Soil acidity

The average pH value was 5.2 (±0.9) indicating the presence of a generalized acidity in the soils. In 64% (n= 125) of the cases, the pH was below 5.3, considered as suitable for the presence of Al⁺³ in the soil solution (^{Espinosa and Molina, 1999}). The Vertisol soil stood out for presenting the highest average pH value with 6.2 (±1.5) and also the highest variability with a CV of 24.1% (Table 2).

According to the Official Mexican Standard NOM-021-SEMARNAT-2000, 47.4% of the sites are classified as strongly acidic and 44.4% as moderately acidic, that is, of the total of the analyzed samples, 92% (n= 180) present problems of acidity, and of these, 52% are strongly acidic with values less than 5. With the exception of Vertisol, in the other types of soils more than 90% of the analyzed sites were located in the strongly acid and moderately acidic categories. Acid soils are characterized by toxicity by Al⁺³, manganese toxicity and deficiency of Ca⁺² or Mg⁺² (^{Tasistro, 2012}), particularly, the presence of aluminum (Al⁺³) in the soil solution occurs when the pH is less than 5.3 (^{Espinosa and Molina, 1999}).

In 88% (n= 208) of the analyzed samples the presence of interchangeable acidity (AInt) was found, defined by ^{Espinosa and Molina (1999)} as the H and Al⁺³ that is retained in the soil colloids by electrostatic forces. The average was 0.96 meq 100 g^-1 (±0.75) with a CV of 78% indicating high variability among the analyzed sites. The possible damages caused by the acidity to the plants were analyzed from the criteria of the AInt content in the soil and the (%) saturation of the CIC by the Al⁺³ (Table 3).

In 57% of the sites, the AInt is above 0.5 meq 100 g^-1, considered the limit for the plants not to be affected (^{Espinosa and Molina, 1999}). For the case of the (%) saturation of Al⁺³ in the CIC, three interpretation criteria are analyzed: the first one indicates that 84% of the sites present values below 25% considered desirable for most of the plants (^{Espinosa and Molina, 1999}); the second, indicates that 55% of the sites present values above 5% considered problematic for the plants (^{Sánchez, 2014}) and the third, indicates that only 16 of the sites present values above 25% considered problematic for the corn cultivation (^{Bertsch, 1998}).

Organic material

In the 236 sites the average content of organic matter (OM) was 1.7% (±1.3) with a coefficient of variation of 79% indicating variability in the studied soils. Based on the fact that most of the studied area is adjacent to the Sierra Madre de Chiapas, originated by volcanic igneous rocks, the soils were considered of volcanic origin (^{DOF, 2016}), 93% of the 215 analyzed sites were located in the category of very low OM content with values lower than 4%, in Cambisol, Acrisol, Planosol and Feozem soils 100% of the sites were located in this class.

Figure 1 shows that in any soil type the OM content was higher than 2.5% and with the exception of Vertisol and Fluvisol, in all soil types the average content was below 2%. In 59.3 of the sites the OM content was ≤1.5% considered a critical limit even for non-volcanic soils.

Figure 1 Content of organic matter by soil type. 

According to ^{López et al. (2018)}, the very low OM content in soils is a factor closely related to the way in which the corn-livestock system practiced by 71% of producers has been worked for years. The practices of cultivation (burning, tillage and grazing of waste) and the non-incorporation of other source of OM, has generated an insufficient recycling of organic waste, which has brought as a consequence, a progressive deterioration of the soil that demonstrates an unsustainable management of the land, in virtue that OM is related to all aspects of soil, such as stability of the structure, ability to retain water, aeration, content and availability of nutrients, pH, cation exchange capacity (^{Benzing, 2001}; ^{Carter, 2002}).

The cation exchange capacity (CIC)

The average value of the CIC in the 236 analyzed sites was 11.4 (±16.8) meq 100 g^-1, with a high variability expressed by the CV of 147%. In all types of soils (with the exception of Vertisol), of the total analyzed sites more than 85% were placed in the low and very low content categories with values below 15 meq 100 g^-1. In Vertisol only 48% of the cases were placed in these categories and the rest in the very high content class with values higher than 40 meq 100 g^-1.

