Environmental Pollution 134 (2005) 267–276 www.elsevier.com/locate/envpol Oil spill in the Rı́o de la Plata estuary, Argentina: 2-hydrocarbon disappearance rates in sediments and soils J.C. Colomboa,b,*, A. Barredaa, C. Bilosa, N. Cappellettia, M.C. Migoyaa, C. Skorupkaa aLaboratorio de Quı́mica Ambiental y Biogeoquı́mica, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata, Av. Calchaqui km 23500 (1888) Florencio Varela, Buenos Aires, Argentina bComisión de Investigaciones Cientı́ficas, Provincia de Buenos Aires, Argentina Received 14 October 2003; accepted 30 July 2004 Hydrocarbon disappearance rates in Rı́o de la Plata sediments and soils were determined. Abstract The 6-month assessment of the oil spill impact in the Rı́o de la Plata described in the preceding paper [Colombo, J.C., Barreda, A., Bilos, C., Cappelletti, N., Demichelis, S., Lombardi, P., Migoya, M.C., Skorupka, C., Suárez, G., 2004. Oil spill in the Rı́o de la Plata estuary, Argentina: 1 – biogeochemical assessment of waters, sediments, soils and biota. Environmental Pollution] was followed by a 13- and 42-month campaigns to evaluate the progress of hydrocarbon decay. Average sediment hydrocarbon concentrations in each sampling include high variability (85–260%) due to contrasting site conditions, but reflect a significant overall decrease after 3 years of the spill: 17G 27, 18G 39 to 0.54G 1.4 mg g�1 for aliphatics; 0.44G 0.49, 0.99G 1.6 to 0.04G 0.03 mg g�1 for aromatics at 6, 13 and 42 months, respectively. Average soil hydrocarbon levels are 100–1000 times higher and less variable (61– 169%) than sediment values, but display a clear attenuation: 3678G 2369, 1880G 1141 to 6.0G 10 mg g�1 for aliphatics and 38G 26, 49G 32 to 0.06G 0.04 mg g�1 for aromatics. Hydrocarbon concentrations modeled to first-order rate equations yield average rate constants of total loss (bioticC abiotic) twice as higher in soils (kZ 0.18–0.19 month�1) relative to sediments (0.08– 0.10 month�1). Individual aliphatic rate constants decrease with increasing molecular weight from 0.21G 0.07 month�1 for isoprenoids and !n-C22 to 0.10G 0.08 month�1 for On-C27, similar to hopanes (0.10G 0.05 month�1). Aromatics disappearance rates were more homogeneous with higher values for methylated relative to unsubstituted species (0.17G 0.05 vs. 0.12G 0.05 months�1). Continued hydrocarbon inputs, either from biogenic (algal n-C15,17; vascular plant n-C27,29) or combustion related sources (fluoranthene and pyrene), appear to contribute to reduced disappearance rate. According to the different loss rates, hydrocarbons showed clear compositional changes from 6–13 to 42 months. Aliphatics disappearance rates and compositional changes support an essentially microbiologically-mediated recovery of coastal sediments to pre-spill conditions in a 3–4 year period. The lower rates and more subtle compositional changes deduced for aromatic components, suggest a stronger incidence of physical removal processes. � 2004 Elsevier Ltd. All rights reserved. Keywords: Oil spill; Hydrocarbons; Degradation; Rı́o de la Plata Doi of original article 10.1016/j.envpol.2004.02.032. * Corresponding author. Fax: C54 11 4275 8266. E-mail address: laqab@arnet.com.ar (J.C. Colombo). 0269-7491/$ - see front matter � 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2004.07.028 1. Introduction The progress of oil disappearance in the aquatic environment is determined by the interaction of well- known petroleum- and environment-specific factors which control the effectiveness of physical, chemical mailto:laqab@arnet.com.ar http://www.elsevier.com/locate/envpol 268 J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 and microbial removal of hydrocarbons, e.g. crude composition, hydro-dynamism, solar irradiation, tem- perature, particle abundance, sediment texture, micro- biological composition and nutrient availability (e.g. Blumer and Sass, 1972; Atlas, 1981; Fusey and Oudot, 1984; Sugiura et al., 1997). The basic knowledge of petroleum geochemistry permitted the development of a solid framework in topics related to hydrocarbon fingerprinting, compositional alteration and weathering stages of oil components spilled in the environment (e.g. Hostettler and Kvenvolden, 1994; Wang et al., 1999, 2003; Pastor et al., 2001; Boehm et al., 2001). However, quantitative information on hydrocarbon removal rates in natural environments under different conditions is much more limited. In the preceding paper (Colombo et al., 2004), the biogeochemical