Potential of bioremediation for buried oil removal in beaches after an oil spill

Marine Pollution Bulletin 76 (2013) 258–265

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Potential of bioremediation for buried oil removal in beaches after an oil spill Joana Pontes a, Ana P. Mucha b, Hugo Santos b, Izabela Reis b, Adriano Bordalo b,c, M. Clara Basto a, Ana Bernabeu d, C. Marisa R. Almeida b,⇑ a

CIMAR/CIIMAR, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre, s/n, 4169-007 Porto, Portugal CIMAR/CIIMAR – Centro Interdisciplinar de Investigação Marinha e Ambiental, Universidade do Porto, Rua dos Bragas, 289, 4050-123 Porto, Portugal Laboratório de Hidrobiologia, Instituto de Ciências Biomédicas, Universidade do Porto (ICBAS-UP), R. Viterbo Ferreira, 228, 4050-313 Porto, Portugal d GEOMA, Marine and Environmental Geology Group, Department of Marine Geosciences, Universidad de Vigo, 36310 Vigo, Spain b c

a r t i c l e Keywords: Buried oil Hydrocarbons Bioremediation Biostimulation Bioaugmentation

i n f o

a b s t r a c t Bioremediation potential for buried oil removal, an application still lacking thorough research, was assessed in a specifically designed system in which an artificially contaminated oil layer of sand was buried in a sand column subjected to tidal simulation. The efficiency of biostimulation (BS, fertilizer addition) and bioaugmentation (BA, inoculation of pre-stimulated indigenous hydrocarbon-degrading microorganisms plus fertilizer) compared to natural attenuation was tested during a 180-day experimental period. The effect of BA was evident after 60 days (degradation of hydrocarbons reached 80%). BS efficacy was revealed only after 120 days. Microorganisms and nutrients added at the top of the sand column were able to reach the buried oil layer and contributed to faster oil elimination, an important feature for effective bioremediation treatments. Therefore, autochthonous BA with suitable nutritive conditions results in faster oil-biodegradation, appears to be a cost-effective methodology for buried oil remediation and contributes to the recovery of oil-impacted areas. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Large quantities of petroleum enter the environment every year through leakage from storage tanks and pipelines or by release in accidental spills, such as that of the Prestige tanker (Albaigés et al., 2006). Recent research on sandy beaches affected by this oil spill revealed the persistence of oil in the sand at depths up to 4 m, a much greater depth than previously expected (Bernabeu et al., 2009). Although the buried oil from this spill is considered very persistent, oil degradation has been observed, as different morphologies of the buried oil were found: tar-balls, particles, oil coatings and emulsions (Bernabeu et al., 2006). The biodegradation of oil, however, is a slow process that requires weeks, months or years (Atlas and Hazen, 2011; Zhu et al., 2001). It is important to study approaches that may accelerate the existing natural attenuation of oil, such as bioremediation. Bioremediation might be a cost-effective treatment tool for accelerating oil removal from contaminated environments. This treatment can be applied by stimulating indigenous microbial assemblages (biostimulation) or by increasing microbial assemblages capable of degrading the oil (bioaugmentation). In fact,

⇑ Corresponding author. Tel.: +351 22 0402571; fax: +351 22 0402659. E-mail addresses: [email protected], [email protected] (C.M.R. Almeida). 0025-326X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.marpolbul.2013.08.029

the growth of hydrocarbon-degrading bacteria and hydrocarbon degradation can be strongly enhanced by fertilization and biostimulation and has proven to be an effective bioremediation treatment on several types of shorelines (Röling et al., 2002). However, the amendment of nutrients to open beach environments is often impractical because water soluble nutrients can be rapidly diluted and leached out of the sediment (Nikolopoulou and Kalogerakis, 2009). An option that avoids these adverse impacts is the use of oleophilic fertilizers, such as Inipol EAP 22 or S200, that are designed to release nutrients continually or intermittently over a period of time on contact with water (Xu et al., 2004). Although biostimulation is considered to be effective, it may still require time to be successful because of the scarcity of indigenous microbes capable of degrading hydrocarbons (Hosokawa et al., 2009). This can be overcome by bioaugmentation. In general, bioaugmentation can be conducted by adding allochtonous microbial assemblages with a known capability of degrading the oil to the impacted beach environment. However, this option can be quite uncertain because it implies the addition of exogenous microorganisms to the environment with unknown effects on the microbial natural diversity and on the environment. In addition, several studies have demonstrated that this type of bioaugmentation is not successful in most cases because exogenous microorganisms are not able to compete with the indigenous ones

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

(Hosokawa et al., 2009). Therefore, another option is bioaugmentation with indigenous microbial assemblages, a protocol that has only been explored in a few reported studies (Hosokawa et al., 2009). Bioremediation is a proven technique for the cleaning and restoration of sites that have been superficially contaminated by petroleum products (Gallego et al., 2007), including marine habitats of a shoreline contaminated by marine oil spills (Swannell et al., 1996; Head and Swannell, 1999). Although bioremediation has been applied to groundwater in terrestrial ecosystems (e.g., Farhadian et al., 2008; Van Stempvoort and Biggar, 2008), to our knowledge, it has still not been applied to the remediation of buried oil in coastal ecosystems. Accordingly, research on this topic is imperative. The problem with buried oil is completely different from that of superficial oil. When bioremediation techniques are applied to superficial oil, the treatments are added directly to the layer of oil. In the case of buried oil, these treatment approaches may not be successful. One must ensure that the bioremediation treatment (i.e., the solution added that contains either nutrients or microorganisms) reaches the buried oil. In this study, bioremediation efficiency for buried oil removal was assessed in controlled laboratory conditions. Two treatments, biostimulation (addition of the oleophilic fertilizer S200) and bioaugmentation (inoculation of an indigenous oil-degrader consortium pre-stimulated in the laboratory combined with the addition of S200 fertilizer), were tested and compared with natural attenuation (oil degradation by indigenous microorganisms present naturally in the sand). Experiments were conducted in a system designed specifically for this study, with a tide regime simulation to mimic the natural conditions of the intertidal area of sandy beaches and to better reproduce the buried oil conditions.

