was classified according to the proposal of Shannon (1997) in four categories. Soils “non-saline” with electrical conductivity (CE) <4 dS m^-1. Soils “moderately saline”, >4 CE <16 dS m^-1. Soils “highly saline”, CE > 16 dS m^-1.

The HadCM3 maps (^{Collins et al., 2001}) are 2.5º × 3.75º (latitude by longitude) of spatial resolution. This correspondsto a regular distribution of 3 687 centroids, with which the continental shelf of the country is covered. In the scenario A1F1 to 2050 (intensive use of fossil fuel intensive), economic growth is projected very fast, declines in the birth rate, and rapid adoption of new and more efficient technologies. In this scenario the population prefers more their personal well-being and has little interest in the care of the environment. The expected global values for the scenario, as well as the previous years, are shown in Table 1.

From Figure 1, the general interpretation of the IC is from very mild (less than 5), to severe (greater than 20), the most compacted areas are agricultural areas. The IS ranges from non-saline to moderately saline soils, it is in agricultural regions where the high extremes of both indices are observed and are assumed to be due to production practices and irrigation water management.

The A1FI scenario groups up to twenty-one environmental indices. Among these, seven were chosen, which we consider to be of greater relevance for evaluating the response of land use and which address its dynamism from the agricultural, livestock and forestry perspectives: 1) annual temperature range; 2) daily rainfall in the wettest quarter; 3) daily rain in the driest month; 4) daily rainfall in the driest quarter; 5) daily rain in the coldest month; 6) daily rainfall in the coldest semester; and 7) daily rain in the wettest month.

According to ^{George et al. (2001)} the rains determine the beginning and end of the growing season of crops, while temperature, ecosystem productivity, crop yield, and to a large extent is responsible for outcrops of salts in the soil. In the selection criterion, the humidity factor, associated with rainfall events and temperature were the referents, since they represent on the one hand the climatic stability, biodiversity and the conformation of the biotic communities in the open ecosystems, the management and use of the water in agricultural production systems, the sustainability of ecosystems, and soil conservation.

A study of this scale on spatial variability/continuity of climatic indices necessarily demands the application of spatial analysis techniques. The semivariogram of each index will be obtained to evaluate the spatial correlation of variables and covariables as a function of distance and stochastic simulation. According to ^{Bohling (2005)}, averaging the squared differences of the variable tends to filter the influence of the mean variant, and according to ^{Rasmussen and Williams (2006)} makes the covariance matrix a crucial element as a predictor (Figure 2).

Figure 2 Typical representation of the covariance function. 

The semivariogram and the resulting response surface show the similarity between centroids (sampling sites), assuming that the points with input x and that are closer to each other, are more likely to be similar to their response value y. Three components are recognized (Figure 2): sill, range and nugget effect; plateau, range and nugget effect. The plateau value has two components: a discontinuity at the origin, often called a nugget effect, and a partial one, the plateau value is one where the value of the semivariation ceases to be representative. Range is the distance at which the model curve equals plateau. With regard to the covariance function, it is considered a scalar function of the correlation (Figure 2).

In order to explore the joint effect of the soil indices and the moisture and temperature of the A1F1 scenario, the response surfaces were obtained by the co-kriging method. This geostatistical technique interpolates the point data of the variable (mapping and curves) of an area. Although it is similar to other linear methods that averages the weight of the elements, this, the weight does not only depend on the distance, but includes direction and orientation of neighboring point data.

Results and discussion

For interpretive support of the response surfaces, a vector layer of the agricultural irrigation area was superimposed (AA-riego.shp).

Annual temperature range (RAT)

According to the American Meteorological Society the annual temperature range is the difference between the average temperature of the hottest month and the coldest month.

From Figure 3, a decreasing gradient in the RAT of up to 28 °C by latitude in the North-South direction would be expected. The response surface was classified into nine categories without a set range. Due to the proximity factor to the oceanic platform, the coastal regions of the central Pacific to the South Pacific, Caribbean Sea and Gulf of Mexico would have the lowest RAT (13-19 °C), while the geographic regions center (34-38 °C) and north (38-41) is where the highest RAT would happen. The caloric capacity of the ocean is decisive for interpreting Figure 3. The continental shelf warms faster than oceanic at daytime and cools faster at night.

Figure 3 Response surface. RAT (ºC). 

