Study area

The research was conducted in an area (32, 919 ha) located in the northeast of MAO (17° 30’ and 17° 42’ N and 97° 00’ and 97° 25’ W) (Figure 1) forming part of the physiographic province of the Sierra Madre del Sur (Southern Sierra Madre) and the physiographic subprovinces Sierras Centrales of Oaxaca and Mixteca Alta. Its climate is temperate sub-humid and the sierra is the predominant topoform. The geology of the area is highly varied, though limestones (36 %) prevail, as well as basic extrusive igneous rocks (18 %) and limestone-shale association (8 %), in addition to other minor rocks on the surface such as sandstones (Rivas, 2008)².

Figure 1 Values of the soil quality indicators of the Mixteca Alta Oaxaqueña. 

Representative soils in the area are Leptosols, Pheozem and Regosoles (^{INEGI, 2007}; ^{IUSS Working Group WRB, 2014}).

Leptosols are thin, stony and poorly developed soils, which may contain a large amount of calcareous material. They are associated with sites of complex orography, are common in mountainous areas and in heavily eroded regions, while their agricultural potential is limited because of their shallow depth and abundant stoniness. The Pheozem are dark and rich in organic matter, they present limitations for agricultural activity as they are susceptible to wind and water erosion and periodic droughts. Regosols are formed on unconsolidated materials and are located within the most fertile soils; their main use is agricultural².

The vegetation of the area, according to the (^{INEGI, 2007}), is mostly induced pasture (25.9 %), wooded areas of oak, pine and pine-oak association (32 %).

The predominant crops are traditional subsistence ones: maize associated with beans or as monocultures; on a smaller scale, wheat, canary seed and barley are planted; maize in cajete (structure built to preserve water from summer rains) and winter wheat grown on residual moisture, irrigated maize and wheat; also, triticale, oats, alfalfa and sorghum. Both milpa intercropped with fruit trees (MIAF) and pastures help to retain the soil and prevent its loss (^{Plan Municipal de Desarrollo (Municipal Development Plan), San Juan Yucuita, Nochixtlán, Oaxaca, 2010}).

Soil sampling

Fifty seven sampling sites were selected (42 agricultural, 12 degraded and 3 forest sites) distributed in five communities of five municipalities of the zone: Tonaltepec (Santo Domingo Tonaltepec), Gavillera, (San Bartolo Soyaltepec), Cerro Prieto (San Miguel Chicahua), Nduayaco (Santiago Apoala) and Pericón (Santa María Apazco). The sites sampled were: 1) ten agricultural sites, four degraded in Tonaltepec; 2) ten agricultural sites, one degraded in Gavillera; 3) seven agricultural sites, three degraded and a forest in Pericón; 4) nine agricultural sites, two degraded in Nduayaco; 5) six agricultural sites, two degraded in Cerro Prieto. In addition, soil in three forest sites were also collected. The communities of each municipality were chosen based on three criteria: 1) belonging to the same basin (Papaloapan), 2) being of high and very high marginalization according to the National Council of Population 2010, and 3) having producers willing to cooperate. The selection did not follow a statistical model, but was rather decided by the researcher experience and the distribution of localities was decided to be directed according to the three criteria above indicated.

In each site we obtained a composite soil sample from a mix of six subsamples collected from the 0 to 30 cm depth, with a fertility sampling auger. The sample was reduced to approximately 2 kg by quartering. The soil was dried in the shade and at room temperature, then ground and sifted through a 2 mm mesh, to perform the chemical analyses in the laboratory. The sites from where we obtained the samples were georeferenced and the coordinates located on a thematic map of degradation. The forest soils were collected and analyzed to have a reference of the current conditions of them.