assessment of aliphatic and aromatic hydrocarbons in waters, sediments, soils and biota permitted to elucidate the magnitude of environmental impact produced by the spill of w1000 tons oil in the coastal ecosystem of the Rı́o de la Plata. Results indicated that offshore waters and sediments were little affected due to the rapid wind-driven transport of petroleum residues to the coast. Six months after the spill, coastal waters, sediments, soils and biota still presented very high hydrocarbon levels exceeding by 1–3 orders of magnitude baseline concentrations. Biotic and abiotic compartments consistently indicated that the most impacted area is the central sector close to Magdalena city, which presented the highest hydrocar- bon levels, specially in low-energy stream embouchures and bays which acted as efficient oil traps. In order to follow the environmental evolution of oil residues, two additional campaigns were carried out 13 and 42 months after the spill. This paper presents results on the rates of hydrocarbon attenuation and changes of the aliphatic and aromatic composition during decay under different environmental conditions in most affected sediments and soils of this temperate, freshwater ecosystem. 2. Methods Sampling was carried out on February 2000 and July 2002, 13 and 42 months after the spill, respectively, covering previously visited stations, the location and description of which are included in Table 1 of the preceding paper. An additional station (Alberdi, ALB) situated midway between LB and Be was included in these campaigns (Fig. 1). Sediments and soils were collected with stainless steel spatulas during low tide along 45 km shoreline. All samples contained in pre-cleaned glass jars were preserved in portable coolers until arrival to the laboratory. The analytical scheme included ultrasonic extraction, silica gel frac- tionation and HRGC-FID determination of individual hydrocarbons as detailed in the previous paper (Co- lombo et al., 2004). In addition, selected samples from the three campaigns were re-analyzed for the determi- nation of aliphatic hydrocarbons, including hopanes (C29 norhopane and C30 hopane), and aromatic components by HRGC-MSD using an Agilent 6850 gas chromatograph coupled to an Agilent 5973N MSD (EI 70 eV, 2.94 scans seg�1, 50–550 amu). The column and chromatographic conditions were the same as those used for the HRGC-FID analyses. The presence of hopane biomarkers, together with terpanes was con- firmed by selected ion monitoring at m/z 191 (e.g. Wang and Fingas, 2003). Fig. 2 shows typical chromatograms of aliphatic and aromatic hydrocarbons and hopanes in soils at 6, 13 and 42 months after the spill. 3. Results and discussion 3.1. Total hydrocarbon averages in surface sediments and soils Table 1 and Fig. 3 present total aliphatic and aromatic hydrocarbon concentrations in sediment and soils collected at 6, 13 and 42 months after the spill. General sediment averages include a 85–260% variability during each sampling due to contrasting site conditions, but reflect a significant overall decrease after 3 years of the spill: 17G 27, 18G 39 to 0.54G 1.4 mg g�1 for resolved aliphatics; 0.44G 0.49, 0.99G 1.6 to 0.04G 0.03 mg g�1 for aromatics at 6, 13 and 42 months, respectively. These general averages suggest stable hydrocarbon levels during the first and second campaign and a strong decrease at 42 months. However, hydrocarbon evolution shows great site to site variations and most contaminated stations show more consistent 1–3 orders of magnitude decrease along the three samplings, i.e. GGi, GGe, JBi. In agreement with the attenuation of hydrocarbon levels, the spatial extension of the impacted area is clearly reduced 42 months after the spill. Except some low-energy, heavily polluted sites such as GGi, Ri, JBi, most stations have declined to previous background hydrocarbon concen- trations during the third sampling (Fig. 3). Average soil hydrocarbon levels are 100–1000 times higher and more homogeneous (61–169%) than sediment values, but show a consistent 2–4 orders of magnitude attenuation along the 3 campaigns: 3678G 2369, 1880G 1141 to 6.0G 10 mg g�1 for aliphatics and 38G 26, 49G 32 to 0.06G 0.04 mg g�1 for aromatics. In spite of the large variability of these general means, they suggest a faster drop of aliphatic hydrocarbons relative to aromatic compounds which show comparable or even higher concentrations during the second sampling campaign. Some degree of residue remobiliza- tion from heavily