The opposite happens with the ocean platform that is heated and cooled relatively slower. This is the reason why the geographic areas near the coast have a RAT of moderate to low whereas in the regions remote to her they have a more extreme climate. According to ^{Pickard and Emery (1990)} the average temperature and salinity of oceanic water is 3.5 °C and 34.7, respectively. Among the expected effects is that the number of species better adapted to the temperature variation increases and that of those with less capacity to tolerate temperature (^{Acklerly et al. (2015}). The geographic space of the northern region is dominated by plant community’s characteristic of arid and semi-arid zones

^{Rzedowski (2006)} groups vegetation types in the arid and semi-arid regions in xerophilous scrub, chaparral, and grasslands (navajita and amacollado); however, the descriptive ecophysiological of the biotic communities, as well as the agroecological requirements of the crops transcend the objectives of the present work.

Daily rainfall expected for the coldest month

From Figure 4, it can be noted that the arid and semi - arid regions (Sonoran desert, Chihuahua desert and Baja California peninsula) are where rainfall of less than 2 mm would be expected. In the South Pacific coast region, daily rains of the order of 1.7 mm to 11 mm would be expected; this volume of rain does not per se represent an extreme event, which suggests a low probability of occurrence of torrential rains. The arid and semi-arid surface in Mexico covers more than 40% of the total territory.

Figure 4 Average daily rainfall (mm) for the coldest month of the season. 

Arid and semi -arid terrestrial ecosystems water is the master limiting factor and in the absence of surface runoff primary production is linearly related to rainfall (^{Pianka, 2012}). This relation of cold with light rains is relevant for open ecosystems where the growth of the plants is slower in comparison when the temperature is high. Given the prevalence of these meteorological conditions for a short period of time, seed germination and plant growth are more dynamic (George et al., 2006) and when rainfall affects the moisture content in the soil, surface temperature, evaporation and evapotranspiration (^{Cong and Brady, 2012}).

Average daily rainfall expected for the hottest month

In arid and semi-arid regions, rainfall events occur in the summer season and are not infrequent. Pulse events are important because they trigger biological activity (^{Huxman et al., 2014}) and the most obvious consequence is that they generate false outbreaks in seed germination in the soil.

The immediate sequel is that these germinated seeds will not survive the subsequent days once the reserve of soil moisture has been exhausted before the high temperature prevalence. In arid and semi-arid ecosystems this concurrence of early rainfall, high temperature and rapid decay is more detrimental to ephemeral species.

According to ^{Turner and Jones (1980)} and ^{Levitt (1980)}, these species are short-lived, which they complete before developing symptoms typical of severe water deficiency and do not possess morphological, physiological or biochemical mechanisms of resistance to drought. From Figure 5, in the Baja California Peninsula it is expected that homogeneous conditions of low probability of occurrence of rain in summer will occur. A major contraction in pulse events is expected, especially in the arid and semi-arid regions, the Bajío and the Central region.

Figure 5 Average daily rainfall (mm) expected in the summer. 

Average daily rainfall expected for the wetter semester

The rainy season is the transcendental period for ecosystem processes (carbon sequestration, plant growth, biotic composition) and therefore the factor that defines the spatial limits of land use. In agricultural regions, it identifies the beginning and end of the growing season in temporary areas, as well as the suitability of water use in irrigation areas, the volume of runoff, the recharge rate of the water table, as well as a number of anthropogenic activities related to ecotourism and ecosystem sustainability and climate change programs.

For the A1FI 2050 scenario, light rains are expected mostly in most of the territory, with two areas being the Baja California peninsula where the North-South latitudinal gradient (0.3 to 11 mm day^-1) is observed, the Pacific region from Sinaloa to Oaxaca and Chiapas with medium intensity rains (6.2 to 16 day^-1). Some high mountain areas in Veracruz, State of Mexico and Puebla are where the highest intensity events are expected to occur, with rains ranging from 16 to 22 mm day^-1. Except for the central regions of Sonora and south of Chihuahua, the Central and North regions expect light rains.

Figure 6 shows the heterogeneity in the spatial distribution of rainfall events. However, it is important to mention that the per map excludes the extreme event factor. ^{Huber and Gulledge (2011)} document in a trend analysis of extreme torrential rains, and that the increase is due to the warming of the atmosphere, the heat waves will become more humid with the consequent increase in the abiotic stress. ^{Medvigy and Beauliu (2012)} the erratic pattern in rainfall events should be monitored day-to-day because fluctuations in radiant flux affect the entire planet.

