Introduction

When pesticides are applied as aerial sprays and powders to the foliage, large amounts of them reach the soil, which acts as a reservoir of these persistent chemicals, from which they move towards the body of terrestrial invertebrates (ants and beetles), passing by air or water. It also mentions that the crop decreases the amount of organochlorinated insecticides volatilized from the soil.

Under the surface, the pesticide may spread to the surface or be transported by soil moisture before it can escape into the atmosphere, causing long-term environmental damage, including soil degradation, irreversible contamination of groundwater, inland and coastal waters, accumulation of pesticides in trophic chains and the development of resistance in organisms such as insects, fungi or rodents (^{Edwards, 1976}; ^{Miglioranza et al., 1999}; ^{Dan et al., 2001}; ^{Barry and Warren 2004}; ^{Vives et al., 2005}; ^{Albert et al., 1989}; ^{Jimenez et al., 2005}; ^{Helberg et al., 2005}; ^{Olafsdottir et al., 2005}; ^{Thomas et al., 2012}; ^{Morales et al., 2017}; ^{Varol et al., 2017}).

Some researchers have conducted studies to determine soil characteristics involved in the diffusion, adsorption and retention of pesticides in the soil. In this regard, ^{Ehlers et al. (1986)} worked on a loamy loam soil to observe the spread of lindane in the soil. They found that the diffusion of lindane is influenced by soil water content, bulk density and soil temperature. On the other hand, soil adsorption and retention of herbicides is influenced by the cation exchange capacity, pH, organic matter, clay content and, to a lesser degree, by soil temperature and soil moisture content (^{Sullivan and Felbeck, 1967}; ^{Yu et al., 2013}; ^{Cederlund et al., 2017}).

Regarding the movement and persistence of pesticides in the soil, reported results are variable, ^{Voerman and Besemer (1986)} determined dieldrin, lindane, DDT and parathion residues in a sandy soil, after repeated applications over a period of 15 years. The results found in soil samples analyzed in layers of 10 cm to 60 cm depth indicated that DDT and díeldrin were much more persistent than lindane. On the contrary, the parathion disappeared very soon. Traces of dieldrin and DDT were found below 20 cm.

Also, ^{Jimenez and Arias (1992)} studied the movement and persistence of atrazine in Luvisols of the Fraylesca, Chiapas, Mexico, applying simulated rainfall to small lysimeters. The results showed that for the low dose of atrazine (1.5 kg ha^-1), the highest concentration was located in the first stratum (2.5 cm deep); However, for the high dose (3 kg ha^-1), the highest concentration was found in the second soil layer (5 cm depth).

^{Harner et al. (1999)} conducted a study of 36 Alabama agricultural soils to determine organochlorine pesticide residues. Compounds determined comprised the alpha and gamma hexachlorocyclohexanes (BHC), heptachlor, heptachlor epoxide, trans and cis-chlordane, trans-nonachlor, dieldrin, toxaphene, DDT and DDE. The highest concentrations (arithmetic averages in ng g^-1 dry soil) were found for toxaphene (285 ±390) and DDT’s (p, p’ DDE, 22.7 ±21.4; p, p’-DDT, 24.6 ±30.5; O, P’-DDT, 4 ±5.86; p, p’-DDD, 2.4 ±2.41) frequently used in the southeastern United States of America. Pesticide residues were not proportional to the organic carbon content of the soil, indicating that the concentration of residues was a reflection of the historical application of pesticides.

In the Lagunera Comarca there have been few studies, in this regard, ^{Garcia et al. (1988)} worked on soil monitoring in localities of the Comarca Lagunera to detect residues of organochlorine insecticides. In their results, the organochlorine insecticides detected were: aldrin, lindane, chlordane, 1, 1-dichlorophenyl-2, 2, 2-trichloroethane (DDT) and 1, 1-dichlorophenyl-2, 2-dichloroethene (DDE); the detection range was 0.05 and 4.98 mg kg^-1 of lindane and aldrin, respectively. With respect to DDT and DDE, the former as technical product and the second as its main metabolite, it was observed that DDE was found in greater quantity; this is reasonable because, due to degradation mechanisms, the amount of DDT decreases tends to increase that of DDE.

