Study area

The study area is located in three adjoining properties belonging to the ejido Gatos Güeros in Linares municipality, Nuevo León, between 25°07'55.2" N and 99°18'59.1" W, with an altitude of 225 m (Figure 1). The selected properties are located in the Subprovince of Llanuras y Lomeríos (Plains and Hills) whose name comes from the presence of plains interrupted by scattered hills, low slopes and soft conglomerates. The type of climate is BS₁ (h') semi-dry very warm-warm, with average annual rainfall of 600-700 mm, average annual temperature of 22.4 °C and that of the coldest month is greater than 18 °C; summer rains and the percentage of winter rain is from 5 % to 10.2 % of the annual total (^{García, 1981}).

Figure 1 Location of the study area. 

Vegetation is represented by the Tamaulipan Thorn Scrub, which is composed of dense and very diverse vegetation of shrub and tree plants, which are distinguished by a wide range of taxonomic groups with differences in growth habits, leaf longevity and phenology (^{González et al., 2007}). In the Subprovince, clear soils are predominant, that are classified like luvic, calcic and haplic Calcisols; these soils are present in all topoform systems but especially in the great alluvial plain (^{Inegi, 1986}). Calcisols are soils in which there is a substantial secondary accumulation of calcareous matter and they are found in arid and semi-arid environments, often associated with parental materials of this kind (^{IUSS, 2015}).

Land use systems and experimental plots

In the research site, three experimental plots with different soil uses were selected to evaluate soil texture, pH, bulk density, mechanical resistance to penetration, electrical conductivity, organic matter and organic carbon.

The first use of the soil, which corresponds to the plot of the Native Vegetation Area (AVN), contains native vegetation (Tamaulipan Thorny Scrub) dominated by bushes, dense and thorny, with a wide range of growth patterns, diverse life times of the leaves, textures and contrasting growth dynamics, as well as its taxonomy and phenology. This system has more than 100 years without human intervention.

The second land use is an Agricultural Area (AA); with seasonal maize cultivation in which only yunta is used and conservation agriculture practices are implemented; the plot is 60 years old.

The third land use was called the Grazing Area (ASP); corresponds to a production system with intensive livestock and grassland rotation and approximately 15 years ago it was changed from a temporary agricultural plot to Buffel grass (Cenchrus ciliaris Fig. & De Not.), Pretoria grass (Dichanthium annulatum (Forssk.) Stapf.) and African Star grass (Cynodon nlemfuensis Vanderyst).

The type of soil in the three land use systems investigated is Calcisol, of the Luvic Calcisol type according to the World Reference Base for Soil Resources Classification (^{IUSS, 2015}).

Soil sampling

In January 2017, four composite samples were collected at two depths (0-5 and 5-30 cm.), with four subsamples each until obtaining approximately 1 kg of soil. Sampling was randomized in each of the three land use systems: Native Vegetation Area (AVN), Agriculture Area (AA) and Grazing Area (ASP). The samples (n = 24) were taken to the Laboratorio de Suelos y Nutrición de bosques de la Facultad de Ciencias Forestales de la Universidad Autónoma de Nuevo León (Laboratory of Soils and Nutrition in the School of Forest Sciences of the Autonomous University of Nuevo Leon), where they were dried in the shade at room temperature, they were screened with 0.2 mm mesh and prepared for chemical analysis.

