Three samples of leaves, bark, and roots of Rhizophora mangle trees (between 1.5 and 4 m height) were sampled from four sampling sites inside each lagoon during the period of June to August 2011 (Figure 1). To avoid differences between sampled trees, individuals within sampled regions, sites adjoined to effluents drainages from aquaculture or agriculture, or without the apparent influence of these effluents were selected. Tissue samples were analyzed in triplicate. The selection criteria of trees within the regions were focused on individuals above 5 m height, healthy, and close to each other in the same sampling site. Bark and roots samples were collected by scraping the mangrove’s tissues with a jack-knife (Naranjo-Sánchez and Troncoso-Olivo, 2008). Three leaves per tree were randomly removed and were selected without yellowish coloration or signals of broken edges and physical or parasitic damage. Each leaf was cleaned with distilled water and packed in an aluminum foil in a cold dark recipient and transported to be stored in a freezer at the CICIMAR Environmental Laboratory (MBAY), CIIDIR SINALOA Research Center (NLAG), or the University of Campeche in Ciudad del Carmen (TLAG). All tissues samples were individually wrapped with newspaper and were dehydrated slowly at 40°C in a botanical incubator for 7 days (Haq et al., 2011). After dehydration, the samples were separated and individually milled in a grain grinder (Krups^® GX 4100, Type F408, Mexico). The powders were stored in black plastic bags under darkness and dry room-temperature conditions.

All samples were transported to the CIIDIR-IPN Research Center in Durango City, Mexico. Three aliquots per sample were used. Since phenolic compounds have many unsubstituted hydroxyl or sugar groups, they are considered polar and soluble in polar solvents, such as ethanol and water. Briefly, 10 ml of ethanol at 5°C (80%, v/v) was added to each milled dry aliquot (0.5 g) and were shaken on a heating plate (Cimarec^®, United States) in darkness and room temperature (30 min) and then filtered (Whatman^® No. 1). Solutions were evaporated to dryness using a water evaporator at 40°C, and a second extraction was done with 10 ml of ethanol at 5°C (80%, v/v). Then, 20 ml of petroleum ether was added to each solution after extraction to remove chlorophyll. All supernatant extracts were kept in darkness and refrigerated until further analysis. The leaf (100 μl), bark (10 μl), and root (20 μl) extracts were placed in a vial and diluted with ethanol (800 g L^–1) at 900, 990, and 980 μl, respectively, at room temperature, to calculate TPC, TFC, ABTS∙ + radical scavenging capacity, and Cu⁺ CA (Saiga et al., 2003).

Absorbance readings (ABS) of TPC (750 nm), TFC (510 nm), and AC (695 nm) were compared with their standard curves with Gallic acid (TPC (Equations 1 and 3), and catechin (TFC) (Equation 2), AC (ABTS∙ +) (Equation 3), and Cu⁺ CA chelating rate (Equation 4) equivalent concentrations, respectively.

Each extract (100 μl) was mixed with Folin-Ciocalteu reagent (750 μl) diluted ten times with distilled water, and kept in darkness (5 min); then, a 60 g L^–1 solution of Na₂CO₃ (750 μl) was added to neutralize the reaction and incubated for 90 min to read the absorbance (750 nm) to estimate the TPC expressed as Gallic acid equivalents (GAE) per 100 g dry basis (db) (mg GAE 100 g^–1 db). A standard curve of Gallic acid (SIGMA^® Cat. G7384) was read at 750 nm in a Thermo Scientific^® Spectrophotometer UV-VIS GENESYS^® 10 s (r² = 0.9968) (Adom and Liu, 2002) (Eq. 1).