Although since 1972, in the USA the production of persistent pesticides, such as DDT and Parathion, among others, continues to have an impact on the environment, since appreciable amounts of their residues are retained in the soil, sludge, atmosphere and biota (^{Harrey et al., 1995}; ^{Harner et al., 1999}; ^{Williams and Guang, 2000}; ^{Zhang et al., 2011}; ^{Mishra et al., 2012}; ^{Yang et al., 2012}; ^{Yingxin et al., 2012}; ^{Yang et al., 2013}; ^{Zhang et al., 2013}; ^{Niu et al., 2017}).

Although more advanced countries are now more aware of the potential danger of persistent pesticides, have imposed major restrictions on their use, or are applying more carefully small amounts, global use is increasing (^{Albert et al., 1989}).

In Mexico, during the 1950s, about 1 000 t of DDT were used per year for agricultural use. Its use and production peaked in the 1960s, during which, in 1968, more than 80 000 t were produced per year. Due to its 50 years of use, its persistence and accumulation in the food chain, diet could be the major source of exposure among the general population; however, other pathways of exposure may be through air, soil, and water (^{Luévano et al., 2003}; ^{Nakata et al., 2005}; ^{Hangxin et al., 2011}; ^{Kuranchie et al., 2012} N; ^{Barakat et al., 2013}; ^{Hellar et al., 2013}; ^{Li et al., 2014}; ^{Yu et al., 2014}; ^{Buah and Humhires, 2017}).

Therefore, it is necessary to have more information about the occurrence of such persistent pesticide residues in the different components (soil, water) of the environment, preferably in those regions or areas subject to an intensive agricultural production system, such as case of the Comarca Lagunera, in the state of Coahuila, Mexico, a region that has been intensively exploited for many years through monoculture of cotton, with the establishment of up to 65 000 ha per cycle and systematic application of large quantities of agrochemicals, since up to eight applications of insecticides have been carried out in one cycle (^{Tijerina and Byerly, 1977}).

Although there is no adequate documentation, serious health problems such as allergies, cancers and malformations are reported in the region, among others, caused by contamination of natural resources by pesticides (^{Wolff et al., 1993}; ^{Costabeber et al., 2000}; ^{Luévano et al., 2003}; ^{Kunisue et al., 2004}; ^{Polanco et al., 2017}). In addition, it is important to note that, however, there is not enough evidence to indicate the origin and magnitude of this problem, not even the one related to the soils where the aforementioned agriculture is practiced. Therefore, the present work aims to detect the levels of pesticides that were most used in the Comarca Lagunera, the organochlorines, at five depths in seven localities of Comarca Lagunera in the state of Coahuila, Mexico.

Materials and methods

General location. The study was carried out in the Comarca Lagunera in the state of Coahuila, which has a geographical location of 25° 30’ 00” north latitude and 103° 30’ 00” west longitude, at a height of 1 100 meters above sea level.

Delimitation of study areas. Samplings were carried out in seven localities corresponding to five municipalities of the Comarca Lagunera of Coahuila. These were the following: The Paredon, municipality of Fco I. Madero; The Lequeitio, municipality of Fco I. Madero; San Roque, municipality of San Pedro; The Consuelo, municipality of Matamoros; Ejido the Partida, municipality of Torreon; Ejido Buenavista, municipality of Viesca; Ejido San Manuel, municipality of Viesca.

Soil sampling. Once the farms were located, a random zig-zag sampling was carried out on the land, taking a sample composed of soil per hectare. Individual soil samples were taken from the first eight centimeters deep, due to the possible presence of a higher content of organic matter and later at intervals of every 15 cm of depth up to 68 cm, forming composite samples, giving a total of five subsamples per hectare and per site.

The composite samples were analyzed in the ecology laboratory of the Phytosanitary Institute of the Graduate School and the physical and chemical characteristics in the laboratory of the chemistry section of the Institute of Natural Resources of the College of Postgraduates in Montecillos State of Mexico.