Physicochemical characterization of the soil

Soil characterization was based on the measurement of the following variables: organic matter (OM) was determined by the modified Walkley /Black method (^{Woerner, 1989}). A wet digestion was carried out with concentrated sulfuric acid (H₂SO₄) and carbon oxidation with 0.07M potassium dichromate (K₂Cr₂O₇), added to 0.5 g of soil, 25 mL of K₂Cr₂O₇ and 25 mL H₂SO₄ and titration of excess dichromate with ferrous sulphate 0.2M (FeSO₄ 7H₂0). With this procedure, the content of organic matter and organic carbon of the soil (OC) was obtained under the assumption that the former contains 58 % carbon (^{Castellanos et al., 2000}). The soil reaction (pH) was measured by the AS-23 method of NOM-021-RECNAT-2000 (^{Semarnat, 2002b}), by means of the soil suspension: CaCl₂ 0.01 M at a 1: 2 ratio, through an electrode of glass. Electric conductivity (EC) by rapid determination in soil-water suspension 1: 5 (^{Woerner, 1989}), both pH and EC were recorded in a Corning pH conductivity meter, model 542. Soil texture was defined by the procedure of Bouyoucos by the AS-09 method of NOM-021-RECNAT-2000 (^{Semarnat, 2002b}). The bulk density (DA) by the cylinder method (^{Woerner, 1989}) and the mechanical measurement resistance to penetration (MRP) (kg cm^-2) was assessed with a soil hardness meter Yamanaka type penetrometer (^{Tada, 1987}).

Statistical analysis

The variables of interest were analyzed statistically in a completely randomized design with factorial arrangement, by land uses (A, 3) and depth (B, 2) with four replications. Normal tests of Kolmogorov-Smirnov and homocedasticity of Levene were performed (^{Sokal and Rohlf, 2003}).

The original variables for pH, apparent density and mechanical resistance to penetration were used and the variables of clay, silt to square root and the variables organic matter, organic carbon and electrical conductivity were transformed to natural logarithm. The mean comparison was made with the Tukey test (p≤0.05) and the original variables were handled for the Spearman correlation analysis. All the statistical analyzes were carried out with the computer package Statistical Package for the Social Sciences, version 13.0 for Windows (^{SPSS, 2013}).

Organic matter

The Tukey test for OM showed only differences for the 0-5 cm depth, in the scrub and grassing land in particular, where the highest content is recorded; in contrast, the opposite happened in the agricultural system (Figure 2). In a Vertisol under three forms of land use (scrub, agriculture and secondary vegetation), ^{Llorente (2004)} identified that OC was different between the systems only for the first 20 cm depth, but was the same for the subsequent ones up to 70 cm. Their results also confirm that shrub concentrates the highest OC content and that agriculture, the lowest. In Turkey, ^{Gol (2009)} concluded that the conversion of natural forest to continuous cultivation caused statistically significant decrease in the OM content.

Means with different letters at the same depth are statistically different (Tukey p ≤ 0.05).

Figure 2 Mean values of soil organic matter content (%) at 0-5 and 5-30 cm depths for all three types of land use systems. 

There are significant differences between the types of land use in the content of organic matter and organic carbon for the 0-5 cm depth, but values are higher in AVN (4.20 and 2.43 %), compared to those of AA (1.19 and 0.69 %). The maximum accumulation of organic matter was in the 0-5 cm depth in AVN and ASP, with a tendency to decrease as depth increases. This difference in the distribution of organic matter in these two systems is mainly due to the fact that in AVN the contribution of natural materials occurs on the surface of the soil in the form of leaf litter, branches and other structures; while in the ASP the contribution of organic matter is produced from the roots of grasses within the upper part of the soil as rhizogenic humus (^{Baldock and Nelson, 2000}).

The rate of reduction of organic matter when cultivating an area depends fundamentally on the intensity of the management and the edaphic and climatic characteristics; the highest reduction rate occurs in the first 5 to 10 years (^{Llorente, 2004}). It was estimated that the decrease of OM in the 0-5 cm depth was 71.6 %; if these lands were cultivated for 60 years, and that originally it was scrub, on the one hand, and that the 15 years ago ASP was cultivated, it is feasible that a process of reaccumulation of organic matter has occurred, as it is proposed in ^{Johnson’s model (1995)}. In the 0-5 cm depth of ASP, a recovery of 21.2 % of OM and OC was estimated over a 15 year-period (2.08 and 1.21 %, respectively). In contrast, for the 5-30 cm depth, no reaccumulation occurred but instead a 2.7 % loss of both during the same period took place.