Chemical analysis

In the soil samples we determined: pH by potentiometry (soil-water ratio 1:2), organic matter (OM) by wet combustion with the Walkey and Black method (^{Jackson, 1976}) and extractable P Olsen (Pt_ext) (^{CSTPA, 1980}). The extraction of exchangeable K and Na was carried out with 1N pH7 ammonium acetate and 1N pH 8.2 sodium acetate for Ca and Mg to avoid the solubilization of free Ca carbonates, present in the soils of the area (^{Cano et al., 1984}, ^{Castellanos et al., 2000}). Exchangeable bases were determined by atomic absorption spectroscopy. The effective cation exchange capacity (ECEC) was estimated by adding the exchangeable bases (^{Cottenie, 1980}). The carbono (C) in the microbial biomass (C_BM) was determined by fumigation and extraction with chloroform (^{Joergensen and Brookes, 1990}). From here on these soil characteristics will be called attributes in order to maintain a terminology consistent with the soil quality articles in which this term is preferred. For each condition of the soil and attribute measured, we obtained the mean value of the repetitions. The mean comparison was made with the Tukey test (p≤0.05).

Indicators of soil quality

Soil fertility integrates physical, chemical and biological soil attributes (^{Pieri, 1989}). In this study, only chemical, physicochemical and biological attributes were used to define indicators and quality classes. The selected attributes were: pH in water (1:2), OM, P_ext Olsen, K, Ca and Mg exchangeable (K_int, Mg_int, Ca_int), ECEC and C in the microbial biomass. We selected these variables because they are frequently used to define soil fertility. To define the SQI and the associated values, we used the methodology proposed by ^{Cantú et al. (2007)}, for which it is necessary to define the maximum (I_max) and minimum (I_min) limits of the selected attributes. As there is no single way to establish these limits, agricultural and degraded soils were defined with the following criteria: theoretical concepts, type of soil, ideal soil conditions for potential crops in the study area, availability of nutrients, presence of carbonates, maximum yields according to current soil conditions, climatic conditions and the experience of researchers who have studied the area (Table 1).

Table 1 Attributes used as quality indicators of agricultural and degraded soils of the Mixteca Alta Oaxaqueña and limit values 

^††I_min: minimum value of the attribute; ^¶I_max: maximum value of the attribute; ^§P_ext: extractable phosphorus; ^ΦCa_int: exchangeable calcium; ^¤Mg_int: exchangeable magnesium; ^††K_int: exchangeable potassium; ^¶¶ECEC: effective cation exchange capacity; ^§§C_BM: carbon from microbial biomass.

The values of the indicators were normalized using a 0 to 1 scale. For the SQI these two values represent the worst and best condition from the point of view of quality, respectively (^{Cantú et al., 2007}). In agricultural and degraded soils, the maximum value of the I_max indicator (normalized value Vn = 1) represents the ideal value which is expected to be reached or the best soil quality situation (^{Cantú et al., 2007}). The minimum value of the same I_min (normalized value Vn = 0) represents the minimum desirable or the minimum acceptable quality. For forest soils, quality indicators were not calculated because they were only used as reference for undisturbed soils.

The calculation of the normalized value of the indicators was made in the following way:

where V_n: normalized value of the indicator; I_m: experimental measurement of the attribute considered as an indicator; I_min: minimum value of the attribute considered as an indicator; I_max: maximum value of the attribute considered as an indicator.

For the interpretation of the quality indicators generated for the soils of the five MAO municipalities, a modification of the scale recommended by ^{Cantú et al. (2007)} is proposed, which has only five quality classes. Our proposal considers seven quality classes for each indicator (Table 2), a class higher than one and another lower than zero, due to the wide range of values presented by the indicators generated for the MAO.

Table 2 Soil quality classes of the Mixteca Alta Oaxaqueña. 

Thematic map

One of the main functions of the indicators is to evaluate conditions or trends over time (^{Cantú et al., 2007}), for which the sites were georeferenced in a thematic map with the types of degradation for follow up changes in space and time. This meets one of the conditions required for soil quality indicators (^{Cantú et al., 2007}). The use of a degradation map was due to the fact that this process has an intimate connection with the loss of soil capacity for production and other services. To generate this map, we used the vector information available in the layer Evaluation of Soil Degradation Caused by Man in the Mexican Republic (^{SEMARNAT-CP, 2002}), scale 1:250 000, and the complementary document called Report on the Situation of the Environment in Mexico (^{SEMARNAT, 2012}). From a layer of municipal demarcations scale 1: 250 000 generated by the ^{CONABIO (2012)}, the municipalities corresponding to the study area were mined. Based on this new map and using the command Extract of ArcGis, version 10.1 (^{ESRI, 2013}), we extracted the polygons from the types of soil degradation that occur in the study area. To this composition a vector layer was superimposed with the points corresponding to the coordinates of the sampling sites. The thematic map of degradation generated indicates the type and main causes of degradation of the areas where the soils selected for the study are located. To verify whether the information generated by this map corresponded with that of the study sites, we checked it with photographs and observations in the field.