polluted sites can be expected in this 269J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 Table 1 Total resolved and UCM aliphatic and aromatic hydrocarbons in sediments and soils after 6, 13 and 42 months of the spill Sediments (mg g�1) Soils (mg g�1) ALI Petro/biog ALI UCM ARO Petro/pirog ARO UCM ALI Petro/biog ALI UCM ARO Petro/pirog ARO UCM LB6 0.1 0.3 1 – – – LB13 0.03 0.0 – 0.14 0.6 – LB42 0.1 0.1 – – – – ALB6 na na na na na na na na na na na na ALB13 0.3 1.6 – 0.14 0.4 – 2729 28.7 63 590 101 0.4 4401 ALB42 0.03 0.6 – – – – 0.4 0.3 0.03 0.6 Be6 1.0 0.7 20 0.43 0.3 8.9 Be13 0.7 0.1 – 0.19 0.8 – Be42 0.01 1.5 – – – – JI6 0.3 0.7 9 0.01 1.0 0.6 JI13 0.1 0.2 – 0.24 0.1 – J142 0.2 0.3 – – – – JIIC6 948 1.9 15 725 22 0.2 966 JIIC13 518 2.8 11 154 18 0.6 528 JIIC42 27 0.03 21 0.06 0.1 JIIS6 13.1 1.9 220 0.23 0.6 36.6 JIIS13 133.7 6.3 3949 3.76 0.7 200.2 2867 2.5 127 148 69 0.5 7480 JIIS42 0.1 0.5 – – – – 5 0.1 247 0.15 0.2 PN6 1.8 0.6 24 0.02 0.2 2.0 PN13 0.1 1.5 – 0.03 1.2 – PN42 0.1 0.7 – – – – GGi6 86.6 1.9 1445 1.34 0.4 205.7 GGi13 37.4 3.1 1256 2.87 0.2 84.0 GGi42 5.3 0.1 43 0.05 0.1 – GGe6 58.9 1.5 311 0.64 1.1 20.7 5187 1.1 20 218 25 0.6 4130 GGe13 1.6 0.7 155 0.10 0.4 5.2 3073 1.0 56 076 21 0.9 1223 GGe42 0.1 0.7 – 0.004 0.2 – 1 0.1 18 0.03 0.2 Ri6 16.4 1.6 689 1.18 0.2 139.1 4900 1.6 37 178 69 0.6 5751 Ri13 71.6 2.5 3553 4.90 0.5 116.4 775 19.9 37 199 30 1.0 2890 Ri42 0.9 0.3 14 0.08 0.1 – 2 0.1 15 0.05 0.1 Re6 1.0 1.7 27 0.04 0.5 4.2 Re13 0.4 0.4 – 0.16 0.5 – Re42 0.1 0.3 – – – – AL6 5.1 1.5 58 0.08 0.7 7.3 AL13 0.7 0.9 46 0.11 0.5 – 1316 2.0 38 910 53 0.3 2414 AL42 0.1 0.5 – – – – 1 0.1 0.05 0.1 JBi6 34.2 1.6 310 0.78 0.5 30.6 JBi13 2.4 0.3 62 0.90 – – JBi42 0.4 0.2 – 0.02 0.1 – Jbe6 6.6 1.9 68 0.06 0.1 0.9 Jbe13 0.2 0.3 – 0.14 0.2 – Jbe42 0.1 0.2 – – – – PE6 0.02 – – – – – PE13 0.4 0.2 – 0.23 0.1 – PE42 0.04 0.1 – – – – AVG6 17.3 1.3 265 0.44 0.59 41.5 3678 1.5 24 374 38 0.47 3616 SD6 27.0 0.6 423 0.49 0.34 67.5 2369 0.4 11 314 26 0.20 2434 AVG13 17.8 1.3 1504 0.99 0.47 101.5 1880 9.5 55 680 49 0.61 3156 SD13 39.2 1.7 1803 1.61 0.30 80.7 1141 11.8 39 435 32 0.29 2510 AVG42 0.54 0.4 29 0.04 0.13 – 6.0 0.1 75.2 0.06 0.22 – SD42 1.40 0.4 20 0.03 0.08 – 10.2 0.1 114.9 0.04 0.21 – ALI: total resolved aliphatic hydrocarbons; petro/biog: !C22C isopr/C15CC17COC23; ALIUCM: aliphatic unresolved complex mixture; ARO: total resolved aromatic hydrocarbons; petro/pirog: methylated/unsubstituted ARO; AROUCM: aromatic unresolved complex mixture; stations as in table 1 of Colombo et al., 2004; sampled at 6, 13 and 42 months after the spill; –: below quantification limits. 270 J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 JI JIIC PN GGe GGi Ri Re JBe AL JBi PE JIIS Atalaya La Balandra Magdalena Gauchito Gil Ricardo Juan Blanco Pearson Argentina Magdalena Río de la Plata Uruguay 0 50 km Buenos Aires La Plata Be 0 4 km ALB Punta Blanca LB Fig. 1. Study area and station location along Rı́o de la Plata coast. dynamic coastal area where tides and specially wind- driven currents facilitate sediment transport. This physical remobilization would also explain the consis- tent 130–470% increase of the ALIUCM from 6 to 13 months (265G 423 to 1504G 1803 mg g�1 in sediments and 24G 11 to 56G 39 mg g�1 in soils). Since this unresolved hump has been long recognized as indicative of degraded petrogenic residues (e.g. Volkman et al., 1984; Killops and Al-Juboori, 1990), the UCM increase suggests remobilization of heavily biodegraded aliphatic residues. Aromatic hydrocarbons show more conserva- tive levels for both the resolved and UCM fractions (Table 1). The more drastic changes of aliphatic hydrocarbons are consistent with their known higher susceptibility to microbial attack relative to aromatic components (e.g. Atlas, 1981; Colombo et al., 1996; Sugiura et al., 1997). In the last campaign all hydrocar- bon fractions, including the UCM are consistently reduced, reflecting the combined effect of biodegrada- tion and physical removal. 