Soil characteristics analyzed. The following physical determinations were made: ^{Klute (1986)} and chemical ^{Page (1982)} to soil samples: texture, by the Bouyoucos hydrometer method; the moisture content of the soil, by the gravimetric method; the pH, with the potentiometer and a water-soil ratio of 2:1; electrical conductivity with the conductivity meter and the 2:1 ratio and organic matter by the method of Walkey and Black.

Preparation and extraction of soil samples. In order to carry out the extraction of compounds in soils, the method 3540B proposed by EPA (Environmental Protection Agency) of the United States of America was used in 1990, following the procedure that is described below: 5 g of dry soil screened in a 2 mm mesh, previously air-dried and placed in a 25*80 mm supelco cellulose cartridge previously washed with acetone, introduced into the extraction column; subsequently, the soil sample was washed for 24 h with 80 ml of a high purity hexane-acetone mixture for chromatography (HPLC) in a 1:1 ratio.

Concentration of samples. For the concentration of the samples a Kuderna Danish kit was used, taking the following steps: the liquid samples, derived from the extraction of the soil, were placed in the Kuderna Danish at a temperature of 65 °C for 2 h, concentrated volume of 10 ml.

Cleaning soil samples. One centimeter of silanized fiber was introduced into the cleaning column, packing it perfectly well, then washing it with 2 ml of hexane and 2 ml of acetone. 13 g of florisil (60-100 mesh Merck) and 4 g of sodium sulfate were added. 40 ml of hexane was added to the column to condition the sodium sulfate and florisil, discarding the eluate. The concentrated extract of the test sample was transferred to the cleaning column. 60 ml of hexane was then added, the eluate being received in an Erlenmeyer flask. The 50 ml of a hexane-diethyl ether mixture was then added in a 9:1 ratio. It was diluted with 20 ml of hexane-diethyl ether in an 8: 2 ratio. The sample was concentrated on a rotary evaporator to a volume of approximately 5 ml.

Standards used. In this work a standard of 17 Hewlett Packard organochlorine pesticides and an internal standard (dibutyl chlorendate) in acetone were used to test the recovery efficiency of the method. The determinations were performed in two replicates.

Equipment and method used. The gas chromatograph used was a Hewlett Packard 5890 series, equipped with electron capture detector. A column HP 608 of 30 m and 0.53 mm in diameter was used. The thermal conditions of the analysis were: injector at 280 °C, column at 290 °C and detector at 280 °C, using nitrogen as auxiliary gas and helium as carrier gas.

Two temperature ramps with three levels were used in order to ensure the identification of the compounds, these were.

First temperature ramp. First level: initial temperature 80 °C, final temperature 190 °C for one minute, at 30 °C per minute. Second level: initial temperature 190 °C, final temperature 280 °C for a minute at 6 °C per minute. Third level: initial temperature 280 °C, final temperature 290 °C for five minutes and speed of 20 °C per minute.

Second temperature ramp. first level: initial temperature 80 °C, up to 210 °C for one minute and 40 °C per minute of speed. Second level: initial temperature 210 °C, up to 280 °C for one minute at a speed of 4 °C per minute. Third level: initial temperature 280 °C, final temperature 290 °C for one minute and speed of 20 °C per minute.

Statistic analysis. Multiple regression analysis was performed for each locality, taking as a dependent variable the concentration of dieldrin in the soil and as independent variables the soil characteristics. The effect of each independent variable was evaluated from the regression coefficients, and the coefficient of determination (r²) was used to evaluate the joint effect of all variables.

Results and discussion

The results obtained in the physical and chemical analyzes of the soil realized in the seven localities corresponding to the five municipalities of the Comarca Lagunera are presented in Table 1.

Table 1 Results obtained in the physical and chemical analyzes of seven localities studied. 

In the Table 2 presents the results of the concentrations of dieldrin (ng g^-1) at each site and depth of soil studied.

Table 2 Concentration (ng g^-1= μg kg^-1= ppb) of dieldrin for each depth and site studied. 

With the information in Tables 1 and 2, multiple regressions were performed in order to determine the influence of the content of organic matter, the percentage of clay and the pH of the soil on the concentration of dieldrin for each studied locality, these results are shown in Table 3. The analysis performed emphasized the coefficients of determination (r²) of the regression and in the coefficients of multiple regression (weight coefficients) of each variable, independent of the same.