Chemical analysis of soils

To interpret the chemical attributes evaluated, we used the information proposed by ^{Castellanos et al. (2000)}. The agricultural MAO soils are from neutral to alkaline (pH 6.8 to 8.2). In the degraded soils of Tonaltepec, Pericón and Gavillera the pH was located in the neutral to slightly acidic range (pH 7.2 to 6.4) (Table 3). In the agricultural soils of this zone the OM varied from moderately low to medium (1.7 to 2.3 %) and in those degraded from moderately low to low (<1.4 %). In contrast, in Tonaltepec the OM was very low (< 0.7 % OM) but reached concentrations >4 % in the agricultural soils of Pericón, which is due to the contribution of organic matter to the soil (added residues and compost). The concentration of P_ext in the degraded soils was low (< 10 ppm P_ext Olsen), but the agricultural soils of Pericón, Nduayaco and Cerro Prieto showed higher presence of P_ext Olsen (10.4 to 20.6 ppm) (Table 3). The contents of Ca_int were high (> 10 cmol kg^-1) in both agricultural and degraded soils. The soils of the municipality of Cerro Prieto were the exception as the content of this cation was low (< 5 cmol kg^-1) in them.

Table 3 Attributes evaluated in the soils of the Mixteca Alta Oaxaqueña 

^¶¶¶¶ Means with different letter are statistically significant (p≤0.05); DMS: minimum significant difference.

^†OM: organic matter; ^¶P_ext: extractable phosphorus; ^§C_aint: exchangeable calcium; ^ΦMg_int: exchangeable magnesium; ^¤K_int: interchangeable potassium; ^††ECEC: effective cation exchange capacity; ^¶¶C_BM: carbon of the microbial biomass.

The concentrations of the other exchange bases were medium and low Mg_int (1 to 5 cmol kg^-1) and K_int (< 0.6 cmol kg^-1) (Table 3). The exception was the degraded soil of Pericón with 16.6 cmol kg^-1 of Mg_int, possibly related to the mineralogy of soils (^{Bautista et al., 2011}).

The ECEC values were defined mainly by the exchangeable Ca and Mg contents, which were high (25 to 41 cmol kg^-1) in Tonaltepec and Pericón, medium (13 to 25 cmol kg^-1) in Nduayaco and Gavillera, and low (4 to 8 cmol kg^-1) in Cerro Prieto (for agricultural and degraded soils). The significant decrease of C_BM in the degraded soils of Pericón, Nduayaco and Cerro Prieto (14.9 to 141.6 mg kg^-1) indicates that the OM used by microorganisms as a C source has been lost in these soils (cf. ^{Xu et al., 2013}), which is the result of the erosion present in the area. The increase of organic C in the soil was accompanied by an increase in C_BM, as observed in soils of different ecosystems and agricultural crops (^{Xu et al., 2013}). But there were not enough experimental data to establish a correlation.

The soils of the forests presented pH values from moderately acid to neutral (5.7 to 6.8), very high OM contents (4.4 to 18.5 %) and C_BM (312 to 1980 mg kg^-1) but low P_ext levels (2.2 at 3.7 ppm), as expected in forest soils. The exchangeable bases (Ca, Mg and K) and the ECEC were notably lower than in the agricultural and degraded sites, which is associated to an exchange complex composed of mostly organic matter, where the exchangeable bases must be low. The high content of Ca_int (42.8 cmol kg^-1) and therefore a high ECEC (49.5 cmol kg^-1) in the soil of Pericón forest denotes the presence of calcareous material more than a high cation exchange capacity.