259

entrance and exit from the columns’ bottom. Filling the columns took ca. 45 min, allowing the water to remain slightly above the sand top. At this stage, each column was fully flooded with seawater and remained flooded for ca. 5 h. Afterwards, the water drained out slowly (ca. 2 h), leaving the column dry for ca. 4 h. The water residence time was selected to mimic the period of sediment inundation in the beach intertidal zone. To reproduce the buried oil, a portion of sandy sediment (1.6 L per column) was mixed with crude oil supplied by a refinery (80 mL per column). The mixture was then left to age in a hood for 3 days to allow the physical adsorption of petroleum hydrocarbons to sand particulates and to eliminate the more volatile petroleum fractions, simulating natural weathering. This contaminated sand was mixed daily to maintain aerobic conditions. Then, it was buried at approximately 30 cm from the surface in each column. The contaminated layer (ca. 5 cm height) had approximately 20 cm of clean sand below and ca. 30 cm of clean sand above (each column contained approximately 17.5 L of sediment) (Fig. 1 B). After the completion of column assemblage, the columns were left undisturbed without any physical shaking, except for the water flow movement. The grain size of the sediment and the flow of water going in and out resulted in oxic conditions, under the assumption that the hydrocarbon degradation processes occurred in oxic conditions. The column system was placed indoors in a dark room under a constant room temperature (at 19 °C, the average temperature expected to be found at the sub-surface sediment layer). In addition, all columns were covered with a black cover, and a low intensity red light was used during column manipulation to prevent photo-oxidation. 2.3. Treatments performed

2. Materials and methods 2.1. Sampling Sandy sediment (between 2 and 20 cm) was collected (in black plastic bags) in August 2011 in the middle tide zone of the São Pedro de Maceda beach (NW of Portugal) at low tide to retrieve wet, but not submerged, sediment. The beach is protected from direct industrial and human contamination and, to our knowledge, has not been affected by any major oil spills. In the laboratory, the collected sediment was carefully combined and homogenized to avoid small-scale variations in its composition that could influence the experiments. A small portion of the homogenized sediment was separated for microbial and chemical characterization. The sediment was also characterized in terms of organic matter content and grain size (as described in Ribeiro et al. (2011)). The natural seawater used was also obtained from the NW Portuguese coast and filtered through activated carbon before entering the system to prevent the input of impurities. The salinity, pH and nutrient levels were also assessed (as described in Almeida et al., 2012). 2.2. Experimental design The experiments were conducted in a system composed of 9 columns divided into 3 sets of 3 columns each (Fig. 1). Each set was individually connected to an automatic pump system, which was connected to a container filled with seawater, working independently of the other sets. At the base of each column, an inlet was connected to an electric water pump, and an outlet was connected to an electrovalve. The inlet and outlet were controlled by a timer (activated independently every 12 h) to induce the water

Following the assembly of the sand columns, the system was left to equilibrate, with tide simulation, for 3 days. After that period, the treatments (performed in triplicate) began. Accounting for the possible variability among the 3 sets of columns from the filling and draining of each set, each replicate of each treatment was placed in a different set (Fig. 1). The treatment solutions were added when the sand columns were void of water. Three sets were assembled as follows: (NA) Natural attenuation: action of indigenous microorganisms naturally present in the sediment; (BS) Biostimulation treatment: addition of the oleophilic fertilizer S200 to enhance the activity of indigenous oil-microorganisms degraders; and (BA) Bioaugmentation treatment: inoculation of an indigenous oil-degrader consortium pre-stimulated in the laboratory, plus S200. Abiotic losses were not controlled and were assumed to be identical in all treatments. The results were interpreted using comparisons of the two tested treatments with natural attenuation. The commercial fertilizer S200 is a widely used bioremediation agent designed to adhere to oil (Díez et al., 2005; Gallego et al., 2006; Jimenez et al., 2006). It contains a saturated solution of urea (nitrogen source) in oleic acid with phosphate esters (phosphorous source) (Díez et al., 2005). The fertilizer solution was applied according to instructions from the manufacturer to attain a final molar ratio of C:N:P equivalent to 120:10:1 in each column. Assuming that the only source of carbon was provided by the addition of crude oil, the ratio of carbon:S200 was 10:1. For the BA treatment, an inoculum of an indigenous oildegrader consortium pre-stimulated in the laboratory, plus S200, was added at the beginning of the experiment. In fact, the BA

260

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

(A)

(B)

Fig. 1. (A) System composed of 9 columns. All columns were covered with a black cloth to prevent photo-oxidation. (B) Schematic of one of the columns. High tidal level (HTL) is the level of water when the sand was fully saturated with seawater, and low tidal level (LTL) represents the column without water.