The decrease in OM was lower (20 %) for both types of land use change (agriculture or grazing land) for the 5-30 cm depth. These results agree with those of ^{Chandran (2009)}, who estimated that 20 to 40 % of organic matter is lost when the lands with native vegetation are converted into cropland; likewise, ^{Llorente (2004)} calculated 37 % of organic carbon that is lost when cultivating the soil that was previously occupied by submontane scrub. ^{Cantú and Yáñez (2017)} recorded OC losses from 2.5 % in grasslands to 65 % in the agriculture area with respect to scrub vegetation for Vertisol soils. These losses are caused mainly by tillage and by the small amount of waste that is reincorporated into the soil.

pH and Electric Conductivity

pH and electric conductivity are indicators of the quality of the soil and of multiple chemical, physical and biological properties that influence its fertility (^{Castellanos et al., 2000}). In general, soils with pH in a range of 5.8 to 7.5 are considered as desirable or less problematic for most crops (^{Rodríguez and Rodríguez, 2002}). The Calcisol registered averages of 7.3 to 8.0 with respect to the soil reaction (pH), which classifies it as moderately alkaline. The ANOVA for pH did not reveal differences between land use (FA); only for FB = depth, significant differences were confirmed (p≤0.01), and an increase of pH in the 5-30 cm depth for the three systems of land use; there are no significant differences in the FA * FB interaction (Table 3).

EC is very important since it is related to the absorption of nutriments (^{Salcedo et al., 2014}). The results obtained in the analysis of this variable showed significant differences for FA = land use (p≤0.01) and the FB = depth (p≤0.05); in contrast, there were no significant differences in the interaction of FA * FB (p> 0.05) (Table 3). The comparison of means for electric conductivity (EC) for each system of land use is illustrated in Figure 3. EC varied from 79.7 to 116.7.0 μS cm^-1 for the 0-5 cm depth; the values of the three types of land use presented a classification of low salinity, which suggests that there is no crop restriction, and there is no presence of salts that exert a negative impact on productivity (Figure 3).

Means with different letters at the same depth are statistically different (Tukey p ≤ 0.05).

Figure 3 Mean values of the soil electric conductivity at 0-5 and 5-30 cm depths for the three types of land use. 

In the AVN and ASP plots, a significant increase in electric conductivity was recorded with values of 155.8 and 255.6 μS cm^-1, respectively at 5-30 cm, while in the AA they were 73.4 μS cm^-1 for the same depth. The highest number of the two areas mentioned above may be due to the presence of feces of cattle in the ASP and the high content of organic matter in AVN; on the other hand, the area with the lowest numbers, AA, has no fertilizer or chemical product at all.

Apparent density and mechanical resistance to penetration

The simple ANOVA for BD showed no significant differences between the systems of land use, and the opposite occurred in the RMP (p ≤ 0.05). The first ranged from 1.24 to 1.32 g cm^-3, while the RMP ranges ranged between 0.78 and 3.1 kg cm^-2 (Figure 4). ^{Fernández et al. (2016)}, identified differences in bulk density between forests and other kinds of land use, and attributed the low content of organic matter to the greatest depth. Generally, an excessive compaction of the soil causes a decrease in the land quality by reducing the macroporosity of the soil and the amount of water and nutriments that are available for the plant roots of, which hinders its growth.

Means with different letters at the same depth are statistically different (Tukey p ≤ 0.05).

Figure 4 Mean values of mechanical resistance to penetration for the kinds of land use: AVN, AA and ASP. 

However, according to the evaluations proposed by ^{Woerner (1989)}, calcic soils show a very low level of compaction, due to the good state of conservation, the absence of trampling of cattle in scrub and the contribution of organic matter.

