Bioremediation of a Crude Oil-Polluted Soil: Biodegradation, Leaching ...

BIOREMEDIATION OF A CRUDE OIL-POLLUTED SOIL: BIODEGRADATION, LEACHING AND TOXICITY ASSESSMENTS C. H. CHAîNEAU1∗ , C. YEPREMIAN1 , J. F. VIDALIE2, J. DUCREUX3 and D. BALLERINI3 1 Muséum National d’Histoire Naturelle, Laboratoire de Cryptogamie, Paris, France; 2 TotalFinaElf, DGEP/SE/ENP, Paris la Défense, France; 3 Institut Français du Pétrole,

Rueil-Malmaison, France (∗ author for correspondence, e-mail: [email protected], Fax: 01 40 79 35 94)

(Received 21 August 2001; accepted 27 September 2002)

Abstract. The combined fate and effects of hydrocarbons (HC) on a soil ecosystem affected by bioremediation were studied during 480 days in a field experiment. The HC removal rates, the HC and metabolites mobility and the potential toxicity were assessed. A clayey soil polluted by 18 000 mg HC kg−1 dry soil, was treated with either static-ventilated biopile or series of five windrows periodically tilled in order to determine the relative influence of nutrients, bulking agents, aeration and soil temperature. HC concentrations were determined by infrared spectrometry, gravimetry, gas chromatography and thermodesorption. Between 70 to 81% of the initial HC were removed through biological processes in fertilized soils, whereas natural attenuation without added nutrients was 56%. When adding fertilizers, residual HC were cyclic compounds poorly biodegraded and strongly trapped on the organo-mineral matter. Leaching of HC and water-soluble metabolites was demonstrated during the first stages of biodegradation. Low levels of the HC were detected in the leachates at day 480. Maximal toxicity was highest immediately after the introduction of oil then decreased as biodegradation proceeded. No toxic effect was recorded on worms survival and on seeds germination at day 480. However growth of plants was reduced in treated soils and a potential residual toxicity was observed on the basis of photosynthesis inhibition and bacterial bioluminescence (Microtox) tests. Keywords: bioremediation, crude oil, hydrocarbons, leaching, oil pollution, sorption, toxicity

1. Introduction Hydrocarbons (HC) introduction into the soil environment can occur from pipeline blow-outs, road accidents, leaking of underground storage tanks, landfarming fields and uncontrolled landfilling. When released on the soil surface, HC adsorb on the organo-mineral matter (OMM) of the soil (Means et al., 1980; Li and Gupta, 1994; Fine et al., 1997; Chaîneau et al., 2000a). HC can be subjected to biodegradation, volatilization and leaching (Bossert and Bartha, 1984; Oudot et al., 1989; Chaîneau et al., 1995). Removal of HC from soils can be performed using biological treatments like bioremediation if the environmental conditions are optimum (temperature, soil moisture, nutrients). The presence of high rates of organic matter and clay may affect the extent of biodegradation due to a priming Water, Air, and Soil Pollution 144: 419–440, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

420

C. H. CHAîNEAU ET AL.

effect on microbial communities (Joergensen et al., 1997) and to a decrease of accessibility to microorganisms (Providenti et al., 1993; De Jonge et al., 1997). High concentrations of HC can eliminate vegetation due to their phytotoxic properties (Amakiri and Onofeghara, 1983; Chaîneau et al., 1997, 2000b). Much of the country’s environmental legislations are now focused on treatment and disposal of polluted soils, especially with respect to hazardous waste management. As there are no universal HC cleanup standards, the remediation end points might be in the evaluation of the impact of residual HC on the soil ecosystem and on the water quality. In this work, the bioremediation of a crude oil contaminated clay soil was investigated in a two-year field experiment in Montmirail (Marne, France) where crude oil is pumped from onshore wells. Contaminated soil was treated with active bioremediation and the biodegradation rate, the toxicity and the leaching of hydrocarbons and metabolites were determined. 2. Materials and Methods 2.1. S OIL An elluviated brown clay textured soil (Typic Hapludalf) was collected from an agricultural area in Montmirail (Marne, France) with no previous history of HC contamination (Table I). Crude oil obtained from the Villeperdue field (Marne, France) was composed of 48% saturated HC, 26% aromatics and 26% polar compounds containing resins (7%) and an hexane insoluble fraction (HIF, 19%) that includes petroleum asphaltenes and unidentified organic molecules. The soil (60 m3 ) was contaminated under controlled conditions with crude oil. The oil was applied at a rate of 26 L m−2 on a 80 cm layer of soil to obtain an initial concentration of 18 000 mg HC kg−1 dry soil. Immediately after spreading, the oil and the soil were thoroughly mixed with a rototiller. 2.2. S OIL TREATMENTS Windrows were formulated by placing the polluted soil in six concrete boxes, each containing 5 m3 of soil. The dimensions of each windrow were: length: 3.5 m, width: 1.5 m, height: 1 m. Four different treatments were tested to evaluate the specific influence of nutrients, bulking agents, soil mechanical treatment and temperature on HC biodegradation (Table II). Treatment 1 was used to evaluate natural attenuation of HC without added fertilizers. Treatment 2 determined the influence of nutrients only on HC biodegradation. Fertilization consisted of supplying agricultural fertilizers (NO3 NH4 : 4.8 kg m−3 ; K × P (20 × 30) : 1.3 kg m−3 ) to obtain a C/N/P ratio of 100/10/1 (Bossert and Bartha, 1984). It was applied on days 0 and 270. Soils in the three treatments (T3 , T4 , T5 ) were amended with nutrients in the same rate as T2 and were mixed with straw (15% v/v) in order to increase