treatment was actually the BA treatment plus the BS treatment. The inoculation of pre-stimulated microorganisms can be more effective if it is supplemented with nutrients (Jimenez et al., 2006; Xu et al., 2004). The fertilizer was added to the inoculum just before addition to the sand columns. The inoculum for the BA treatment was prepared by mixing non-contaminated sediment (from the São Pedro da Maceda beach) with Bushnell Hans medium (supplemented with 2% NaCl and 20 mM NO3, in KNO3 form) and crude oil, at a final sediment:medium:oil ratio of 20:40:1 (v/v/v). This mixture was kept in constant agitation in the dark and at room temperature for 4 days. This microbial inoculum presented an abundance of 8.3 log10 cells ml1 of a very diverse community composed of 270 operational taxonomic units, as determined using an Automated rRNA Intergenic Spacer Analysis (unpublished results). Afterwards, this inoculum (a sediment slurry) was mixed with seawater in a 1:2 ratio and was then added to each BA column. During the experiment, a solution containing the S200 fertilizer dissolved in seawater was applied once a week to both the BS and BA treatments to prevent nutrient depletion. 2.4. Sample collection during the assay Each column had holes at the bottom, middle and top (Fig. 1), which allowed for the collection of sand samples from the contaminated layer (at the middle) and from the layers of sand above and below the contaminated one, allowing the evaluation of a possible spread of hydrocarbons over the sand column during the assay. A sediment portion (approximately 15 g) from each column (at different heights) was consistently sampled when the column had no water. The sampling was performed at the beginning of the assay (T0) and after 7 (T7), 15 (T15), 30 (T30), 60 (T60), 90 (T90), 120 (T120) and 180 (T180) days. Thus, the experiment lasted 6 months. To check for the possible losses of hydrocarbons due to the water flow, seawater effluents were also sampled in the first two weeks of the experiment (before and after treatment solution addition and after 7 (T7) and 15 days (T15)). The sampling material was washed with Teepol detergent and rinsed several times with natural seawater. Sediments were analyzed for hydrocarbon contents and microbial abundance determination, whereas seawater samples were analyzed only for the hydrocarbon measurements, as described below. For the hydrocarbon analysis, sediments (stored in aluminum

foil) and water were first frozen (20 °C) to reduce microbial degradation activity and to prevent the loss of volatile compounds. For microbial determinations, samples were stored at 4 °C until processed.

2.5. Hydrocarbon measurements To obtain total petroleum hydrocarbon (TPHs) measurements, a previously optimized method was used (Couto et al., 2013). Briefly, thawed sediment or water were mixed with tetrachloroethylene (P99% spectrophotometric grade, Sigma–Aldrich) (1:10 (w/v)), and an ultrasonic (Elma, Transsonic 460/H model) extraction was performed for 30 min. The extracts were then cleaned with deactivated silica gel (70–230 mesh, Macherey–Nagel). The sample extracts were analyzed by Fourier transform infrared spectrophotometry (Jasco FT/IR-460 Plus) using a quartz cell with a path length of 10 mm (Infrasil I, Starna Scientific). To obtain a chemical profile of the hydrocarbons, analyses by gas chromatography with flame ionization detection (GC-FID) of the sediments were also performed. Sediments were ultrasonically extracted with a mixture of n-hexane (95%, UV-IR-HPLC, PAI-ACS, Panreac) and acetone (P99.9%, CHROMASOLVÒ plus, HPLC, Sigma–Aldrich) (1:1 (v/v)) for 30 min. The extracts were cleaned with FlorisilÒ (60–100 mesh, Fluka) and analyzed by GC-FID (Varian 3800) with a 30 m  0.25 mm  0.25 lm column (Varian Factor Four, VF-5ht), following the conditions described in the literature (Saari et al., 2007). The hydrocarbon range window measured was between nonane (C9) and tetracontane (C40).

2.6. Microbial analysis The most probable number (MPN) of culturable hydrocarbon degrading microorganisms (HD) in sediments was determined according to the protocol developed by Wrenn and Venosa (1996). Pre-filtered (0.2 lm) Arabian Light crude oil was the selective substrate, and BH medium supplemented with 2% NaCl was used as the growth medium. MPN plates were incubated for 2 weeks at room temperature. After incubation, filtered and sterilized tetrazolium salt (p-iodophenyl-3-p-nitrophenyl)-5-phenyl tetrazolium chloride (INT) (3 g L1) was added to each plate well. Positive wells were scored after overnight incubation at room temperature.

261

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

2.7. Statistical analysis

natural degradation

biostimulation

bioaugmentation

TPHs concentrations obtained by FT/IR analysis and HD abundance (log10 MPN g1 wet sediment) in sediment samples were comparatively evaluated. Sediment samples from each column were independently analyzed. The mean values and respective standard deviations were calculated for each set of three replicates. Statistically significant differences (p < 0.05) were evaluated using a parametric one-way ANOVA (analysis of variance). If any significant difference was detected, a multiple Tukey comparison test was performed to locate the difference. All statistical tests were performed using the commercial software SPSS.

