The mangroves of Nariño Department, Colombian Pacific coast, have a high structural development, with average tree density between 308 and 1124 individuals/ ha and basal areas between 8.2 and 35.4 m²/ha, dominated by the Rhizophora genus (^{Tavera, 2010}). The cover of mangrove forest in 2014 (117,469 ha) represented 56% of the total cover of mangroves in the Colombian Pacific region (209,403 ha) and 41% of total national mangrove cover (289,122 ha; ^{Gómez-Cubillos et al., 2015}). This ecosystem is considered strategic for the Colombian Pacific due to ecosystem function and the provision of goods and ecosystem services that sustains the economy of coastal inhabitants of the Department and municipality of Tumaco (^{Uribe and Urrego, 2009}). Therefore, local populations are vulnerable to the deterioration of these ecosystems (INVEMAR et al., 2017).

The mangroves of Nariño, and especially those of Tumaco, are affected by different environmental degradation factors. These factors include the discharge of sewage and solid waste due to the poor sanitation conditions of the municipalities (INVEMAR, 2017); logging of mangroves and agricultural activities that cause loss of coverage in certain areas of the forest (^{Tavera, 2010}); water contamination in the higher river basins that drain into the Tumaco Bay and the Pacific Ocean, product of the mining of nickel, gold and other precious metals (Corponariño, 2015; Agencia Nacional de Minería, 2017); and recurrent crude oil spills that affect the marine environment and mangroves, such as 410000 gallons of crude oil spilled as a result of attacks to the Trans-Andean oil pipeline in June 2015 (INVEMAR, 2015; INVEMAR et al., 2017). Oil spills are the most important factor related to the environmental degradation of Tumaco's mangroves due to its high environmental impact (INVEMAR et al., 2017).

According to ^{Duke (2016)}, after an oil spill in mangroves, the ecosystem has a lethal response in a few days, reflected in the death of the most sensitive organisms (mollusks, crustaceans, fish, seedlings, among others), and the death of mature mangroves during the first six months. In addition, the mangrove has a sub-lethal response in the long term which helps us to understand the impacts and possible trajectories of recovery or loss of ecosystem (Duke, 2016). The hydrocarbons analysis of mangrove sediments in Mira River estuary, two years after the crude oil spill is relevant to assess of the impacts and environmental effects in the ecosystem; and for design and implementation of ecological restoration measures.

The mangroves of the Mira River estuary develop in places with different levels of hydrodynamic energy, where processes of sediment erosion and deposition vary. This is associated with the flow and the presence of numerous river meanders flowing into the Tumaco Bay and the Pacific Ocean (^{Restrepo and López, 2008}). Fine silt and clay sediments are deposited in these mangroves and accumulate part of the organic material imported by the river as well as that generated by the mangroves themselves. These processes favor the retention of organic contaminants, including petroleum hydrocarbons that due to their characteristics of low solubility and hydrophobicity, adhere to the sediments and persist for a long period of time (^{Olguín et al., 2007}). To determine the mangrove sediment quality of Mira River estuary along the southern coast of Tumaco that was affected by the crude oil spill, the chemical quantification of petroleum hydrocarbons was made, and the adverse effects risks of these compounds on benthic biota associated to the mangroves were identified.

The study area is located in the Mira River estuary, southern coast of Tumaco municipality, Nariño Department, Colombian Pacific, between the coordinates 1°30'0" y 1°50'0" N y 78°40'0" y 79°10'0" W (Figure 1). The mangrove species in this zone are Rhizophora mangle, R. Harrizonii, R. racemosa, Avicennia germinans, Laguncularia racemosa, Pelliciera rhizophorae and Mora oleifera (^{Tavera, 2010}). The climate regime is monomodal, with rainfall throughout the year, reaching multi-year average values between 2500 and 7000 mm. Increased rainfall occurs in the first half of the year. The multi-year averages of temperature oscillate between 24 and 28°C; the relative humidity between 85% and 95%; and the evaporation between 700 and 1300 mm (IDEAM, 2014). The tide in the area is semi-diurnal, with a daily cycle with two high tides and two low tides, and tidal ranges average between 2.46 and 2.58 m (^{Restrepo and Otero, 2007}).