Indicators of soil quality

The values of the indicator pH were close to 1 in most agricultural and degraded soils; this value is considered as optimum (pH ≈ I_max) for the development of potential crops in the area (^{Cantú et al., 2007}). This chemical attribute does not seem to present a problem in the current state of soil fertility. Only the agricultural sites of Tonaltepec and Pericón, with indicators of pH quality greater than 1 (i.e. 1.28 and 1.16, respectively), require corrective management actions to lower the soil pH and avoid problems of availability of Fe, Zn, Mn and Cu (Figure 1A). These values are attributed to the presence of high levels of exchangeable Ca, coming from the calcareous material in that area. The quality indicator pH was also used by ^{Bi et al. (2013)} in their study of soil types.

The OM and P_ext indicators were negative in the degraded soils of the five communities (Figures 1B and 1C). According to ^{Cantú et al. (2007)} values lower than I_min (less than zero) indicate the worst quality situation and report very low values for the soil organic carbon (SOC) indicator (0.18) in agricultural soils of Argentina. In addition, ^{Gómez et al. (1996)}, ^{Campos et al. (2007)}, ^{Reynolds et al. (2009)} and ^{Bi et al. (2013)} used the SOC indicator. ^{Gregorich et al. 1997}; ^{Lal et al., 1999}; ^{Govaerts et al., 2006} and ^{Ghaemi et al., 2014} also considered it a key indicator for soil quality. The negative values of these indicators indicate that OM and P_ext Olsen levels are lower than the minimum required for the development of crops in the area, so they must be corrected. This could be partially achieved through the addition of composted manures available in the area and crop residues.

The low concentrations of P_ext can be due to the fact that it is not usual to fertilize with phosphorus and to the restrictions regarding the high presence of calcareous material present in these soils. Therefore, to improve the use of phosphorus, the use of phosphate fertilizers with a low Ca/P ratio is recommended.

The value of the C_BM indicator was positive but low since it varied from 0.1 to 0.5 in agricultural soils (Figure 1D). These values are explained by the low concentration of OM in these soils, a condition that does not favor the development of microbial biomass. ^{Campos et al. (2007)}, ^{Bi et al. (2013)}) and ^{Liu et al. (2013)} have also used C_BM as an indicator of soil quality. An exception was the agricultural land of Pericón, where the indicators of OM, P_ext and C_BM were 0.5, 1.0 and 1.0. The values of 1 are considered high and are the result of the addition of compost to the soil for fertilization, as we found during field visits. The Ca_int indicators (0.90 and 0.94) (Figure 1E) and ECEC (1.06 and 0.96) (Figure 1F) of the agricultural soils of Tonaltepec and Pericón are considered as ideal conditions to maintain an adequate soil fertility. However, the potential availability of this Ca is limited by the formation of calcium carbonates, judging by the pH values in these soils (8.2 and 7.9, respectively) (^{Castellanos et al., 2000}). The majority of the soils in Tonaltepec correspond to Leptosols that are characterized by the presence of calcareous material. In Nduayaco and Gavillera, the value of the Ca_int indicator (0.5) is 50 % lower than the ideal value defined for the agricultural soils of the area, and the ECEC indicator is between 70 and 60 % lower than the ideal value. The values of the Ca_int and ECEC indicators were lower in the degraded soils than in the agricultural ones in both locations, which is a clear indication of a lower amount of Ca in the degraded soils. The negative values of the Ca_int and ECEC indicators in the agricultural and degraded soils of Cerro Prieto show that these soils require special attention to improve their quality and fertility. In this site the degraded soils also presented deficit conditions in terms of the exchangeable Mg and K contents (Figures 1G and 1H, respectively) as indicated by the negative values of the indicators of these variables. The most extreme case was the K_int indicator with values lower than 0.1 in agricultural soils, which is why it is 90% lower than the ideal value (Vn=1). This condition is worrisome because K as well as P are essential elements for crops. This defines the need to correct soil management and apply fertilizers with potassium.