The highest average values of mechanical resistance to penetration are concentrated in ASP and AVN with values of 3.10 kg cm^-2 and 2.98 kg cm^-2, respectively. It should be noted that AA had the lowest depressed figures (0.78 kg cm^-2), with a ratio of almost 4 to 1 with respect to the other kinds of land use. These values can be explained, mainly in the case of ASP, by the trampling of the animals and in the agricultural area (AA), by the constant clearing of the land with yoke for sowing. ^{Mogollón (2012)}, in his study in La Sabana de Bogotá, observed critical compaction values of 20.3 kg cm^-2 in the first 9 cm of depth due to animal walking; the lower values of 18.5 kg cm^-2 in the first 30 cm of depth were due to the preparation labors for sowing, which break the compacted structures, and thus, diminish the mechanical resistance to penetration and bulk density.

Texture

The results of the ANOVA of the percentage variables of sand, clay and silt presented significant differences for the factor: FA = land use (p≤0.01), while in the FB factor = depth, only the clay registered significant differences (p < 0.05), and the interaction of both FA * FB factors (p> 0.05), did not show significant differences (Table 3). The results of the Tukey test (p <0.05) for each texture fraction at two depths are illustrated in Figure 5. Sand and silt only showed differences between uses for the 5-30 cm depth, where sand was the fraction with the highest content in AVN and for ASP the lime was the fraction with the highest and different contents from the other types of use.

Means with different letters at the same depth are statistically different (Tukey p ≤ 0.05).

Figure 5 Mean values (%) of a) sand, b) clay and c) silt at 0-5 and 5-30 cm depths for AVN, AA and ASP. 

Regarding clay, differences are recognized in both soil depths; AA stands out for the highest content in both, while ASP values are the lowest. With the triangle of textures proposed by the International Society of Soil Science (ISSS) (^{Mathieu and Pieltain, 1998}), it was determined that the soil for the depth 0-5 cm is of the loam type (C) for the AVN, clay silt loam (Crl) for AA and silty loam (Cl) for ASP; in depth 5-30 cm was loam type (C) for AVN, for AA silty clay (Rl) and for ASP silty loam (Cl).

The granulometric analysis for both depths allowed to determine the values as: very high in the proportion of silts (from 45 to 72.5 %), high in the clay (18.4 to 48.6 %) and low for the sand (6.3 to 31.9 %). The obtained numbers agree with what was consigned by ^{Miralles (2006)}, since for the profiles of Calcisol type franc clay and sandy loam textures were distinguished, mainly, and to a lesser extent franc and silty loam textures.

Some studies show that the use or management of the soil does not cause an alteration to the ecosystems in the proportion of mineral particles (^{Cruz-Ruiz et al., 2012}). However, it is observed that in the trawl in the most superficial layers of the soil (0-5 cm) of the Calcisol there is a greater proportion of sands, probably due to the water drag of the finer fractions of the soil from the upper layers, in an eluvial process and the iluviation of these clays at greater depth.

^{Yüksek et al. (2010)} identify significant differences in the contents of sand, silt and clays in different systems of use with respect to a natural forest in Turkey.

^{Dieckow et al. (2009)} mention that soil texture plays an important role in the reduction of carbon stores after the conversion of native vegetation to conventional farmland; it is less adverse in clay soils than in sandy soils, due to the interaction with the surface (organ-mineral interaction) and the physical protection of the internal aggregates.

Table 4 shows the results of the correlation analysis for the 0-5 cm depth. The electric conductivity had a positive and highly significant correlation with the silt content (r = 0.768). Organic matter registered significance in both positive and negative correlations with organic carbon (r = 1.000), clay content (r = -0.832), bulk density (r = -0.594) and mechanical resistance to penetration (r = 0.660). Likewise, sand had a negative and significant correlation with clay (r = -0.661). The pH, clay, silt and BD did not define any correlation for the 0-5 cm depth.

The results of the correlation analysis for the 5-30 cm depth indicate that electric conductivity was negatively correlated with the clays (r = -0.625) and positively with the pH (r = 0.588). Also, the correlation between clay and silt was negative (r = -0.838). Likewise, no correlation was detected with the sand (Table 5).