TPH (mg/g)

15

10

5

0

3. Results 3.1. Seawater and sediment characteristics The collected sediment was an oxic sand with a very low organic matter content (0.13 ± 0.03%), a grain size higher than 0.25 mm, an HD abundance of 2.4 ± 0.3 log10 MPN g1 and TPHs content below the detection limit (LOD = 0.033 mg g1). The natural seawater used in the experiment exhibited an average salinity of 35, a slightly alkaline pH (pH of 8), nutrient levels of 1 0.52 mg L1 NO3, 5.0 lg L1 NO2, 25 lg L1NHþ 4 and 62 lg L PO3 and TPH levels below the limit of detection 4 (LOD = 3.3 mg L1). 3.2. Periodicity of nutrient addition The total nitrogen values in the seawater ranged from ca. 0.73 mg L1 before entering the columns to ca. 0.55 mg L1 after draining out. The total nitrogen in the S200 solution was ca. 237 mg L1 before entering the columns and approximately 137 mg L1 after draining out. After one week, the water coming out of the columns of the BS and BA treatments had ca. 1.5 mg L1 of total nitrogen, indicating that the S200 solution should be added every week. 3.3. Hydrocarbon contents throughout the experiment The effectiveness of the different treatments on the degradation of hydrocarbons was evaluated using the TPH levels in the contaminated middle sand layer throughout the experiment and through complementary chromatogram profiles of saturated hydrocarbons. 3.3.1. TPH concentrations in sediments The TPH concentrations in the contaminated sediment middle layer revealed that the hydrocarbon elimination/degradation occurred throughout the assay (Fig. 2). In fact, after 7 days, the TPH concentrations were reduced by ca. 50% in all treatments. The marked effect of the BA treatment occurred mainly after 60 days, which is when the TPH degradation reached approximately 80%, indicating that this treatment had a faster hydrocarbon degradation rate (although the differences were not statistically significant, p > 0.05). Significant differences among treatments (p < 0.05) were observed only after 120 days, and the TPH decrease was more accentuated in the BS and BA amendments (significantly lower TPH concentrations in BS and BA treatments than in NA), with degradations levels of 93% and 96%, respectively. At the end of experiment (T180), the TPH removal in all treatments was similar, reaching 93%, 97% and 96% in the NA, BS and BA treatments, respectively. The variability in the TPH concentrations among the replicates contributed to the high standard deviation of the mean values re-

T0

T7

T15

T30

T60

T120

T180

Days Fig. 2. TPHs concentrations throughout the 180-day experimental period, in the natural attenuation scenario and in the biostimulation and bioaugmentation treatments (mean values and respective standard deviations, n = 3).

corded (Fig. 2). This variability could be related to differences among the water flows in the three sets of columns. Water flow can influence the oil and/or nutrient diffusion (Bernabeu et al., 2010), leading to a heterogeneity of conditions, influence on the composition of local bacterial community (Head et al., 2006) and variable hydrocarbon removal rates. However, the triplicate replicates (receiving the same treatment) indicated that TPH removal, in general, had similar patterns throughout the course of the assay. 3.3.2. Spatial distribution of hydrocarbons The TPHs were analyzed not only in the middle contaminated sand layer but also in the sand layers below and above the contaminated layer to assess the possible diffusion and spatial distribution of hydrocarbons in the sand column. Overall, the diffusion of oil was different among microcosms receiving different nutritional treatments (Table 1). In NA, TPHs were detected in the bottom sand layer until day 60, but in both the BS and BA treatments, TPHs were registered only up to 15 or 30 days (the TPH concentrations in this layer were 15–50 times lower than those in the middle layer). In comparison to the top layers, TPHs were detected in NA after 7 days and up to 30 days (TPHs ca. 0.20 mg g1). Conversely, in both the BS and BA amendments, TPHs were detected mainly after 30 days, although they were detected in only one replicate per treatment. 3.3.3. TPHs in water effluents TPHs were measured in the water coming out of the columns before (T0) and after (T1, after a high tide simulation) the addition of the treatments and after 7 (T7), 15 (T15) and 30 (T30) days. Before treatment addition, TPHs were generally detected in the water effluents of the different columns (TPHs at T0 ca. 7 mg L1). For NA, TPHs were detected not only at T1 but also at T7 in water effluents, with similar values to those obtained at T0. After 15 days, no more TPH losses were observed (at T15 and T30, the TPH concentrations were below the detection limit). For the water effluents at T1 in the BS treatment, TPHs were generally below the detection limit (TPHs were only detected in the effluent of one replicate). However, for the BA treatment, TPHs in the water effluents at T1 were ca. 24 mg L1, with a high variability among replicates. These high TPH levels likely occurred because the inoculum solution contained oil-hydrocarbons (the TPH concentration in the inoculum solutions was 60 ± 4 mg L1). After a full 7 days, no TPHs were detected in the water effluents of these two treatments as the TPH concentrations were below the detection limit at T7, T15 and T30.

262

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

Table 1 TPHs concentrations (mg g1, mean values and respective standard deviations, n = 3) along the 180-day experimental period in the sand layers above (top) and below (bottom) the petroleum contaminated middle sand layer in natural attenuation and in the biostimulation and bioaugmentation treatments. Time of the experiment (days)

Sand layer

Natural attenuation

Biostimulation

Bioaugmentation

T7

Above Below




T15

Above Below Above Below

0.18 ± 0.09 0.4 ± 0.3 0.20 ± 0.05 0.3 ± 0.2



T60

Above Below




T120

Above Below




T180

Above Below




T30

a b

Limit of detection (LOD) of 0.033 mg g1. TPHs were detected in one of the replicates (ca. 0.15 mg g1).