The fact that more than 85% of the analyzed samples show low CIC, indicates that the soils have limited capacity to retain cations (nutrients) due to their sandy texture and low content of organic matter; under these conditions, in addition to the limited capacity to store readily available K, Mg and Ca, these cations are unprotected against the leaching process (^{Benzing, 2001}; ^{Arcila and Farfán, 2010}).

Figure 2 shows that 65% of the behavior of the CIC is explained by the acidity level of the soil; as the pH increases, new negative charges are generated in the exchange complex and is reported by ^{Anzorena (1995)}; ^{Benzing (2001)}, as variable CIC, which is more common in the CIC coming from the OM, than from the clays. ^{Foth and Ellis (1996)} point out that, in mineral soils with low organic matter content, the influence of pH on the CIC is minimal.

Figure 2 Influence of pH on the CIC. 

On the other hand, the OM explains 32% (R²= 0.3816) of the behavior of the CIC while (%) of clay 68% (R²= 0.684). Figure 3 shows the joint influence of these two variables on the CIC, the larger points represent greater value of the CIC and vice versa. It is observed that most of the soils are located in the lower left quadrant, in which the values of OM and (%) of clay are lower and, therefore, also have the lowest CIC values.

Figure 3 Influence of clay content (%) and OM (%) on the CIC. 

In the lower right quadrant, there are soils with high CIC value due to the higher clay content since their OM content is low. In the upper right quadrant, soils with high CIC are observed due to their higher OM and clay content. It highlights the fact that in the left upper quadrant practically no site was located due to the low OM content that most soils have. Based on the data shown in Table 4, the plots studied were grouped into 4 main subgroups.

Subgroup 2: represents 33% (n= 77) of the analyzed sites. They contain very low OM contents (<4%) and clay between 15-25%, in which the CIC average is 6.81 meq 100 g^-1. It is the second group of soils cultivated with corn.

Subgroup 3: represents 8.9% of the sites analyzed. They contain very low contents of OM (<4%) and clay between (25-35%), in which the CIC average is 11.63 meq 100 g^-1.

Subgroup 4: represents 6.3% of the analyzed sites. They contain very low OM contents (<4%) and clay greater than 35%, in which the CIC average is 48.9 meq 100 g^-1.

These four subgroups together represent 94% (222 of 236) of the analyzed sites that are characterized by having very low OM levels, in which the CIC depends significantly on the clay content they have. Increasing clay levels also increases the capacity of soils to store nutrients, for example, soils with more than 35% clay have on average a CIC 10 times higher compared to those with less than 15% clay.

Figure 4 shows that Vertisols contain the highest values of clay with an average of 43.8% (± 26.7%), therefore 60% of the cases that have the highest values of CIC (>40 meq 100 g^-1) belong to this type of soil where only 14% of the corn area is located. According to ^{Roy and Barde (1962)}, Vertisols have a high CIC due to the predominance of type 2:1 clay, for example, a soil with 50-60% clay has a CIC of 50-66 meq per 100 g of soil. According to these authors, the montmorillonite clay content gives an approximate estimate of CIC because 1 g of clay imparts on 0.5 to 0.7 meq of CIC to the soil and the contribution of organic matter to the CIC is quite low in the tropical zone (^{Roy and Barde, 1962}).

Figure 4 Clay content (%) according to soil type (n= 216). 

On the other hand, the soils of the Luvisol and Cambisol groups that are the most representative of the maize areas, 34.5% contain on average 21.5% clay (±16.4%). In the rest of the soil units, the average content is below 20%.

The content of phosphorus (P)

The average content of the 129 sites analyzed was 26.1 ppm (±22.8), with a CV of 87.1%, that indicates high variability among the studied soils, so it is recommended in subsequent studies to map the areas according to their content. Figure 5 shows the values of P and its location with respect to the critical level of 15 ppm considered critical by NOM-021-SEMARNAT-2000 and IPNI (2017).

Figure 5 Content of P in the sampled sites. 

Only 40% of the cases could be expected response to the application of phosphate fertilizer because they have contents below the critical limit, 60% remaining presented medium and high levels of P.