Based on the aluminum chloride colorimetric method (Sultana et al., 2009), 100 μl of each extract was mixed with 75 μl of NaNO₂ (50 g L^–1). After 5 min, 75 μl of AlCl₃ (100 g L^–1) was added, and 6 min later, 500 μl of NaOH (40 g L^–1) and 600 μl of distilled water were added. The solution was mixed and the absorbance (510 nm) was read within 30 min in the Thermo Scientific^® and expressed in milligrams of catechin per gram dry basis (mg CE 100 g^–1 db). The standard curve was performed with a standard solution (0.1 mg ml^–1, catechin hydrate, SIGMA^®, Cat. QO125) (Eq. 2).

The method described by Saiga et al. (2003) was used. A mixed solution of 290 μl of 50 mM acetate buffer (pH 6.0), 6 μl of 4 mM pyrocatechol violet, copper (20 μl), and 1 μg (10 μl) (CuSO₄ 5H₂0) was added to each aliquot sample (by triplicate). The Abs of change of color from blue to yellow was measured at 632 nm in a Thermo Scientific^® Spectrophotometer. The standard was an aliquot of Kupfer (II)—sulfate standard solution (Merck^®, Cat. 202790) (Eq. 4):

Where: A₀ = absorbance of the control (blank, without extract); A₁ = absorbance in the presence of the extract; and; Cu = chelating activity assay.

The particle size and clay type of all sediment samples were analyzed in the Environmental Pollution Laboratory of the CIIDIR-IPN Research Centre in Sinaloa (CIIDIR-SINALOA), Mexico. Four sediments samples were collected from each collecting point in the three lagoons. The textural class of the soil was determined with the Bouyoucos method (Bouyoucos, 1962). Three replicas for a given sample (50 g) of soil (dry and 2-mm sieved) per sample were analyzed. Each replica was placed in a cup filled with distilled water up to 5∼6 cm above the sediment supplied with 35 ml of 1 N sodium hexametaphosphate. Each replica was thoroughly mixed (for high clay soils: allow to slake (soak) for 15 min; for sandy soils: mix for 5 min, and for finely textured soils: mix for 15 min). The soil suspensions were transferred to a sedimentation cylinder and diluted in 1 L of water. A hydrometer was placed inside the cylinder, filling with water to the 1,150 ml mark. The hydrometer was removed and was sealed each time at the top of the cylinder with parafilm. The cylinder was tilted back and forth for thorough mixing several times and set on a stable surface. The hydrometer was placed into it, and the 40 s timing was immediately recorded when it was read. The hydrometer was removed from the cylinder. The time and temperature were recorded and started a 2 h setting period. A control blank cylinder (water; 5 ml of l N sodium hexametaphosphate) was prepared, and both were read as follows:

Temperature correction factor (T°C) = (observed temperature – 20°C) × 0.03.

Corrected 40 s reading: 40 s (T°C): 40 s reading – Blank reading + T°C.

Corrected 2 h reading: 2 h (T°C) = 2 h reading – blank reading + T°C.

Particle percentages: Sand (%) = (2 - 0.05 mm): [Oven dry soil weight – 40 s reading (c)] + [oven dry soil weight] × 100%; Clay (%) = (< 0.002 mm): [2 h reading (c)] × [oven dry soil weight]^–1 × 100%; Silt (%) = (0.05 - 0.002 mm): 100 - (% sand + % clay).

We determined the textural classes of the soil sample using the textural triangle.

Eight sediment samples per coastal lagoon, two samples per sampling point (n = 72), were collected with a Teflon spoon and immediately placed into polyethylene bags, stored in an icebox, and then kept in a freezer for preservation until analysis in the Environmental Pollution Laboratory in the CIIDIR-SINALOA.

Acidic digestion was used to determine Fe, Cu, Pb, Ni, Cd, and Mn contents (Breder, 1982). Each sediment sample was dried (60°C) for 48–96 h and homogenized by grinding. Then, 0.5 g of each sample was placed on the heating plates with 5 ml aqua regia (1:3 HCl: HNO3) and kept until dryness was achieved again. After cooling, the solution was adjusted to 50 ml with deionized water and stored in plastic vials until analysis. The total content of metals was determined by flame atomic absorption spectroscopy (GBC^® Avanta) (air-acetylene flame), recovering over 90%. Reference materials (PACS-2, Marine Sediment Reference Materials for Trace Metals and other Constituents, National Research Council, Canada) were used to ensure standardization.