3.3.4. Profiles of hydrocarbon contamination The identification of individual components present in a sample contaminated with oil is not viable using FT/IR; however, GC is a good alternative that provides information about the hydrocarbon profile. In fact, each type of oil has a different hydrocarbon distribution that is considered its ‘‘oil fingerprint’’. The portion of sediment artificially contaminated with oil (T0) had an n-alkanes distribution mainly between C10 and C22, with well-resolved peaks (Fig. 3). The distribution extended to C40, with the resolved peaks becoming smaller as they approached C40. In addition, a large hump of unresolved mixture of hydrocarbons was observed. Chromatogram profiles of replicates (same treatment at a given time period) exhibited similar responses (results not shown). In

fact, despite some differences in the resolved peak intensities, the same distribution pattern in all replicates was generally found. Pertaining to the obtained results, only one replicate per treatment was processed during the experiment, each of which was consistently taken from the same column set. Replicate variability was already expected, as observed from the TPH values. Increasing biodegradation effects were visible in the chromatograms and were mainly focused on lighter n-alkanes (C10–C22) (Fig. 3). The degradation of n-alkanes was considerable despite the lack of evidence from the TPH values. The differences observed between the results obtained using FT/IR and the GC/FID analyses could be related to the capacity of the FT/IR method to measure a more extensive range of hydrocarbons than the GC/FID method.

Fig. 3. Chromatograms of non-polar hydrocarbons within the range C9H20 (n-nonane, retention time 9 min) –C40H82 (n-tetracontane, retention time 29 min) observed in the sand layer contaminated with petroleum in natural attenuation (NA) and the biostimulation (BS) and bioaugmentation (BA) treatments after contamination (T0) and after 7 (T7), 30 (T30), 60 (T60), 120 (T120) and 180 (T180) days.

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

Therefore, a TPH FT/IR analysis can include more compounds than a GC analysis, which leads to the apparent difference in the results between the methods of analysis. In addition, the chromatograms did not reveal the resolved peaks from the hydrocarbons that were included in the TPH data. A significant decrease in hydrocarbon contents in all treatments was found after just 7 days, with a similar decrease within treatments (Fig. 3) in accordance with TPH results. At day 30, a further decrease in the resolved and unresolved peaks occurred in all amendments. However, at this time, hydrocarbon removal was higher in the BA and BS amendments than in NA. In addition, the decrease of n-alkanes was much more pronounced in the BA treatment, with a lower intensity in the peaks. After 60 days, a major decrease of hydrocarbon concentrations occurred in the C10–C40 range for the BS and BA treatments compared to NA. In fact, for the BS treatment, only a few resolved peaks were observed in the chromatogram. However, no peaks were found for the BA treatment, which implies that a complete removal of saturated hydrocarbons was achieved. A significant decrease in the intensity of the resolved and unresolved peaks in NA was only observed after 120 days. At the end of experiment (T180), chromatographic analysis revealed a complete removal of n-alkanes from both the BS and BA treatments and a significant degradation of saturated hydrocarbons in NA. Consequently, a gradual disappearance of the n-alkanes between C10 and C40 throughout the experimental period was observed. The n-alkanes were completely depleted in the BS and BA treatments by the end of experiment, whereas the NA profile still presented a few peaks. Chromatographic analysis clearly indicated that bioaugmentation increased the degradation rate of the saturated hydrocarbons, with the earliest effects observed after 30 days. Conversely, the efficaciousness of biostimulation was only verified after 60 days. In NA, larger differences were registered only after 120 days into the experiment. 3.4. Most probable number of hydrocarbon degraders After the addition of the oil to the sediment, the HD abundance rose to about double the amount initially present in the sediment (T0, Fig. 4). The mean values of HD abundance were always constant in NA, whereas a larger variation along the assay occurred in the BS and BA treatments. An increase of HD abundance was registered in the BA treatment from day 30 (T30) until day 60 (T60), which is when it reached methodological saturation (Fig. 4), a time period

Natural attenuation

14

Biostimulation

*

Log 10 MPN g -1

12

Bioaugmentation

*

10 8 6 4 2 0 T0

T7

T15

T30

T60

T120

T180

Days Fig. 4. Hydrocarbon degrading microorganisms estimated using the most probable number (log10 MPN g1, mean and standard deviation, n = 3) in sand contaminated with petroleum over time. *: methodological saturation.

263

in which a significant reduction in the saturated hydrocarbons was observed. The same increase occurred in BS, but only from day 60 (T60) to day 120 (T120). Therefore, both the BS and BA treatments were effective in enhancing microbial oil-degrading assemblages in the sediment-contaminated layer compared to NA. At 120 days (T120), the abundance of HD abruptly decreased in the BA treatment and continued to decline until the end of experiment, a period in which no saturated hydrocarbons were detected at all and in which a significant decrease in TPHs concentration was observed. This behavior was also observed for the BS treatment but only after 180 days. In fact, in the BA and BS amendments after 180 days (T180), the HD levels were practically identical to those in NA.