BIOREMEDIATION OF A CRUDE OIL-POLLUTED SOIL

421

TABLE I Composition of the elluviated brown soil (Typic Hapludalf) Clay Silt Sand pH Organic carbon C/N N-NO3 N-NH4 K-K2 O P-P2 O5

(%) (%) (%) (%) (mg kg−1 ) (mg kg−1 ) (mg kg−1 ) (mg kg−1 )

48 46 6 8 11 11.46 2.23 1.12 0.2 0.02

air diffusion, stimulate microorganisms growth and improve the soil structure. T4 was amended with freshly harvested Rye-grass (Lolium perenne L., 15% v/v). T5 was covered with a greenhouse to increase the temperature of the soil surface. Temperature in the first 15 cm soil layers increased 15 ◦ C in summer and 3 ◦ C in winter as compared to T3 under sunny conditions. Rainfall and temperature on the site are given in Figure 1. A control soil (T0 ) without HC received fertilization and straw in the same amounts as the other treatments. Every month, all soil was mechanically tilled to renew and facilitate air diffusion. A second method for HC biodegradation was evaluated. A biopile (Bp) composed of 30 m3 of polluted soil (2.5 m height, 4 m length, 3 m width) was set up on a plastic liner. Polluted soil was mixed with straw, amended with nutrients in the same proportion as the windrows and covered with a black plastic liner to protect the soil from the rain. At three depths (60, 120, 180 cm), four slotted pipes were placed in the soil and were connected to a blower (surpresser, LAMSON minitron 4/21S, 0.75KW) which aerated the soil continuously. The incoming airflow was 10 m3 h−1 . Two lysimeters (30 cm height, 30 cm length, 20 cm width) were set near the windrows. They were filled with 15 kg of the soils from T0 and T3 to sample the drainage waters during rainy periods. 2.3. S OIL AND WATER SAMPLING At 0, 90, 180, 270, 330 and 480 d, 10 individual soil samples were randomly collected in each windrow with a 5 cm diameter steel corer. In Bp, 5 individual samples of soil were taken respectively in the layers 0–60, 60–120 and 120–180 cm. For each treatment, an average sample (2000 g) was prepared by homogenizing and mixing an aliquot of each replicate. Each sample was wrapped with an aluminum

422


TABLE II Characteristics of the different soil treatments Treatments T0 T1 T2 T3 T4 T5 Bp

Oil contaminationa – + + + + + +

Fertilizationb + – + + + + +

Strawc + – – + + + +

Rye-grassc – – – – + – –

Greenhouse – – – – – + –

a Hydrocarbons content was 18 000 mg HC kg−1 dry soil. b Fertilization provided 4.8 kg NO NH m−3 dry soil and 1.3 kg KxP (20, 30) m−3 dry soil. 3 4 c 15% v/v.