3.2. Modeling of total average hydrocarbon loss rates The disappearance of oil residues in coastal environ- ments has been modeled as an exponential decrease characterized by first-order rate constants (e.g. Page et al., 2002; Grossi et al., 2002). However, in contrast with controlled laboratory or field experiments, in real spill situations possible confounding factors such as continuous inputs from commercial ships, boating, coastal discharges or biogenic sources introduce some uncertainty in the calculations. The oiled area around Magdalena is relatively remote from large urban centers and is not subjected to heavy chronic impact, thus providing a good opportunity to study oil disappearance rates in a natural environment without intensive human intervention. To obtain a first estimation of the rates of hydrocarbon attenuation in this coastal ecosystem, average levels in sediments and soils for the three campaigns were modeled to a first-order rate equation (exponential decrease): HCZHC0 e �kt where HC is the hydrocarbon concentration, HC0 is the initial concentration, k is the first-order rate coefficient (degradation constant: month�1) and t is time (months). Fig. 4 presents the results for average sediment and soil hydrocarbon concentrations along the three campaigns (note the logarithmic scale). Three rate equations were obtained since the AROUCM was not quantifiable in the third sampling. Exponential fittings are in general good (R2Z 0.65–0.99), but results should be interpreted with caution since 3 sampling times is the lowest resolution needed for these models. Also the exponential fit of the ALIUCM is somewhat poorer due to its remobilization-mediated increase in the second cam- paign. The average rate constants which represent total losses (both biotic and abiotic) are comparable for 271J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 Pr Ph Np C29 C30 Hopanes C29 Hopane C20-21 Terpanes C30 Hopane C23-24 Terpanes 14 15 Fn 19 21 23 25 27 23 29 3117 33 Months 6 13 42 35 R el at iv e R es po ns e MChr TMPh Chry DMPh MPyr Fla Months 6 13 42 BzPer Per Pyr Retention time (min) Fig. 2. Chromatograms of aliphatic, aromatic and hopane (m/z Z 191) hydrocarbons in a soil sample (Ricardo) collected after 6, 13 and 42 months of the spill. Fn: farnesane, Np: norpristane, Pr: pristane, Ph: phytane, 14–35: C14–35 n-alkanes, Fla: fluoranthene, Pyr: pyrene, DMPh and TMPh: di and trimethylphenanthrenes, MPyr: methylpyrene, Chry: chrysene, MChr: methylchrysene, Per: perylene, BzPer: benzo( ghi)perylene. 272 J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 0.001 0.010 0.100 1.000 10.000 100.000 1000.000 10000.000 LB AL B Be JI JI IS PN G G i G G e R i R e AL JB i JB e PE AL Bs JI IC s JI IS s G G s R s AL s A L I ( u g / g d w ) 6 13 42 Sediments . Soils 0.001 0.010 0.100 1.000 10.000 100.000 LB AL B Be JI JI IS PN G G i G G e R i R e AL JB i JB e PE AL Bs JI IC s JI IS s G G s R s AL s A R O ( u g / g d w ) 6 13 42 Sediments Soils . Fig. 3. Total resolved aliphatic and aromatic hydrocarbon concentrations in sediments and soils after 6, 13 and 42 months of the spill. aliphatics and aromatics, but they are twice as high in soils relative to sediments (kZ 0.18–0.19 vs. 0.08–0.10 month�1 or 0.006 vs. 0.003 day�1). Due to differences in the degree of oiling and environmental factors, the rate constants of individual sites show large variability ranging from 0.07–0.16 month�1 in sediments to 0.10– 0.31 month�1 in soils. These rate constants are comparable or lower than those reported in the literature for field or laboratory experiments. For example, a field experiment with Arabian light crude oil spread on surface seashore sediments and treated with different fertilizers yielded rate constants of 0.11–0.27 month�1 for saturated and aromatic components (Fusey and Oudot, 1984). Shorter term field experiments performed with chemically dispersed oils in a wetland environment report 2–10 times higher rate constants, i.e. 0.012–0.052 day�1 (Page et al., 2002). Another field experiment performed in bioturbated Mediterranean coastal sediments treated with Arabian light crude oil yielded rate constants of 0.12–0.39 month�1 for C17–30 n-alkanes and isopre- noids (Grossi et al., 2002). In a 3-month biodegradation experiment of 10% crude oil-contaminated soils in- oculated with different microflora, the rate constants ranged from 0.09 to 0.49 month�1 for aliphatic hydro- carbons (Colombo et al., 1996). According to the intercept values of the model, the average initial (time 0) sediment hydrocarbon concentrations would be 1.3 and 45 mg g�1 for ARO and ALI, respectively. These estimated initial levels are about 3 times higher than average values measured 6 months later. The corresponding initial concentrations in soils, are 251 and 14 610 mg g�1, respectively, which are 4–7 times higher than the 6-month levels. The average percent loss from initial concentrations de- termined from the model equations (Fig. 3c) indicate that more than 40% of sediment hydrocarbon residues were lost 6 months after the spill with an almost complete recovery at about 4 years. The more rapid recovery in soils appears to occur at 3 months and less than 3 years, respectively. 