Table 3 Results of regression analysis. 

The results of Table 1 show the diversity of the physical and chemical characteristics of the soils under study, since these correspond to five municipalities of the Comarca Lagunera. It also emphasizes that private farms have the highest percentages of organic matter, mainly in the depth 0-8 cm, due to the best practices of soil management and the applications of organic materials, which generally carry out in the small properties.

The farms did not present problems of salinity; on the contrary, CE was generally low due to the good management and quality of irrigation water. On the other hand, the sites Lequeitio and Ejido Buenavista presented a relatively high pH in some of their depths (8.5 and 8.7 respectively), indicating the presence of excess of interchangeable sodium, since the sodium soils generally have a pH greater than 8.5.

In Table 2, it can be observed that dieldrin was the only pesticide consistently found in the seven localities studied in the Comarca Lagunera, out of a total of 17 that contained the standard used, which differs with the work done by ^{García et al. (1988)} in the Lagunera Region, in which they report the presence of aldrin in considerable concentrations. This is explained by the fact that aldrin is transformed over time by the effect of sunlight on dieldrin, this is the main cause for which aldrin was not detected. Another important aspect is that the higher concentrations of dieldrin occurred in the particular properties, in contrast to the low concentrations that were found in the Ejido the Partida.

This is due to the fact that in private farms they apply higher doses and more frequent applications of pesticides, due to poor socioeconomic conditions. Also, Table 2 indicates that the particular farms have the highest concentrations of dendrin in the second depth (8-23 cm) and the Buenavista Ejido, in the third depth (23-38 cm). These results coincide with those reported by ^{Voerman and Besemer (1986)} who mention that below the first 20 cm of soil depth only traces of dieldrin are found.

The highest concentrations in the surface soils of the small properties and ejidos analyzed were 2.57 and 1.22 ng g^-1 (ppb) respectively. In Mexico, there is no Standard that establishes maximum permissible limits for this product; however, the EPA (Environmental Protection Agency) establishes 0.04 mg kg^-1 (^{Romero et al., 2009}).

In the Table 3 shows that the highest r² corresponds to the small property El Consuelo, Municipality of Matamoros, Coahuila, and to Ejido The Partida, Municipality of Torreón, Coahuila (r²= 0.99). These localities were the only ones that presented significant difference in the weight of the organic matter: in the first case, a positive value was observed and in the ejido, a negative value; it was expected a positive value for both cases; however, possibly due to the fact that the small property presented the highest values of organic matter and concentration of dieldrin, coupled with the fact that in the Ejido The Partida presented significant difference for the weight coefficient of pH (r²= 0.96), diluted the effect of organic matter. For the rest of the localities there were no significant differences in the weight coefficients, which indicates that the soil characteristics (organic matter, clay content and pH) contribute similarly in the respective coefficients.

Also, according to the linear regressions, the highest coefficients of determination for the organic matter content of the soil correspond to the small property El Consuelo (r²= 0.99) and to the Ejido The Partida (r²= 0.83), which indicates a greater effect of the organic matter on the concentration of dieldrin in the particular farms than in the ejidos.

These results coincide with that reported by ^{Seybold et al. (1994)}; ^{Yang et al. (2005)}, who indicate that the organic carbon content of soil horizons correlates significantly with the distribution of atrazine (r²= 0.84); DDT (r²= 0.71) and with that found by ^{Mishra et al. (2012)}; ^{Yu et al. (2013)}.

Conclusions

In the soils and depths studied only the organochlorine pesticide dieldrin was found. In general, the surface layer of the soils (depth of 0-8 cm) has the highest concentrations of the product mentioned. The soil characteristics that most correlated with their concentration were organic matter content, pH and clay content. The soils of the particular farms studied had a higher concentration of dendrin than those of the ejidos evaluated.

The highest concentrations of the product in the surface soils of the small properties and ejidos analyzed were 2.57 and 1.22 ng g^-1, respectively. Although these concentrations are lower than the maximum permissible limits established by the EPA, there is a possibility of incorporating this pesticide into the food chain and its bioaccumulation in living organisms.