NUTRITIONAL CHARACTERIZATION OF COCOA SOILS IN MEXICO † [CARACTERIZACIÓN NUTRIMENTAL DE SUELOS CACAOTEROS DE MÉXICO]

:


INTRODUCTION
Financial benefits derived from cocoa cultivation provide a good alternative for developing tropical and subtropical areas of Mexico and the world.In 2021, 52,993.9ha of cocoa were cultivated in Mexico, with the states of Tabasco and Chiapas being the most important due to their area and production (SIAP, 2022).It is estimated that more than 5,000 producers are growing the crop, with an average yield of 540 kg/ha and a production value of 1,176,811 pesos (SIAP, 2022).However, in the last fifteen years, production has decreased due to different factors, such as pests and diseases, nutrition, quality, and conventional soil management.Cocoa production in Mexico is carried out in different types of soils, such as Cambisols, Luvisols, Nitosols, and Andosols (Suárez- Venero et al., 2021).Bautista et al. (2004) established that the indicators of good quality soil should allow (a) an analysis of the current situation and identify critical points concerning sustainable development, (b) an analysis of possible impacts before an intervention, (c) monitoring of the impact of anthropogenic interventions; and (d) help determine whether the use of the resource is sustainable.
The cocoa agroecosystem in Mexico, as in all cocoa-producing areas in the world, presents problems in its production because it is unknown if the plantations were established on ideal soils and if the right varieties were used.Among several deficiencies in the management of the cocoa production system, constant soil nutrient removal throughout the harvest stands out (Villason and Olguera, 2020).It has been observed that, although this is a closed system where initially the crop was planted in fertile soils previously occupied by forest the soil has become impoverished and is currently not able to supply enough nutrients to the crop to achieve good yields (Hartemink, 2005).Regarding its yield and mineral fertilization, there are still doubts since, in some cases, fertilization has shown little effect; but this is because it is necessary to know about the specific nutritional requirements of the tree, as well as the availability of nutrients in the soil (Rodríguez, 1992).These nutrient deficiencies can be resolved through rational crop fertilization (Rodríguez et al., 2001) once the nutritional removal standards of the cocoa tree are known (Alonso et al., 2020).Moreover, for a specific site, the chemical analysis of the soil will allow the formulation of a rational mineral fertilization dose (González, et al., 2018).It is therefore imperative to know the factors that affect its growth and performance, one of them is mineral fertilization depending on the type of soil and its fertility status.Regarding the above, evaluating and understanding the soil's properties is necessary based on the cocoa agroecosystem's nutritional condition and requirements (Villason and Olguera, 2020).
Soil fertility is considered the most significant indicator for the sustainable production of agroecosystems this is fundamentally derived from its physical, chemical, and biological properties that reveal the current and future condition of its productivity (Hanks and Ritchie, 1991;Wibawa et al., 1993).For the chemical properties of the soil in the cocoa agroecosystem and its relationship with the values determined in the entities of Chiapas and Tabasco, this vital information should lead to determining the relationship of these characteristics between the types of soil present in the cocoa agroecosystem and also inferring its fertility condition (Wadt, 2005).Therefore, the soil data series allows for inferring ideal soil conditions in this agroecosystem (Bockheim, 2008).On the other hand, multivariate analysis allows us to reduce the number of variables to understand the variation of soils due to their nutritional status (Du et al., 2008).For this reason, this work aimed to characterize the soils under cocoa plantation in Chiapas and Tabasco, Mexico.

Site selection
The study was carried out in different cocoagrowing municipalities of Chiapas and Tabasco, Mexico.The properties were located where the representative cocoa plots over fifteen years old were selected, with a surface area of 1 hectare and this was the reason for selecting the soils.

Sampling of cocoa soils
Within the plots, at each sampling site, the soil surface was cleaned of leaves and other residues, and using a straight shovel 15 cm wide by 30 cm long with a sharp tip, a 30 cm x 30 cm stump was opened.x 30 cm deep, in each vine and from the walls facing the sun, the samples were collected from top to bottom and packaged in 30 cm x 40 cm plastic bags (Thong and Ng, 1978;de Oliveira and Valle, 1990;Wood and Lass, 2008).Each sample was identified by numerical order, locality, entity, and georeferenced (Njukeng and Baligar, 2016) (Supplementary 1).

Physical-chemical characterization of soils
Once in the laboratory, the samples were dried for eight days in outdoors, under the shade.Subsequently, the dried samples were sieved with a 2 mm mesh, and 800 g of each were weighed; their chemical and physical characteristics were determined (Okoffo et al., 2016).Phosphorus oxide (P2O5) and Potassium (K) were determined by atomic absorption; Nitrogen (N) and Nitric Oxide (NO) by visible light spectrophotometry with a colorimeter, pH by a potentiometer and Organic Matter (OM) by the method of Walkley and Black (1934).Other determinations: Calcium (Ca) and Magnesium (Mg) (by complexometry), Iron (Fe), Manganese (Mn), Zinc (Zn), Copper (Cu) (by Atomic Absorption), Sulfur (S) (turbidimetry) and Boron (B) (Colorimetric Method and Oxalic Acid).Toxic elements such as Chlorine (Cl) (turbidimetry or potentiometry), Bicarbonate (HCO3) (ICP), Carbonate (CO3) (by combustion) and Cationic Exchange Capacity (CIC) (Ammonium Acetate).The texture was determined by Bouyoucos's method (1962).