3.3. Modeling of individual hydrocarbon loss rates The total average loss rates calculated previously reflect the mean disappearance of aliphatic and aromatic hydrocarbons, but the individual components show different behaviors. The detailed analysis of each fraction by HRGC-MSD in selected soils (Rs, JIICs, GGs) and sediments (GGi) permitted to model com- pound-specific rates, including hopane biomarkers. Fig. 5 shows the rate constants of some aliphatic and aromatic hydrocarbons grouped according to expected sources and reactivity. The results indicate lower disappearance rates in GGi sediments relative to soils, and contrasted differences between compounds. Lower 273J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 molecular weight petrogenic n-alkanes and isoprenoids show the highest rate constants whereas heavier n- alkanes and hopanes disappear 2–3 times more slowly. Aromatic hydrocarbons show more homogeneous, in- termediate rates. Fig. 6 shows the relationship of rate constants and molecular weight of individual aliphatic hydrocarbons. The rates show significant inverse correlations with molecular weight (R2Z 0.47–0.84), and consistently lower values for GGi sediments relative to soils (0.08G 0.05 vs. 0.19G 0.08 month�1). Norpristane, pristane and C15–19 n-alkanes show the highest rates (0.22G 0.07 month�1) whereas higher molecular weight C27–29 n-alkanes present very low values (0.10G 0.08 month�1), similar to hopanes (0.10G 0.05 month�1). The decrease of biodegradation with increasing a y = 1233.22e-0.08x R2 = 0.65 y = 44.97e-0.10x R2 = 0.96 y = 1.31e-0.08x R2 = 0.83 0.01 0.10 1.00 10.00 100.00 1000.00 10000.00 0 10 20 30 40 50 60 S e d i m e n t A v g ( u g / g d w ) ALI ALIUCM ARO AROUCM ALI ALIUCM ARO AROUCM b y = 182398.37e-0.18x R2 = 0.91 y = 14610.40e-0.18x R2 = 0.99 y = 250.91e-0.19x R2 = 0.95 0.01 0.10 1.00 10.00 100.00 1000.00 10000.00 100000.00 0 10 20 30 40 50 60 Time (months) Time (months) S o i l A v g ( u g / g d w ) c 0 20 40 60 80 100 0 10 20 30 40 50 60 Time (months) L o s s ( % ) SoilALI SoilALIUCM SedALI SedALIUCM Fig. 4. Rate loss modeling of average aliphatic and aromatic hydrocarbon concentrations in sediments (a), soils (b) and estimated aliphatic hydrocarbon disappearance (c). molecular weight of aliphatics has been reported for laboratory and field experiments (e.g. Colombo et al., 1996; Grossi et al., 2002), and probably reflects the reduced membrane permeability of heavier molecules (Sugiura et al., 1997). The intense biodegradation of isoprenoids in the Rı́o de la Plata suggests the existence of well-acclimated microbes using hydrocarbons as the sole carbon source. This is favored by the 6-month period elapsed since the spill. In the Aegean Sea spill in Galicia coast, alkanes and isoprenoids were also degraded within 6 months of the accident (Pastor et al., 2001). The continued inputs of biogenic n-C15–17 from periphytic algae growing on macrophytes, sands and coastal soils are the most probable explanation for the apparent lower rates of these compounds (Fig. 6). A similar interference from terrestrial vegetation cuticular wax inputs would explain the low rates of C27,29 n-alkanes compared with n-C26– 28. This highlights the importance of characterizing other natural and anthropogenic hydrocarbon sources to estimate oil disappearance rates. In this context, isoprenoid rates may be considered as a more precise estimation of bulk oil residue attenuation. As hopanes are well known, very slowly-degrading compounds, they are used as ‘‘internal standards’’ to discriminate the relative importance of biodegradation from physical removal of oil residues (e.g. Grossi et al., 2002; Page et al., 2002; Wang and Fingas, 2003). Following the decreasing trend of rate constants with molecular weight (Fig. 6), the importance of biodegra- dation calculated as the difference with hopane values decreases from 58G 14% for isoprenoids and lower molecular weight n-alkanes to 31G 14% for n-C23–26. There is no apparent biodegradation of higher molec- ular weight n-alkanes (9.2G 15%) which disappear at similar rates than hopanes. In GGi sediments, where overall hydrocarbon reduction is lower, the relative importance of biodegradation is higher (60–75% for isoprenoids to n-C21). In contrast, biodegradation is responsible for only 