Figure 6 Spatial distribution of daily rainfall (mm) expected in the rainy season. 

Average daily rainfall expected for the driest quarter

The summer season in Mexico is distinctive because of: a) increase in the demand of water in specific in the industrial zones and of agriculture of irrigation; and b) increased demand for energy for comfort in anthropogenic activities. Mexico is of special scientific interest this season by the presence of the monsoon.

According to ^{Guido (2016)} the monsoon season is a condition of dramatic events, with heavy rains, regrowth of vegetation, strong winds, and a large number of heat strokes. The monsoon begins in May in the northern region. Increases evaporation in coastal areas (Gulf of Mexico and Gulf of California), and generates moisture conditions on the continental shelf, which will eventually produce rainfall. This perspective can be better identified in the flat areas of the country and contrast in high areas where average daily rainfall of up to 4.7 mm is expected.

From the interpretation of Figure 7, the highest rainfall intensity in summer is expected to occur in the coastal region of the Gulf of Mexico and the Yucatan peninsula, Tabasco and Chiapas uprising. On the other hand, the expected rainfall deficit would have its greatest effects in the agricultural areas of Sonora (Valle Yaqui- Mayo, Caborca and Valley of Mexicali), Sinaloa (all agricultural area) and Comarca Lagunera. The analysis of this response area is a call for attention to promote improvements in the practices of production systems that tend to be efficient in the use of water, as well as the selection of drought tolerant materials, as mitigation measures of climate change.

Figure 7 Spatial distribution of the average rainfall (mm) expected in the dry season. 

Average daily rainfall expected in the coldest quarter

The beginning of the winter season and its duration is important for the development of crops, especially fruit trees that demand a certain number of cold units to complete their cycle, the condition of soil moisture and its storage, the recharge of the water table, the distribution of species in open ecosystems (forest, rangeland, or grassland), as well as in the provision of environmental services (biomass production, carbon sequestration, water harvesting, and organic carbon formation, among others).

In Figure 8, a decrease in rainfall would be expected and arid and semi-arid ecosystems would be expected to be the most affected, resulting in fragmentation of the geographic landscape, where biotic communities would tend to migrate to more stable climate zones. This would favor the integration in the biotic community of new species and the disappearance of some others without sufficient time to adapt. In forest and livestock ecosystems, the low availability of water in the soil and the very likely increase in air temperature would affect their ability to provide environmental services

Figure 8 Expected forecast for rainfall average (mm) in winter time. 

Average daily rainfall expected in the wettest month

Strongly associated with the growing season and the demand for water by cultivated species, the information represented in Figure 9 is key to appropriate programs on climate change and food security. For open ecosystems, forecasting low-intensity rainfall events is key to species dominance. Heavy storms infiltrate more than light rains; however, recurrent events of light rain can infiltrate at greater depth than a single event, if the soil moisture storage capacity favors its accumulation between events. This happens when rainfall is greater than the evapotranspiration rate.

Figure 9 Average daily rainfall (mm) expected in the wettest month of the rainy season. 

In agricultural areas, some indirect effects that would cause light and frequent rains would be expected to be reflected in the migration, incidence, and population dynamics of pests and diseases (^{Villalobos and Retana, 2017}), soil moisture content and dynamics, among others. From Figure 9, rainfall of less than 6 mm / day is expected for the north central region of the country, including the Baja California Peninsula, up to 11 mm in the southern Pacific coastal region and up to 22 mm in specific regions of Veracruz and North of Puebla.

What is observed on the map is contradictory to the global trend that has documented the increase in the frequency and intensity of extreme rainfall events.

Conclusions

Climate change and its synergistic effects are a determining factor in defining the dynamism and spatial boundary of land use. This reality has been amply documented in this and other works. The most catastrophic climate change scenario A1F1 2050 has provided important evidence of how the confluence of climate and soil management factors are rapidly deteriorating already unstable ecosystems.

The best projections give evidence of an increase in water scarcity, light and sporadic and anticipated rains in the arid and semi-arid regions of Mexico, which precede the consequent fragmentation of the ecosystem where new species would be incorporated into the biotic communities and disappearance of those with reduced adaptability to prevailing conditions.

The use of agricultural land would require a greater volume of irrigation water with the consequent exploitation of an already overexploited water table, while the areas producing rainforest would be under the influence of an increasingly erratic rainfall regime. Soil stands as an essential strategic resource on what, unlike climatic factors, there is no awareness of its degradation