Statistical analysis
Multivariate analysis was used to analyze the data: principal component analysis (PCA) was carried out using the PRINCOMP procedure of SAS (2020), where the eigenvalues and eigenvectors were considered.The principal components 1, 2, and 3 were plotted on a Cartesian plane to observe the distribution of the samples.In addition, hierarchical cluster analysis was used, and through the squared Euclidean distance and the semipartial correlation coefficient, the groups of the different ecotypes were separated.The groups were compared using the Tukey test of means with an alpha=0.05.

Principal component analysis
According to the principal components analysis, 70.9% of the variation in cocoa soils in Mexico was explained with the first three components and 96% with the first five components (Table 1).
The variables that contributed the most to the variation within each principal component were for CP1 pH, NO3, Ca, Mg, Fe, and B; for CP2, the MO, N, Zn, Mn, and Cu; and CP3, the P2O5, K, and S. (Table 2).
The variables that contributed the most to the distribution of the cocoa soils of Chiapas and Tabasco were MO, N, NO, pH, Mg, and Ca (Figure 1A) and P2O5, S, Fe, and B (Figure 1B).
The nutritional content of Chiapas and Tabasco, Mexico's cocoa soils, shows a wide dispersion.The grouping tendency is mainly due to nutrient concentration rather than origin.In quadrant I, samples from Chiapas are mainly grouped, and their dispersion is due to the content of OM and N (Figure 2, Figure 1A).In quadrant II, the content of B and Fe groups them; in quadrant III, the content of P and Cu; and finally, in quadrant IV, the grouping is by pH, Mg, Ca, and NO (Figure 2, Figure 1A).
Figure 3 shows how the samples tend to concentrate in the center, resulting in little dispersion.In quadrant I, five samples are characterized by their higher content of Ca and Mg (Figure 3, Figure 1B).In quadrant II, two samples stand out for their P and S content (Figure 3, Figure 1B).

Hierarchical cluster analysis (HCJ)
According to the ACJ, three large groups and eight subgroups were formed (Figure 4).
The grouping of the soil samples was presented with a tendency to their origin rather than due to altitude or nutrient content.The general altitude of all samples ranged from 5 m to 516 m and, on average, 151.4 m.  , 2019).The other subgroups are within the recommended pH range for cocoa (Table 3).3).
The most significant sources of P come from the mineralization of phosphate rocks, which, when weathered, decompose and release phosphates.These are absorbed by plants, subsequently by animals through ingestion, and finally return to the soil to be transformed into orthophosphates, through the microorganisms present (Ruttenberg, 2003).K, for its part, is one of the elements most in demand in cocoa (Reetz, 2016).Subgroup IIa presented the highest values of P and K (207.5 and 810 kg/ha, respectively).The samples come from Escuintla and Huehuetán, Chiapas, and subgroup IIIc with the lowest values of P (5.46 kg/ha) with samples mainly from Tuxtla chico, Chiapas and subgroup Ia for K (191.53 kg/ha), with samples from Tabasco (Conduacán, Teapa and Cárdenas) and Chiapas (Pichucalco and Tecpatán) (Table 4).
The husk is the part that has a high requirement in K; this could be one of the reasons why these soils have had a lower presence of this element.The husk is usually discarded due to its potential to cause fungal diseases (Pascual-Cordova et al., 2017;Castillo et al., 2018).
Subgroup IIIa, where all samples are from Maravilla Tenejapa, Chiapas, presented the highest values of Ca (6671.5 ppm) and Mg (2409.3ppm), and subgroup IIa the lowest values of Ca (5.8 ppm) and Mg (1.3 ppm); two samples from Escuintla and Huehuetán, Chiapas (Table 4).Ca is an element usually found in high concentrations in cocoa plantations.This is usually a nutrient that is little required by the cocoa plant, so it is found in high quantities in certain regions (Pascual-Cordova et al., 2017;Singh et al., 2019).Mg is also usually found in high concentrations due to the contribution of this element, which is made by tree litter (Furcal-Beriguete, 2017).7).Finally, B is an element used to promote flowering and reduce the effect of some viral and fungal diseases in cocoa plants (Krauss and Soberanis, 2002).It has been reported that the foliar application of this element during the flowering of the cocoa plant favors tolerance to the cocoa swollen shoot virus disease (Kouadio et al., 2017) was highest in the IIc subgroup and lowest in IIIa with samples from Maravilla Tenejapa, Chiapas (Table 4).The highest altitude of the plots was for subgroup IIIa of Maravilla Tenejapa, Chiapas, and the lowest for Escuintla and Huehuetán (Table 4).
The characterization of the groups carried out using this methodology on the chemical properties and nutritional concentration of the soils of Chiapas and Tabasco, referenced with the standards reported in the specialized literature for the sustainable production of the cocoa agroecosystem in the world, confirm that its chemical characteristics of pH and MO, the reference standards show the following ideal values for this production system: pH (5.5-7.0) and MO (2.93-5.52).Likewise, concerning the macronutrients (N, P, K, Ca, Mg and S) determined in these soils, they are located within the ideal reference standards for the sustainable production of cocoa in these entities: N (105-280 kg/ha), P (32-48 kg/ha), K (234-938 kg/ka), Ca (800-3600 ppm), Mg (110-488 ppm) and S (>50 ppm), (Wessel, 1971;Ritung et al., 2007;Njukeng and Baligar (2016);PBI, 2015;McDonald, 1934).Also, these references indicate that the concentrations of micronutrients are ideal in the cocoa agroecosystem of Chiapas and Tabasco: Fe (19-45 ppm), Zn (3-12 ppm), Cu (0.4-1.80 ppm), Mn (0.5-2.20 ppm), B (0.16-0.90 ppm) and Al (9-133 ppm).Based on the results, except for some values on the chemical properties (pH and MO), as well as the macronutrients and micronutrients, of the soils in this agroecosystem, supported by this classification system within the framework of their pedological properties, and compared with reference standards, both systems confirm that the cocoa agroecosystem in Mexico maintains ideal conditions for its sustainable production.