To assess the quality of mangrove sediments in the Mira River estuary, nine sampling stations were established, distributed in the Coba, Bocagrande, Mira, Terán, Congal and Chontal areas (Figure 1). Sediment samples were collected between February and March 2017, during the low tide, when the mangrove substrate was not flooded. At each station, in the mangrove intertidal zone, simple surface sediment samples were collected (up to 10 cm depth) using a metallic shovel.

Figure 1 Study area with sampling stations of mangrove sediments (red circles) in the Mira River estuary, Tumaco municipality, Nariño Department, Colombian Pacific. 

Sample analyses were made in the Laboratories Unit of Marine Environmental Quality (LABCAM) at Marine and Coastal Research Institute (INVEMAR). Granulometry was determined using the gravimetric method with dispersed in sodium hexametaphosphate, screening in different sieves (^{Dewis and Freitas, 1984}). Grain-size classification was completed using the granulometric scale of ^{Wentworth (1992)}. The content of organic matter was determined by wet oxidation (ICONTEC, 2013). The concentration of petroleum hydrocarbon equivalent to chrysene (HP) was done by extraction with dichloromethane-acetone (1:1) and quantification by fluorescence spectroscopy (UNESCO, 1982); and 16 hydrocarbons aromatic polycyclic (PAHs) of toxicological importance (naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo (a) anthracene, chrysene, benzo (b) fluoranthene, benzo (k) fluoranthene, benzo (a) pyrene, indene (1,2,3-cd) pyrene, dibenzo (a, h) anthracene and benzo (ghi) perylene), through extraction with dichloromethane-acetone (1:1) (EPA, 1996) and quantification by gas chromatography coupled with mass selective detector (EPA, 2014).

To determine estuarine sediment quality based on the concentrations of HP by dry weight (d.w.), results were compared with classification values proposed by ^{Massoud et al. (1998)} as follows: unpolluted areas (<15 μg/g d.w.), slightly oil polluted (15-50 μg/g d.w.), moderately oil polluted (>50-200 μg/g d.w.) and heavily oil polluted (>200 μg/g d.w.). The concentrations of 16 PAHs were compared with the values of threshold effect level (TEL) and probable effects level (PEL) suggested by ^{MacDonald et al. (1996)} for estuarine sediments. The method of MacDonald et al. (1996) is based on the coincidences between chemical concentrations with previously known biological effects. This method is widely used as a tool for initial detection of the distribution of environmental importance of chemical substances and the risk of adverse effects to benthic organisms (^{Birch, 2018}).

Figure 2 shows the granulometry, organic matter content and HP concentration in the mangrove sediments of the River Mira estuary. At most stations, silt and clay grains predominate with proportions between 57.7 and 94.1%, with the exception Bocagrande (E4) where fine sands predominate (58.9%), and in Teran (E6), Congal (E7) and Chontal (E9) where more than 95% of the sample collected corresponded to vegetal material (Figure 2a). The high percentage of vegetal material in the samples is normal in mangrove soils, where leaf litter produced by forest high primary productivity and low organic matter decomposition rates in anoxic conditions, accumulates to form peat (^{Saraswati et al., 2016}) which can reach about 2 m thick in mangroves of the Pacific (^{Fujimoto and Miyagi, 1993}; ^{Ezcurra et al., 2016}). The levels of organic matter were between 46 and 106 mg/g d.w. The highest values were registered in the stations Congal (E8) and Chontal (E9) (Figure 2b), which are areas of mangrove preservation that presents a high structural development (^{Tavera, 2010}; INVEMAR et al., 2017).

Figure 2 Maps with the results of: (A) granulometry, (B) organic matter and (C) petroleum hydrocarbons equivalent of chrysene -HP in mangrove sediments in the Mira River estuary, Tumaco municipality, Nariño, Colombian Pacific. 