Eight sediment samples per coastal lagoon, two samples per sampling point, were collected with a Teflon^® spoon, wrapped carefully in aluminum foils, and stored in polyethylene bags. Samples were kept in a cooler with ice for preservation and transported to the Environmental Pollution Laboratory of the CIIDIR-SINALOA.

The USEPA 8270, 508, 8081b, 3660, 3500, and 3600 extraction technics were conducted (Usepa, 1995, 1996a, 1996b, 1998, 2007a, 2007b, 2007c). In general, 5 g of dried sediment was added to an Erlenmeyer flask adding enough dichloromethane to cover the sediment and immediately was sonicated for 15 min (repeated three times). The dichloromethane was removed at ∼150 ml each time (method 3550). After that, a sodium sulfate tap was placed in a fiberglass funnel, removing the rest of the dichloromethane particles and placing the flask in a rotavapor until dryness to recover 2 ml of hexane.

Gas chromatograph Perkin Elmer^® AutoSystem XL, with an electron capture detector (ECD), coupled to the Navigator software © V. 6.3.1., provided with a J&W^® DB 5–30 m (0.25 mm internal diameter capillary column, 5% diphenyl-95% dimethylpolysiloxane), was used to determine the OCPs. Briefly, 1 μl of the hexane-diluted OCP sample was injected into the chromatograph. OCPs were identified by their retention times, and quantitation was calculated on peak height/area and by comparison with the results of the reference standard (SUPELCO^® 48858-U Sigma-Aldrich) that included the OCP analytes: α–BHC, β–BHC, lindane, δ–BHC, aldrin, 6–hydroxy-2–naphthyl disulfide, 1,1–dichloro–2,2–bis (4–chlorophenyl) ethane, 4,4′–DDT, dieldrin, α–endosulfan, β–endosulfan, endosulfan sulfate, endrin, endrin aldehyde, heptachlor, heptachlor exo-epoxide, and methoxychlor.

In the Environmental Pollution Laboratory of the CIIDIR-SINALOA, the chromic and sulfuric acid methods (Walkley and Black, 1934) were used. A 100 g sample of soil was weighed and placed into a 500 ml Erlenmeyer flask containing 10 ml of 1 N potassium dichromate solution. Then, 20 ml of sulfuric acid was added and mixed by gently rotating for 1 min. Samples were left for 30 min and then diluted with deionized water (200 ml). After dilution, 10 ml of phosphoric acid, 0.2 g ammonium fluoride, and 10 drops of diphenylamine indicator were added. The mixture (drop by drop) was titrated with 0.5 N ferrous ammonium sulfate solution until the endpoint from dull green to a brilliant green. Three replicas per sample and one quality control sample was analyzed for each set of samples with the following equation:

% Organic Matter = 10[1(S÷B)] × 0.67

Where, S = sample titration; B = blank titration.

One-way ANOVA to analyze the differences in TPC, TFC, ABTS∙ +, and Cu⁺ CA, OCP, metal, OM, and phosphate concentrations in sediments was applied, and if significant differences were found (p < 0.05), a post hoc honestly comparison tests (HSD) Tukey’s, Scheffé’s, and Bonferroni and Holm multiple were used (Vasavada, 2014). Pearson’s analysis was performed using SIGMA PLOT^® software to obtain the coefficient of correlations (r) of TPC, TFC, AA, and Cu⁺ CA per tissue among sites and tissues within sites, among bio compounds, and AA and Cu⁺ CA activities with trace metal concentrations.

The mean concentration and standard deviation (SD) were calculated for trace metals and OCPs and compared with the average continental crust metal concentration (Taylor, 1964) to determine the enrichment rate. A regression analysis was carried out to determine the coefficient of determination (r²) among trace metals.