The indicators that best defined the low fertility of the agricultural and degraded soils of the MAO were OM, P_ext, exchangeable Mg and K. Of these indicators, ^{Liu et al. (2013)} used Kint, and ^{Bi et al. (2013)} used P_dis; in addition, ^{Bautista et al. (2011)} reported that the Mg _int indicator was the best one to evaluate the quality of semiarid soils of Oaxaca, but they point out that the inclusion of this chemical attribute should be taken care of because its variations may be affected by the mineralogy of the soil. Other indicators selected by ^{Bautista et al. (2011)} are SOC, pH, C_BM, Mg_int, Ca_int, K_int, which explained most of the variability of the properties of semiarid soils. According to ^{Bautista et al. (2011)}, the SOC is a very sensitive SQI and is highly associated with farming systems. ^{Campos et al. (2007)} also selected the Mg_int and ECEC to monitor the quality of the soil in a transect of Cofre de Perote, Mexico.

Quality and degradation indicators

The degradation of the soils was confirmed in the field by the presence of gullies, channels or mass movement due to the removal of the soil (Figure 2). In Cerro Prieto and Nduayaco, the main cause of degradation has been associated with agricultural activity (Figure 2A and Figure 2B). In Pericón, the burning and removal of crop residues could decrease nutrients and organic matter (Figure 2C). On the other hand, in Tonaltepec (Figure 2D) degradation is considered to be the product of overgrazing and deforestation, since this municipality demands important quantities of firewood for the elaboration of mud crafts. Sixty two percent of the SQI of agricultural and degraded soils present low quality, very low, excess and deficit classes. This indicates that a considerable percentage of these soils do not have the fertility conditions required for the adequate development of potential MAO crops.

Figure 2 Sampling sites of the Mixteca Alta Oaxaqueña. (2A): Cerro Prieto, San Miguel Chicahua; (2B): Nduayaco, Santiago Apoala; (2C): Pericón, Santa María Apazco; (2D): Tonaltepec, Santo Domingo Tonaltepec. 

The location in the thematic map of degradation (Figure 3) of the quality classes associated with the values obtained for the MAO soil quality index allowed to define a relationship between the current state of soil degradation, the SQI values and the quality classes associated with them. The SQI values of the agricultural soils indicated deficient quality classes (SQI < 0), moderate, low and very low (SQI < 0.5), as well as very high (Vn=1) in the sites of Gavillera, Nduayaco and Tonaltepec. The sites of Gavillera were located on the map of degradation in areas with water erosion and land deformation, caused by overgrazing, agricultural activity and overexploitation of vegetation for domestic use (Figure 3). The Nduayaco and Tonaltepec sites were located in areas of water erosion with loss of surface soil; and degradation is attributed mainly to deforestation with removal of vegetation for domestic use, overgrazing and agricultural activity (Figure 3).

Figure 3 Location of sites and soil quality classes of the Mixteca Alta Oaxaqueña in the thematic degradation map. 

Information on the Pericón and Cerro Prieto sites is discussed separately. In Pericón, agricultural and degraded soils are located in areas that present degradation due to the decline of fertility and reduction of OM content, as a result of deforestation with vegetation removal and agricultural activity (Figure 3).

However, the values of the P_ext and C_BM indicators for the agricultural soil of this locality presented high values (SQI ~ 1), which defines a very high-quality class. This confirms that the management of agricultural soil comprising the addition of organic material is adequate. In Cerro Prieto, the SQI of agricultural and degraded soils are located in stable areas under normal conditions (Figure 3). However, in both conditions the values of the SQI were very low or negative, like the SQI of the degraded soils of Tonaltepec, Pericón, Nduayaco and Gavillera. This confers them very low quality and deficit classes, and therefore a great decline of fertility.

The previous results indicate that most of the evaluated sites coincided in the state of the soil degradation (thematic degradation map) and the values of SQI and quality classes. Only in Cerro Prieto and Pericón this coincidence was not clear. In both cases, the values of the SQI indicated different soil fertility conditions, which is deduced from the degradation map. The above is attributable to the scale of the documents used to generate the thematic degradation map (1: 250,000). Previous matches can be improved if larger scale maps are used.

However, the development of cartography at the regional level requires a considerable investment due to the large amount of field work that must be done. In addition, in the country there are considerable cartographic resources, which can be complemented with new information, as was the case with the current study. In this way, the use of current cartography can be expanded, which would allow to make recommendations at the regional level.