CONCLUSIONS
The

Figure 1 .
Figure 1.A: Contribution of the variables for CP1 and CP2; B: Contribution for CP1 and CP3.

Figure 2 .
Figure 2. Dispersion of the 107 soil samples according to components CP1 and CP2.

Figure 3 .
Figure 3. Dispersion of the 107 soil samples according to components CP1 and CP3.

Table 2 . Eigenvectors and Pearson correlation of the original variables with the principal components.
(González-Gordon et al., 2018;Argüello et al.IIa, were 4 samples from Chiapas, the altitude ranged between 230 m and 516 m and on average 381 m; pH 7.3; MO 4.85%; NO 25.83 ppm; N 169.95 kg/ha; P 10.68 kg/ha; K 519.45 kg/ha; Ca 6671.5 ppm; Mg 2409.3 ppm; S 43.03 ppm; Fe 27.08 ppm; Zn 0.7 ppm; Mn 3.38 ppm; Cu 3.13 ppm; and B 0.1 ppm.G IIIb, were 14 samples from Chiapas, the altitude ranged between 84 m and 492 m and on average 284.1 m; pH 6.5; MO 4.42%; NO 10.93 ppm; N 155.26 kg/ha; P 19.45 kg/ha; K 475.77 kg/ha; Ca 3934.6 ppm; Mg 584.4 ppm; S 16.91 ppm; Fe 52.51 ppm; Zn 2.16 ppm; Mn 28.18 ppm; Cu 4.34 ppm; and B 0.1 ppm.Finally, G IIIc were 8 samples from Chiapas, the altitude ranged between 83 m and 500 m and on average 376.5 m;The pH is one of the main variables that indicate the general state of the soil; in this work, this parameter ranges between a pH of 5.3 to 7.3.When comparing the groups and subgroups, subgroup IIIa, where all samples come from Maravilla Tenejapa, Chiapas, presented the highest pH(7.3)andNO(25.83),abovewhat is recommended for cocoa and subgroup IIa from Escuintla and Huehuetán, Chiapas had the lowest pH (5.3) below the recommended ranges for cocoa.Which could limit the availability, absorption, concentration of nutrients, and yield in cocoa cultivation(Rosas- Patiño et al., 2021).On the other hand, pH values between 7 and 7.5 are related to high Cationic Exchange Capacity and, therefore, greater nutrient intake(González-Gordon et al., 2018;Argüello et al.

Table 3 . Comparison of means of 15 variables in 107 soil samples from Chiapas and Tabasco, México.
When analyzing the S concentration, subgroup IIa presented the highest values (328.49ppm) and IIIb the lowest (16.91 ppm), the latter with samples from Chiapas (Maravilla Tenejapa, Marquéz de Comillas, Pichucalco, Tecpatán and Salto de Water) (Table4).Regarding Fe, subgroup IIc, with samples from Chiapas, presented the highest value (104.02ppm)IIIa and c, and Ib the lowest value (Table6).Subgroup IIb, with samples from Tabasco and Chiapas, presented the highest value of Zn (4.97 ppm), and subgroups IIIa and Ib with the lowest value (0.7 ppm).The highest Mn value (28.18 ppm) was for subgroup IIIb and the lowest (3.38) for subgroup IIIa (Table4).
cocoa soils of Mexico present nutritional ranges primarily according to the requirements of cocoa.Likewise, the nutrients that systematically differentiate the cocoa soils of Mexico are pH, NO, Ca, Mg, Fe, B, MO, NI, Zn, Mn, Cu, P, K, and S.