**3. Methodology**

#### **3.1. Sampling**

Throughfall deposition consists of collected solutes in wet deposition under the forest canopies. This method is widely used to estimate atmospheric deposition inputs to the forest ecosystems, since it includes both, wet and dry deposition. Because of the high cost and difficulty of the measurements of dry deposition fluxes in forest stands, throughfall collectors constitute a good choice to obtain a reliable estimation of N and S atmospheric inputs in a given ecosystem [11]. However, in regions where N deposition is chronic, throughfall collec‐ tors underestimate the total N fluxes (wet plus dry) [12]. This underestimation is attributed to the uptake and retention of atmospheric N compounds in tree canopies, especially during dry deposition [13]. In spite of this restriction, throughfall estimations of N and S deposition are useful to establish a base line in sites where atmospheric deposition data are not available as is the case of the tropical forests in Mexico. Since, a worldwide database of atmospheric N and S inputs is available; inputs quantified in a given region can be compared with these reference values to obtain a diagnosis about the severity of N and S deposition and the possible vulner‐ ability of the ecosystems.

Automatic wet/dry collectors are expensive and need to satisfy certain requirements for their installation; therefore, passive sampling devices as throughfall collectors constitute a good sampling alternative in a given region; in addition, it is possible to increase the density of the sampling grid at a low cost. Throughfall deposition can be defined as the hydrologic flux of ions to floor contained within a solution. This work used passive throughfall collectors developed and tested by [14], constituted by an ionic exchange resin mixed bed within a column.

Throughfall deposition in Carmen Island was collected on one- year basis, from July 2013 to June 2014; passive sampling devices were exposed during two months for six periods through‐ out the year in each of the thirteen sampling points. Samples were collected at the end of each period.

Samples were collected with a funnel; the solution was channeled to the mixed resin bed through the column (particularly, in this work, a mixed AmberliteTM IRN150 ion exchange resin bed was used)., where ions were retained (Figure 3). At the end of the sampling period (two months), retained ions within the same column were extracted by using extraction solutions to recover the sampled elements. Sulfate and nitrate retained within the resin column were selectively extracted with a 2 N KCl solution and analyzed by turbidimetry and color‐ imetry, respectively.

Extraction efficiency was calculated as the percentage of the loaded ions on the columns to the recovered in sequential extractions. It was found that recovered N and S percentage from the third extraction was insignificant, therefore; only two extractions were considered for the recovery process, reaching an efficiency of approximately 96.8%. Similar extraction

was analyzed by molecular absorption spectrometry using the blue indophenol method

colorimetrically by using the brucine method [18]. The weight of the extractant was converted to volume by using the specific density of the extractant solution (2N KCl = 1.05 g ml-1). The amount of each collected ion in a given column was determined as the factor of extractant volume multiplied by ionic concentration in the extractant. The surface area of the funnel opening and the sampling period were used to estimate the deposition fluxes of N


was analyzed

[16]. Sulfate was determined by turbidimetric method [17], whereas NO3

**Figure 3. Throughfall sampling devices used in this study. Figure 3.** Throughfall sampling devices used in this study.

and S per land area per year (Kg ha-1 yr-1).

**3.2 Chemical Analysis** 

NH4 +

efficiencies have been reported by other authors [15].

**Figure 2.** Sampling sites location along Carmen Island, Campeche, Mexico.

Throughfall deposition consists of collected solutes in wet deposition under the forest canopies. This method is widely used to estimate atmospheric deposition inputs to the forest ecosystems, since it includes both, wet and dry deposition. Because of the high cost and difficulty of the measurements of dry deposition fluxes in forest stands, throughfall collectors constitute a good choice to obtain a reliable estimation of N and S atmospheric inputs in a given ecosystem [11]. However, in regions where N deposition is chronic, throughfall collec‐ tors underestimate the total N fluxes (wet plus dry) [12]. This underestimation is attributed to the uptake and retention of atmospheric N compounds in tree canopies, especially during dry deposition [13]. In spite of this restriction, throughfall estimations of N and S deposition are useful to establish a base line in sites where atmospheric deposition data are not available as is the case of the tropical forests in Mexico. Since, a worldwide database of atmospheric N and S inputs is available; inputs quantified in a given region can be compared with these reference values to obtain a diagnosis about the severity of N and S deposition and the possible vulner‐

**3. Methodology**

ability of the ecosystems.

**3.1. Sampling**

150 Current Air Quality Issues

Extraction efficiency was calculated as the percentage of the loaded ions on the columns to the recovered in sequential extractions. It was found that recovered N and S percentage from the third extraction was insignificant, therefore; only two extractions were considered for the recovery process, reaching an efficiency of approximately 96.8%. Similar extraction efficiencies have been reported by other authors [15].

#### **3.2. Chemical analysis**

NH4 <sup>+</sup> was analyzed by molecular absorption spectrometry using the blue indophenol method [16]. Sulfate was determined by turbidimetric method [17], whereas NO3 was analyzed colorimetrically by using the brucine method [18]. The weight of the extractant was converted to volume by using the specific density of the extractant solution (2N KCl = 1.05 g ml- 1). The amount of each collected ion in a given column was determined as the factor of extractant volume multiplied by ionic concentration in the extractant. The surface area of the funnel opening and the sampling period were used to estimate the deposition fluxes of N and S per land area per year (Kg ha- 1 yr- 1).

#### **3.3. Data interpolation and mapping**

All methods based on point measurements (e.g., wet deposition, micrometeorological and dry deposition measurements, throughfall deposition), cannot be directly connected to emission inventories. Maps can only be produced directly from these measurements if the network is enough dense to account for spatial and temporal variations. This may be the case for networks measuring compounds with little spatial variation or for measurements of deposition in areas of simple terrain as the case of Carmen Island. A point measurements network should be interpolated using statistical methods as the kriging technique, which allows including monitoring data from neighboring stations for interpolation in a national scale [19]. Consid‐ ering the characteristics of measured data and the expected interpolation results, diversified methods have been employed with atmospheric deposition network data [10, 20- 22].

To obtain N and S deposition maps in Carmen Island, a geostatistical procedure was used to interpolate field measurements into a continuous spatially pattern, where data were interpo‐ lated using the kriging interpolation technique. Kriging is a statistical method that provides unbiased estimates of variables in regions where the available data exhibit spatial autocorrelation and these estimates are obtained in such a way that they have minimum variance. The first step was to obtain the statistics summary of primary data. From this, a descriptive report of data was carried out, including mean, maximum, minimum values and their frequency distribution. This analysis was carried out in order to determine if it was necessary to make a data transformation since the results of the kriging interpolation are more reliable when data are normally distributed [20].Data interpolation was used in this work to obtain a continuous variable by isolines to increase the number of points in the sampling grid. Thresh‐ old values definition helps to obtain isolines or imaginary lines in which studied variables become continuous, this method is useful since it takes into account the spatial behavior of the variable from distance and the variability between points.

Values were obtained from a linear combination of the original points with known data. As a result, greater compact areas around the variable maximum values were estimated. Once additional points and isolines were obtained, deposition data were mapped to assess their spatial and temporal distribution along the Island.