4. Discussion The persistence of oil at depths up to 4 m could be found in sandy beaches affected by the Prestige oil spill (Bernabeu et al., 2006, 2009). The existence of this subsurface, or ‘‘deep’’, contamination has a direct repercussion on the present cleaning strategy and the recovery of beaches. Bioremediation, a cost-effective technology that eliminates or reduces the concentration of contaminants, is frequently used in petroleum clean-up, with successful results. However, to our knowledge, it has not yet been evaluated for its use in cleaning buried oil. The present work focused on buried oil and tested biostimulation and bioaugmentation techniques to increase the yield of hydrocarbon removal. Despite the gaps and differences between laboratory and field studies, laboratory assays provide useful information for the manipulation of environmental factors and are a very reliable first option for the assessment of oil bioremediation (Jimenez et al., 2006). Therefore, laboratory-controlled conditions were used in the present work. The experiments were designed to simulate the conditions of the middle intertidal area of sandy beaches. The contamination was reproduced through the burial of a portion of sand that had previously been mixed with oil in a sand column that experienced the effects of simulated tidal cycles with natural seawater. In the present work, a reduction of approximately 50% in the hydrocarbon concentrations in the contaminated middle sand layer was observed in all treatments, including NA, after just 7 days. After 7 days, a small but continuous decrease of hydrocarbon concentrations occurred until the end of the experiment. This degradation kinetic was previously observed in other studies but only with superficial oil contamination (Bento et al., 2005). There are three main pathways for the loss of hydrocarbons from sediments: evaporation, leaching and biodegradation. In our case, the evaporation was negligible because the mixture of the oil with the sediment was weathered for 2 days, and the oil was buried, which prevented evaporation processes. The marked decrease in hydrocarbon concentrations after 7 days was most likely associated with lixiviation, as hydrocarbons were detected in the effluents. The continuous flux of seawater provided by the simulation of the tidal cycles led to the diffusion of the oil along the sand column and to hydrocarbon losses. These losses could also justify the lower TPH concentrations observed after 7 days in NA compared to the BS and BA amendments, as hydrocarbons were only detected in both the BS and BA amendments effluents after the first day. This may be related to the particular characteristics of the added fertilizer. S200 is an oleophilic fertilizer with characteristics that allow adhesion to the oil, and it therefore exhibits a great resistance to water action by avoiding quick dilution and becoming easily washed out with tide cycle simulations (Nikolopoulou and Kalogerakis, 2009). Thus, the S200 fertilizer may have fixed the oil by not allowing it to be swept away

264

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

with the water flux in the BS and BA treatments where it was applied. The particular characteristics of this fertilizer were also observed in the distribution of hydrocarbons along the sand column. Generally, hydrocarbons were detected in the bottom sand layer for a longer period of time than in the top sand layer. In NA, hydrocarbons were detected for a month more in the bottom of the column compared to the BA and BS treatments. However, these hydrocarbon losses by lixiviation and by distribution of oil along the sand column were low and did not significantly affect the comparison among treatments. In fact, it is thought that after the first week, the continuous decrease in the hydrocarbon concentrations until the end of the experiments was mainly due to hydrocarbon biodegradation. The differences observed among treatments, particularly in the last samplings, were due to the differences among hydrocarbon biodegradation rates. The HD abundance increased in the BS and BA amendments compared to the HD value in the contaminated sand; this difference was likely due to the addition of the fertilizer. Furthermore, the highest increase of HD abundance recorded in the BA treatment was likely due to the HD already present in the added inoculums; these microorganisms were able to reach the middle layer of contaminated sand. In both the BS and BA amendments, treatment solutions were added to the top of the sand layer to simulate what really occurs in an actual environment and confirming, with the obtained results, that both nutrients and the added microorganisms’ consortium reached the buried oil layer. The HD abundance increase was consistently related to a hydrocarbon concentration decrease. In fact, after 30 days, hydrocarbon removal was higher in the BA and BS amendments than in NA. The degradation primarily occurred in the lighter fraction of n-alkanes (C10–C22), which has been commonly observed in other studies of surface oil contamination (Bento et al., 2005). Because saturated hydrocarbons constitute the largest fraction of crude oil by mass, their biodegradation is quantitatively the most important process in the removal of crude oil from the environment (Head et al., 2006). This effect was more pronounced in the BA treatment, in which there was an increase in the HD abundance compared to the other treatments. This increase attained a maximum after 60 days; at this time, an almost complete removal of saturated hydrocarbons and a decrease in TPH concentration was observed. Regarding the BS treatment, an HD abundance increase occurred after 60 days (a delay of 1 month relatively to the BA treatment) and reached a maximum after 120 days, which is the time period that exhibited the greatest decrease in hydrocarbon concentrations. After 120 and 180 days in the BA and BS treatments, respectively, the HD abundance declined, most likely due to the decline of the carbon source because the hydrocarbon levels were also quite reduced at these times. At the end of experiment (T180), the extension of hydrocarbon degradation in all treatments was similar, which indicates that the nutrients that are naturally present in seawater were sufficient to produce the same so-called ‘‘bioremediation endpoint’’ (Head et al., 2006). Biodegradation in NA, although occurring at slower rates, revealed that autochthonous microorganisms had the capacity to both tolerate and degrade the added oil, a feature already reported for superficial oil contamination (Jimenez et al., 2006). Assemblages of oil-degrading microorganisms previously exposed to oil spills or at sites with chronic pollution exhibit a higher percentage of hydrocarbon degraders and, consequently, higher biodegradation rates compared to the microbial communities of pristine environments (Zhu et al., 2001). However, even the microbial communities of pristine environments exhibit tolerance to environmental modifications associated with oil pollution and display a capacity to degrade oil (Bordenave et al., 2007). For instance, the São Pedro da Maceda beach is considered an un-impacted system, and the sand microbial communities present in the intertidal