41–64% average of the stronger hydrocarbon decrease observed in soils, indicating enhanced physical removal. The rate modeling of individual aromatic hydro- carbons is limited by the reduced detections in the third campaign. In general, aromatics show more homoge- neous lower values relative to aliphatics (0.04–0.26 month�1), and confirm the lower rates at GGi sedi- ments. Methylated aromatics disappearance rates are somewhat higher relative to unsubstituted compounds (0.17G 0.05 vs. 0.12G 0.05 month�1) but the differ- ences are small, perhaps reflecting the slow disappear- ance of heavy aromatics summed to confounding factors such as continued atmospheric inputs of pyrogenic PAHs. Petrogenic chrysene and methylchrysene present homogenous higher rates (0.17G 0.04 month�1) relative to pyrogenic fluoranthene and pyrene (0.10G 0.02 274 J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 y = 9736.98e-0.30x R2 = 0.99 y = 10902.19e-0.30x R2 = 0.98 y = 2399.53e-0.25x R2 = 0.97 y = 37.70e-0.15x R2 = 0.99 0.01 0.10 0.10 1.00 y =1164.85e-0.16x R2 = 0.95 y = 7671.35e-0.22x R2 = 1.00 y = 379.53e-0.08x R2 = 1.00 y = 8.54e-0.03x R2 = 0.55 1000.00 10000.00 0 10 20 30 40 50 Time (months) Time (months) Time (months) > n - C 2 3 ( u g / g d w ) < n - C 2 2 ( u g / g d w ) 60 10.00 1.00 10.00 100.00 100.00 1000.00 10000.00 0.01 0.10 1.00 10.00 100.00 1000.00 10000.00 0 10 20 30 40 50 I s o p r e n o i d s ( u g / g d w ) 60 y = 311.62e-0.19x R2 = 0.90 y = 5978.01e-0.27x R2 = 1.00 y = 1901.03e-0.19x R2 = 0.98 y = 32.71e-0.12x R2 = 1.00 0 10 20 30 40 50 60 Rs GGs JIICs GGi y = 98.62e-0.14x R2 = 0.96 y = 97.61e-0.14x R2 = 0.95 y = 18.12e-0.08x R2 = 0.98 y = 2.06e-0.04x R2 = 0.98 0.1 1.0 10.0 100.0 0 10 20 30 40 50 60 Time (months) Time (months) Time (months) H o p a n e s ( u g / g d w ) Rs GGs JIICs GGi y = 36.99e-0.22x R2 = 1.00 y = 14.50e-0.18x R2 = 0.94 y = 12.61e-0.18x R2 = 0.97 y = 0.37e-0.10x R2 = 1.00 0.001 10.000 100.000 0 10 20 30 40 50 60 M e t h y l a t e d A R O ( u g / g d w ) 1.000 0.100 0.010 10.000 100.000 1.000 0.100 0.010 y = 27.81e-0.15x R2 = 1.00 y = 13.09e-0.15x R2 = 0.94 y = 7.10e-0.12x R2 = 0.98 y = 0.29e-0.05x R2 = 0.98 0 10 20 30 40 50 60U n s u b s t i t u t e d A R O ( u g / g d w ) 1 Rs GGs JIICs GGi Fig. 5. Rate loss modeling of aliphatic and aromatic components: isoprenoids (farnesane, norpristane, pristane and phytane), !n-C22 alkanes, On-C23 alkanes, hopanes, methylated and unsubstituted aromatic hydrocarbons. month�1). The pentacyclic aromatic perylene shows the lowest rate constants (0.06–0.08 month�1) probably reflecting its major natural, continuous diagenetic input. The comparison with hopane disappearance rates, supports the lower importance of biodegradation for aromatics. For methylated aromatics and chrysene the relevance of biotic removal is higher (22–64%), compared to pyrene (0–33%), suggesting a prevailing physical attenuation. At JIIC the percentages of bio- degradation are very homogenous for chrysene and y = -0.001x + 0.30 R2 = 0.84 y = -0.0006x + 0.43 R2 = 0.47 y = -0.001x + 0.47 R2 = 0.81 y = -0.0009x + 0.51 R2 = 0.69 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 150 200 250 300 350 400 450 Molecular Weight k ( m o n t h - 1 ) GGs Rs JIICs GGi Np Pr C15 16 17 19 n-alkanes C29-30 Hopanes C21 23 25 26 27 28 29 n-alkanes Fig. 6. Relationship between disappearance rate constants and molecular weight of individual aliphatic hydrocarbons in soils (GGs, Rs, JIIC) and sediments (GGi). 275J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 methylated phenanthrene, pyrene and chrysenes (52– 58%). As observed for aliphatic hydrocarbons, the highest biodegradation corresponds to chrysene and its methylated species in GGi sediments (60–65%), which present the lower overall hydrocarbon losses. The observed trend in disappearance rate constants is broadly consistent with reported patterns for field or laboratory experiments (e.g. Hostettler and Kvenvol- den, 1994; Colombo et al., 1996). Major differences as high isoprenoid rates and low aromatic disappearance, appear to be related to the already advanced alteration of oil residues in the first 6-month sampling, i.e. the rapid attenuation of lighter n-alkanes (!n-C17) and aromatics (i.e. naphthalene, phenanthrene and methyl- ated species) was basically missed and results reflect advanced alteration stages of more persistent oil components. 