The HP concentrations in mangrove sediments were between 3.8 and 846 μg/g d.w. (Figure 2c). Comparing these results with the reference values of ^{Massoud et al. (1998)}, mangrove sediments in E1 are considered heavily oil polluted (846 μg/g d.w.); E7 station is moderately oil polluted (86 μg/g d.w.), and station E8 (85 μg/g d.w) and E3 (33 μg/g d.w.) is slightly oil pollution. The other stations had HP concentrations below 15 μg/g d.w. and were considered as not contaminated (Massoud et al. 1998). The high HP values (heavily oil polluted) at Coba (E1 station) may be associated with its location in an area of sediment deposition, evidenced by the formation of small beaches of mud at the edge of the mangroves with contents of clay of 82%, silts 87% and organic matter of 65.2 mg/g d.w. These condition favor contaminant persistence in sediments (^{Olguín et al., 2007}). Sites found not to be contaminated are (Mira (E5) and Teran (E6)) are located on the principal channel of the Mira River, where further erosion of the mangrove bank is observed which favor the removal of sediments affected by the crude oil spilled in June 2015 and their subsequent transport to the Pacific Ocean. In the case of the Chontal station (E9), it was the least affected by the crude oil spill and after two years was found unpolluted.

The highest concentration of Σ16PAHs (1590 ng/g d.w.) on mangrove sediments was also found at the Coba station (E1). Between the HAPs quantified, chrysene had the highest concentration (1154 ng/g d.w.), a value that exceeded the reference for probable effect level (PEL) in benthic biota (Table 1), followed by pyrene (424 ng/g d.w.), at concentrations greater than the threshold effect level (TEL, Table 1). In the Bocagrande station (E4) was quantified acenaphthylene (7.8 ng/g d.w.); nevertheless, this concentration is below the references of adverse effects on benthic biota (Table 1). At the other stations, the PAHs were below the limit of quantitation of the analytical method used and the adverse effects on benthic biota references (Table 1). The chrysene and pyrene have chemical characteristics that caused them to be retained in the sediments for a long time, due to the four-ring structure of benzene, high molecular weight (chrysene 228.29 g/mol and pyrene 202.25 g/ mol) and high hydrophobicity (^{Olguín et al., 2007}). In the mangroves, the chrysene, pyrene and other PAHs can be re-suspended into the water by the action of tides, and can oxidized resulting in metabolites with greater toxicity such as hydroxycrylene and hydroxypyrene, among others, that have an impact on the development of fish and mollusks (^{Xiu et al., 2016}; ^{Diamante et al., 2017}). Therefore, it is necessary to assess the presence of these metabolites in the mangrove sediments of the Mira River estuary, and their adverse effects on benthic organisms and the mangrove plants. This information is important for the detection of hazards and environmental risks, which can then facilitate the design and implementation of preventive measures and strategies of bioremediation and restoration of the mangrove of Tumaco.

Table 1 Polycyclic aromatic hydrocarbons (PAHs) concentration in mangrove sediments of the Mira River estuary, Nariño Pacific coast, two years after the crude oil spill in June of 2015 and threshold effect level reference values (TEL) and probable effect level (PEL) of PAHs in estuarine sediments proposed by ^{MacDonald et al. (1996)}. Values with symbol (<) are the limit of analytical method quantitation used for each compound. 

In summary, the contamination by petroleum hydrocarbons persists in mangrove sediments at Coba (E1 and E2) and Bocagrande (E3 and E4) stations two years after the attack on the Trans-Andean pipeline in June 2015 during which 410000 gallons of crude oil were spilled on the Mira River. Especially in Coba, the chrysene and pyrene represent a high risk of adverse effects to benthic organisms associated with mangroves, therefore other specific studies are required to determine the presence of hydroxylated PAHs and to assess the chronic effect of petroleum hydrocarbons to benthic organisms, especially those of commercial importance, such as the piangüas (Anadara sp). In addition, in order to rehabilitate and conserve the mangroves in Tumaco, it is necessary to design and implement measures to prevent actionably pollution, and design and implement effective bioremediation measures.