beach zone displayed a potential for the degradation of petroleum hydrocarbons in suitable controlled conditions (Almeida et al., 2012). The sediment of this beach was used in the present experiments, in which the microorganisms exhibited a response to the oil contamination by increasing the HD abundance. In fact, the HD abundance doubled when the sand was mixed with petroleum. In addition, the HD abundance was similar in NA throughout the experimental time period, indicating that the nutrients naturally present in the seawater were enough to maintain the degraders. Thus, the presence of hydrocarbon-degrading microorganisms in the initial sediment collected at the São Pedro Maceda beach appears to have contributed to the occurrence of biodegradation, although natural attenuation could have been different if seawater with a different amount of nutrients was used. Nevertheless, both the BA and BS amendments clearly contributed to a faster removal of oil-hydrocarbons from the sand than NA. It was previously reported that the addition of oleophilic fertilizers as nitrogen and phosphorous sources, such as the S200 fertilizer, significantly enhanced the extension of oil-biodegradation in laboratory (Díez et al., 2005) and in situ experiments (Gallego et al., 2006; Jimenez et al., 2006) at the surface or shallow depths because of the reduction in the response period of the degrading microorganisms (lag period) (Sabaté et al., 2004). This lag period can be further reduced if bioaugmentation is used. Bioaugmentation is supposed to overcome the lag phase that indigenous microbial assemblages often require, i.e., the time that microorganisms take to respond to the presence of a contaminant in the environment during natural attenuation processes, which leads to a prolonged bioremediation process (Lendvay et al., 2003). In fact, in the present work, the addition of a pre-stimulated inoculum (BA treatment) resulted in a faster hydrocarbon removal rate, with a high hydrocarbon concentration decrease occurring earlier in the assay, whereas the effects of the addition of the S200 fertilizer alone (BS amendment) were verified later. In the present case, a notorious BA effect was visible after 30 days. In the present work for BA, a microbial consortium with indigenous oil-degrader microorganisms (stimulated under optimal nutritional conditions in the lab) from the collected sand was added to the respective sand columns. From an applied perspective, using an autochthonous microbial consortium rather than a pure culture for bioremediation is more advantageous because it provides the metabolic diversity and robustness needed for field applications (Tyagi et al., 2011). It is clear that individual microorganisms can metabolize only a limited range of hydrocarbon substrates, so mixed microbial assemblages with broad enzymatic capacities and synergistic/co-metabolic relationships are required to attack different compounds. Moreover, the use of autochthonous microorganisms avoids the ecological and ethical problems associated with the addition of exogenous or genetically modified organisms. Both the bioaugmentation and biostimulation techniques, when applied individually, can present limitations. Therefore, these techniques can be used simultaneously for the remediation of hazardous compounds, namely oil pollutants (Nikolopoulou and Kalogerakis, 2009; Tyagi et al., 2011), thus resulting in a faster hydrocarbon degradation rate, as observed in this work.

5. Conclusions The selection and stimulation of oil-degrader microorganisms from the selected beach, coupled with optimal environmental conditions (i.e., the addition of oleophilic fertilizer to enhance biodegradation and avoid oil diffusion), was the best approach for the effective remediation of buried oil. These results indicate that bioremediation techniques can be used to increase the pace of bur-

J. Pontes et al. / Marine Pollution Bulletin 76 (2013) 258–265

ied oil remediation. It should be emphasized that the microorganisms and nutrients added to the surface of the sand were able to reach the deep buried oil layer. This was the first study to demonstrate that autochthonous BA with suitable nutritive conditions can be a cost-effective methodology for buried oil remediation that contributes to oil removal and to the recovery of an area impacted by oil spills.

Acknowledgments This research was partially supported by the European Regional Development Fund (ERDF) through COMPETE – Operational Competitiveness Program and by national funds through FCT – Foundation for Science and Technology, under project ‘‘PEst-C/MAR/ LA0015/2011’’ and through project ERA-AMPERA/003/2007. We would also like to thank the Spanish MICINN (Contract No. ERACCT2005-016165) and FCT for the Izabela Reis BI scholarship (OILDEBEACH-AMPERA/2008-016), co-financed by POPH/FSE.

References Albaigés, J., Morales-Nin, B., Vilas, F., 2006. The prestige oil spill: a scientific response. Mar. Pollut. Bull. 53, 205–207. Almeida, C.M., Reis, I., Couto, M.N., Bordalo, A.A., Mucha, A., 2012. Potential of the microbial community present in an unimpacted beach sediment to remediate petroleum hydrocarbons. Environ. Sci. Pollut. Res. 9, 1–9. Atlas, R.M., Hazen, T.C., 2011. Oil biodegradation and bioremediation: a tale of the two worst spills in U.S. history. Environ. Sci. Technol. 45, 6709–6715. Bento, F.M., Camargo, F.A.O., Okeke, B.C., Frankenberger, W.T., 2005. Comparative bioremediation of soils contaminated with diesel oil by natural attenuation, biostimulation and bioaugmentation. Bioresour. Technol. 96, 1049–1055. Bernabeu, A.M., Nuez de la Fuente, M., Rey, D., Rubio, B., Vilas, F., Medina, R., et al., 2006. Beach morphodynamics forcements in oiled shorelines: coupled physical and chemical processes during and after fuel burial. Mar. Pollut. Bull. 52, 1156– 1168. Bernabeu, A.M., Rey, D., Rubio, B., Vilas, F., Domínguez, C., Bayona, J.M., et al., 2009. Assessment of cleanup needs of oiled sandy beaches: lessons from the prestige oil spill. Environ. Sci. Technol. 43, 2470–2475. Bernabeu, A.M., Rey, D., Lago, A., Vilas, F., 2010. Simulating the influence of physicochemical parameters on subsurface oil on beaches: preliminary results. Mar. Pollut. Bull. 60, 1170–1174. Bordenave, S., Goñi-Urriza, M.S., Caumette, P., Duran, R., 2007. Effects of heavy fuel oil on the bacterial community structure of a pristine microbial mat. Appl. Environ. Microbiol. 73, 6089–6097. Couto, M.N.P.F., Borges, J.R., Guedes, P.R., Almeida, R., Monteiro, E., Almeida, C.M.R., Basto, M.C.R.P., Vasconcelos, M.T.S.D., et al., 2013. An improved method for determination of petroleum hydrocarbons from soil using a simple extraction process and Fourier transform infrared spectrophotometry. Petrol. Sci. Technol.