3.4. Changes of hydrocarbon composition According to the different loss rates of individual components and the stage of microbiological alteration, hydrocarbon composition of sediments and soils showed compositional changes over the 42 months covered by the study (Fig. 2). The aliphatic fraction showed a consistent shift from higher isoprenoid abundance at 6 and 13 months of the spill to a higher molecular weight n-alkane and hopane predominance at 42 months. Aromatic composition was more conservative but showed a significant reduction of methylated species over time. The average relative abundance of isoprenoids in sediments along the three samplings decrease from 24G 15, 25G 22 to 7.6G 6.2%, whereas On-C23 alkanes increase from 34G 16, 42G 22 to 58G 17%, at 6, 13 and 42 months, respectively. The higher hydrocarbon disappearance rates and direct inputs of terrestrial plant material in soils, produce more con- trasted changes for isoprenoids (34G 12, 57G 20 to 3.2G 3.2%) and higher molecular weight n-alkanes (33G 7.3, 21G 17% to 84G 12%). The intense weath- ering in soils is also reflected by a decrease in the proportions of !n-C22 alkanes (26G 5.4, 19G 5.8 to 10G 6.4%). An intermediate enrichment in isoprenoids at 13 months is also indicated by these general means. Fig. 7 presents the percent composition of aliphatic and aromatic hydrocarbons and total concentrations in selected soil and sediment samples. The initial reduction of total aliphatic concentrations between 6 and 13 months, is paralleled by a decrease of !C22 n-alkanes and relative enrichment in isoprenoids, reflecting their selective preservation during biodegradation. This in- termediate isoprenoid enrichment is consistent with the interpretation of heavy biodegradation indicated by the UCM increase at 13 months. During the 42 month attenuation to basal levels, isoprenoids disappear almost completely and higher molecular weight n-alkanes from plant waxes predominate together with residual petro- genic hopanes. The petrogenic/biogenic ratio !n-C22C isopre- noids/n-C15C n-C17COn-C23 (Table 1), reflects the shift of the aliphatic composition along the three campaigns. In the first sampling, the general average ratio in sediments is high (1.3G 0.57) close to a fresh crude oil value (1.6) reflecting the impact of the spill, but non-affected sites such as LB show baseline values (0.3). At 13 months, the general average of the ratio continues high but more variable (1.3G 1.7) because in polluted sites the ratios increased (2.5–6.3) due to the relative enrichment of isoprenoids. After 42 months, the biotic and abiotic removal of oil residues and selective preservation in addition to continued inputs of vascular plant waxes reduce the ratios to background values in nearly all stations (0.44G 0.37). In soils, the shift of petrogenic/biogenic ratios is more drastic. They change from intermediate values similar to a crude in the first sampling (1.5G 0.4), to high variable ratios in the second campaign due to the isoprenoid enrichment (9.5G 12), and very low background values in the third sampling (0.13G 0.10). 0% 20% 40% 60% 80% 100% R s6 R s1 3 R s4 2 JI IC 6 JI IC 13 JI IC 42 G G s6 G G s1 3 G G s4 2 G G i6 G G i1 3 G G i4 2 C o m p o s i t i o n C o m p o s i t i o n 0.1 1 10 100 1000 10000 T o t a l A L I ( u g / g ) T o t a l A R O ( u g / g ) C15,17 C23 Hop 0% 20% 40% 60% 80% 100% R s6 R s1 3 R s4 2 JI IC 6 JI IC 13 JI IC 42 G G s6 G G s1 3 G G s4 2 G G i6 G G i1 3 G G i4 2 0.01 0.10 1.00 10.00 100.00 MethylARO UnsubARO Per Fig. 7. Changes in aliphatic (a) and aromatic (b) hydrocarbon composition after 6, 13 and 42 months of the spill in soils (GGs, Rs, JIIC) and sediments (GGi). Total aliphatic and aromatic concen- trations are also shown. 276 J.C. Colombo et al. / Environmental Pollution 134 (2005) 267–276 In contrast to the aliphatic results, the chemical composition of the aromatic fraction is more conser- vative, specially during the first and second campaign. Effectively, following the almost constant levels at 6 and 13 months (Fig. 7), the aromatic composition in soils and GGi sediments show essentially no changes showing a 40–50% abundance of methylated species, basically phenanthrene, chrysenes and pyrenes. Only after 42 months of the spill, with the drastic reduction of total aromatic levels, there is a significant decrease in the relative abundance of methylated species in favor of unsubstituted PAHs, including perylene. The methylated/unsubstituted ratios indicative of petrogenic versus pyrogenic inputs are thus higher and variable during the first and second campaigns and decrease to lower background