The high variability in the phosphorus content found in the soils is due to the fact that of the total of the producers that fertilize the corn crop, only 43% apply phosphate fertilizers in doses ranging between 6 and 69 kg ha^-1.

The basic cations (K, Ca, Mg, Na)

The content of these cations showed high variability in the soils analyzed with CV of 182, 110, 112 and 191% for Ca, K, Mg and Na, respectively (Table 1). Because only 17% (n= 40) of the analyzed sites found Na presence and that in 97 of the cases the (%) saturation of the CIC was less than 5% (considered optimal for the plants), this cation was excluded from the subsequent analyzes. According to the NOM-021-SEMARNAT-2000 criteria, 68, 58 and 58% of the analyzed sites were located in the low and very low content classes for Ca, Mg and K respectively.

a) The % saturation of Ca in the CIC

Table 5 shows a classification by type of soil according to the criteria indicated by the ^{IPNI (1990)} for 216 analyzed sites, in general 57.41% of the cases were deficient when presenting levels below 65% of saturation considered as optimal level.

The Vertisol presented the lowest cases of deficiency in the (%) of saturation of Ca with only 26%, while in the rest they were above 54%, highlighting the Cambisol with 78%. In deficient soils it is necessary to apply amendments (cal) to raise the saturation (%) to its optimum level of 65% (^{Sánchez, 2014}).

b) The % Mg saturation in the CIC

Of the 216 sites analyzed, 53% were located in the range of 10-20% of saturation considered optimal, 30% in high levels with values greater than 30% and only 17% were placed in the category of deficient with values lower than 10% (Table 6); Vertisol and Luvisol soils stand out in this last category. In all these deficient soils it is necessary to apply Mg amendments to raise the saturation (%) to its optimum level of 10%.

c) The (%) saturation of K in the CIC

In the case of this cation, 58% of the sites analyzed were in the deficient category with values below 5% saturation. Several studies have shown that K is an important element to increase grain yield and reduce corn stem lodging, its deficiency influences the magnitude of the photosynthetic rate and premature senescence of plants (^{Smith and White, 1988}), as well as in the progressive deterioration of the three basal knots of the corn stem, and the consequent lodging of the plant (Liebhardth and Murdock, 1965; ^{Kayode, 1986}).

Although all soil types showed saturation deficiencies of the CIC by K, the highest frequency of sites was observed in Vertisol, Feozem, Planosol and Cambisol soils (Figure 6).

Figure 6 Frequency of cases with deficiency in the % saturation of K. 

d) Cationic relations

Cationic relations are another criterion of support in the interpretation of soil analyzes, particularly among the bases (^{Chávez, 2012}). Taking as a reference the general critical values suggested by ^{Bertsch (1998)}, in 70, 34 and 29% of the cases optimal values were found in the Ca/Mg (2-6), Ca/K (10-15) and Ca+M /K (20-30) respectively. In the case of the relationships involving K, the lowest percentage of cases with optimal values is due to the fact that this cation presents low contents in 57% of the plots studied.

The micronutrients (B, Cu, Fe, Mn and Zn)

100 and 94.6% of the cases presented adequate levels of Iron and Copper respectively. Regarding Zinc, 46% were placed in the categories of deficient and marginal (Table 7).

The most poor micronutrient in soils cultivated with corn was Boron, since 90% of the cases analyzed (n= 80), were located within the categories of very low and low content (Table 8).

The deficiency of B in the soils may be due to coarse texture, poor organic matter, its formation on igneous rocks and leaching by rain, especially in acid and coarse-textured soils (^{Violic, 2001}).

Analysis by soil type

Figure 6 clearly shows that the plots located in soils of the Vertisol class have the best fertility conditions, since they have higher values of pH, OM, clay and CIC; however, in this type of soil only 14% of the maize area is sown. Due to its capacity to store nutrients (CIC) and clay content, it is followed by Luvisol type soils, in which an estimated 34.5% of the area cultivated with corn is located.

Figure 6 Average characteristics by soil type. 

In the rest of the types of soils (Cambisol, Acrisol, Planosol and Feozem) that represent 51.1% of the area cultivated with corn, the situation is critical in terms of its productive capacity, due to the low levels of CIC, OM and clay.