265

Díez, S., Sabaté, J., Viñas, M., Bayona, J.M., Solanas, A.M., Albaigés, J., 2005. The prestige oil spill. I. Biodegradation of a heavy fuel oil under simulated conditions. Environ. Toxicol. Chem. 24, 2203–2217. Farhadian, M., Vachelard, C., Duchez, D., Larroche, C., 2008. In situ bioremediation of monoaromatic pollutants in groundwater: a review. Bioresour. Technol. 99, 5296–5308. Gallego, J.R., González-Rojas, E., Peláez, A.I., Sánchez, J., García-Martínez, M.J., Ortiz, J.E., et al., 2006. Natural attenuation and bioremediation of prestige fuel oil along the Atlantic coast of Galicia (Spain). Org. Geochem. 37, 1869–1884. Gallego, J.R., Fernández, J.R., Díez-Sanz, F., Ordoñez, S., Sastre, H., González-Rojas, E., et al., 2007. Bioremediation for shoreline cleanup: in situ vs. on-site treatments. Environ. Eng. Sci. 24, 493–504. Head, I.M., Swannell, R.P.J., 1999. Bioremediation of petroleum hydrocarbon contaminants in marine habitats. Curr. Opin. Biotechnol. 10, 234–239. Head, I.M., Jones, D.M., Roling, W.F.M., 2006. Marine microorganisms make a meal of oil. Nat. Rev. Microbiol. 4, 173–182. Hosokawa, R., Nagai, M., Morikawa, M., Okuyama, H., 2009. Autochthonous bioaugmentation and its possible application to oil spills. World J. Microbiol. Biotechnol. 25, 1519–1528. Jimenez, N., Vinas, M., Sabate, J., Diez, S., Bayona, J.M., Solanas, A.M., et al., 2006. The prestige oil spill. 2. Enhanced biodegradation of a heavy fuel oil under field conditions by the use of an oleophilic fertilizer. Environ. Sci. Technol. 40, 2578–2585. Lendvay, J.M., Löffler, F.E., Dollhopf, M., Aiello, M.R., Daniels, G., Fathepure, B.Z., et al., 2003. Bioreactive barriers: a comparison of bioaugmentation and biostimulation for chlorinated solvent remediation. Environ. Sci. Technol. 37, 1422–1431. Nikolopoulou, M., Kalogerakis, N., 2009. Biostimulation strategies for fresh and chronically polluted marine environments with petroleum hydrocarbons. J. Chem. Technol. Biotechnol. 84, 802–807. Ribeiro, H., Mucha, A., Almeida, C.M., Bordalo, A., 2011. Hydrocarbon degradation potential of salt marsh plant–microorganisms associations. Biodegradation 22, 729–739. Röling, W.F.M., Milner, M.G., Jones, D.M., Lee, K., Daniel, F., Swannell, R.J.P., et al., 2002. Robust hydrocarbon degradation and dynamics of bacterial communities during nutrient-enhanced oil spill bioremediation. Appl. Environ. Microbiol. 68, 5537–5548. Saari, E., Perämäki, P., Jalonen, J., 2007. A comparative study of solvent extraction of total petroleum hydrocarbons in soil. Microchim. Acta 158, 261–268. Sabaté, J., Viñas, M., Solanas, A.M., 2004. Laboratory-scale bioremediation experiments on hydrocarbon-contaminated soils. Int. Biodeterior. Biodegr. 54, 19–25. Swannell, R.P., Lee, K., McDonagh, M., 1996. Field evaluations of marine oil spill bioremediation. Microbiol. Rev. 60, 342–365. Tyagi, M., da Fonseca, M., de Carvalho, C., 2011. Bioaugmentation and biostimulation strategies to improve the effectiveness of bioremediation processes. Biodegradation 22, 231–241. Van Stempvoort, D., Biggar, K., 2008. Potential for bioremediation of petroleum hydrocarbons in groundwater under cold climate conditions: a review. Cold Reg. Sci. Technol. 53, 16–41. Wrenn, B., Venosa, A., 1996. Selective enumeration of aromatic and aliphatic hydrocarbon degrading bacteria by a most-probable number procedure. Can. J. Microbiol. 42, 252–258. Xu, R., Lau, N.L.A., Ng, K.L., Obbard, J.P., 2004. Application of a slow-release fertilizer for oil bioremediation in beach sediment. J. Environ. Qual. 33, 1210–1216. Zhu, X., Venosa, A.D., Suidan, M.T., Lee, K., 2001. Guidelines for the bioremediation of marine shorelines and freshwater wetlands. U.S. Environmental Protection Agency.