values in the third sampling. Average ratios in sediments after 6 and 13 months of the spill are high (0.6G 0.3 and 0.5G 0.3), comparable to fresh crude oil values (0.75). The ratios decrease sharply at 42 months (0.1G 0.1) due to the selective elimination of methylated species relative to pyrogenic PAHs and natural perylene. The relationship in soils shows a similar decrease along the three samplings (0.5G 0.2, 0.6G 0.3 to 0.2G 0.2). Taking together the information on disappearance rates and compositional changes, the aliphatic data support an essentially microbiologically-mediated re- covery of coastal sediments to pre-spill conditions in a 3–4 year period. Isoprenoids and shorter chain n-alkanes decrease over time whereas higher molecular weight n-alkanes are as persistent as hopanes. In contrast, the more subtle compositional changes de- duced for aromatic components, support the interpre- tation of a more efficient physical removal process. Methylated aromatics decrease over time whereas pyrogenic components disappear at rates similar to hopanes. For both, the aliphatic and aromatic fractions, soil disappearance rates are consistently higher, mainly due to enhanced physical removal. The presence of continued inputs, either from biogenic or combustion related sources, appear to contribute to decreased rates. References Atlas, R.M., 1981. Microbial degradation of petroleum hydrocarbons: an environmental perspective.MicrobiologicalReviews 45, 180–209. Blumer, M., Sass, J., 1972. Oil pollution: persistence and degradation of spilled fuel oil. Science 176, 1120–1122. Boehm, P.D., Page, D.S., Burns, W.A., Bence, A.E., Mankiewicz, P.J., Brown, J.S., 2001. Resolving the origin of the petrogenic hydrocarbon background in Prince William Sound, Alaska. Environmental Science and Technology 35, 471–479. Colombo, J.C., Cabello, M., Arambarri, A.M., 1996. Biodegradation of aliphatic and aromatic hydrocarbons by natural soil microflora and pure cultures of imperfect and lignolitic fungi. Environmental Pollution 94, 355–362. Colombo, J.C., Barreda, A., Bilos, C., Cappelletti, N., Demichelis, S., Lombardi, P., Migoya, M.C., Skorupka, C., Suárez, G., 2004. Oil spill in the Rı́o de la Plata estuary, Argentina: 1 – biogeochemical assessment of waters, sediments, soils and biota. Environmental Pollution. Fusey, P., Oudot, J., 1984. Relative influence of physical removal and biodegradation in the depuration of petroleum-contaminated seashore sediments. Marine Pollution Bulletin 4, 136–141. Grossi, V., Massias, D., Stora, G., Bertrand, J.-C., 2002. Burial, exportation and degradation of acyclic petroleum hydrocarbons following a simulated oil spill in bioturbated Mediterranean coastal sediments. Chemosphere 48, 947–954. Hostettler, F.D., Kvenvolden, K.A., 1994. Geochemical changes in crude oil spilled from the Exxon Valdez supertanker into Prince William Sound, Alaska. Organic Geochemistry 21, 927–936. Killops, S.D., Al-Juboori, M.A.H.A, 1990. Characterization of the unresolved complex mixture (UCM) in the gas chromatograms of biodegraded petroleums. Organic Geochemistry 15, 147–160. Page, C.A., Bonner, J.S., McDonald, T.J., Autenrieth, R.L., 2002. Behavior of a chemically dispersed oil in a wetland environment. Water Research 36, 3821–3833. Pastor, D., Sanchez, J., Porte, C., Albaigés, J., 2001. The Aegean Sea oil spill in the Galicia coast (NW Spain). I. Distribution and fate of the crude oil and combustion products in subtidal sediments. Marine Pollution Bulletin 42, 895–904. Sugiura, K., Ishihara, M., Shimauchi, T., Harayama, S., 1997. Physicochemical properties and biodegradability of crude oil. Environmental Science and Technology 31, 45–51. Volkman, J.K., Alexander, R., Kagi, R.I., Rowland, S.J., Sheppard, P.N., 1984. Biodegradation of aromatic hydrocarbons in crude oils from Barrow sub-basin of western Australia. Organic Geochem- istry 6, 619–632. Wang, Z., Fingas, M., Page, D.S., 1999. Oil spill identification. Journal of Chromatography A 843, 369–411. Wang, Z., Fingas, M.F., 2003. Development of oil hydrocarbon fingerprinting and identification techniques. Marine Pollution Bulletin 47, 423–452. Oil spill in the Racuteo de la Plata estuary, Argentina: 2-hydrocarbon disappearance rates in sediments and soils Introduction Methods Results and discussion Total hydrocarbon averages in surface �sediments and soils Modeling of total average hydrocarbon loss rates Modeling of individual hydrocarbon loss rates Changes of hydrocarbon composition References