Figure 1. Monthly measurements of air temperature () and rainfall ().

foil and kept at –20 ◦ C until analysis for HC composition and concentration and at 0 ◦ C until toxicity and leaching potential experiments. The drainage water from the two lysimeters was regularly collected in glass bottles. Samples were kept at 0 ◦ C until HC and dissolved organic carbon (DOC) analyses. 2.4. C HEMICAL ANALYSIS Soils samples were sieved (2 mm diameter openings) before HC analysis. The total amount of HC in the soil was measured with three methods : infrared spectrometry, microgravimetry and thermodesorption (Pollut-Eval). Detailed analyses


423

of chemical composition of residual crude oil were made by capillary gas chromatography (GC-FID). Infrared measurements (IR) consisted of the extraction of 10 g of dried (60 ◦ C/24 hr) replicates soil samples with 20 mL of carbon tetrachloride for 30 min by ultrasonication (40 kHz) at 40 ◦ C. Extraction efficiency was about 95%. The IR-absorption was measured at 2925 cm−1 using a Perkin-Elmer 683 infrared spectrophotometer equipment with 1 mm cells. The calibration curve was developed with the initial crude oil. The amount of HC was calculated as the difference between the amount of total HC in the polluted soil minus the biogenic HC content in the control. Statistics were determined by analysis of variance (ANOVA, F-test). The microgravimetry (GR), described in Chaîneau et al. (1995), was applied to the average sample of soil. Soxhlet extraction of 50 g dried soil sample with chloroform was conducted for 8 hr. The chloroform extractable organic matter (EOM) was weighed after evaporation of the solvent. It contained HC and some biogenic lipids. The EOM extracts was separated in hexane-insoluble, saturated, aromatic and resin fractions by liquid-solid chromatography (Oudot et al., 1987). Each fraction was weighed. The total HC extracts were analysed by computerized gas-chromatography (GC) with a Delsi DI 200 chromatograph fitted with a direct injection port and a FID detector both set at 340 ◦ C; carrier gas was He under 0.8 bar; column was a CP Sil 5 CB (Chrompack) capillary column (50 × 0.32 mm, film thickness 0.25 µm); temperature programming was 100–320 ◦ C, 3 ◦ C min−1 . Acquisition and numerical treatments of data were performed using custom-made computer programs (Oudot, 2000). Thermodesorption measurements (Pollut-Eval) allowed the determination of HC concentration and composition in the average soil sample of the windrows (Ducreux et al., 1997; Haeseler et al., 1999). A pyrolysis analyzer allowed the thermo-vaporization and the pyrolysis of organic carbon. The FID detection provided three characteristic peaks corresponding to the fractions constituting the pollutant. In the case of the crude oil, 3 peaks were detected corresponding respectively to S1 : distillable hydrocarbons with a maximum carbon number of 20, S2 : other distillable hydrocarbons (C20 –C40 ) and S3 : non-distillable hydrocarbons with a carbon number over 40. In the control soil, the amount of S1 , S2 , S3 was negligible. In drainage waters, the concentrations of dissolved organic carbon (DOC) and extractable organic matter (EOM) including HC and some metabolites were measured to determine whether hydrocarbons and/or metabolic by-products leached out from the soil. They were measured according to Chaîneau et al. (1996) and represented the difference between the treated and the control waters. The EOM was recovered by CCl4 liquid-liquid extraction, evaporated and concentrated to 1 mL before IR and GC analyses. 2.5. M ICROBIOLOGICAL COUNTS The number of colony-forming units (CFU) of total heterotrophic bacteria (THB) and hydrocarbon-adapted bacteria (HAB) were determined in the average soil samples after 0, 30, 90, 180 and 480 d according to the most probable number method

424


(MPN) using three tubes per dilution. The method previously described (Chaîneau et al., 1995) and used by different authors (Wrenn and Venosa, 1996) consisted in the inoculation of appropriate media (trypcase-soy 30 g L−1 for THB, mineral medium for HAB) with decimal dilutions of soil or culture medium. Incubation was run for 3 d for THB and 21 d for HAB at 24±1 ◦ C. The number of viable microorganisms was obtained from standard MacCrady tables after examination of growth positive tubes. 2.6. L EACHING POTENTIAL TESTS At 0, 30, 90, 180, 330 and 480 d, leaching potential tests were performed on the average sample of soil from T0 and T3 according to Chaîneau et al. (2000a). Samples were sieved (2 mm diameter openings). Tests were realized using a 100 g/1000 mL soil to distilled water ratio. The suspensions of soils were placed in sealed glass vials and agitated on a rotary shaker for 24 hr at 20 ◦ C. The concentrations of DOC and EOM in leachates were measured as for the drainage water samples. 2.7. E COTOXICITY TESTS Toxicity tests (phytotoxicity, worm mortality, photosynthesis inhibition, bacterial luminescence perturbation) were carried out on average samples of bulk soils from T0 and T3 at 0, 30, 90, 180, 330 and 480 d. The phytotoxicity (inhibition of germination and inhibition of plant growth) was assessed with two standardized methods (NFX 31-201, NFX 31-202) as described in Chaîneau et al. (1997). The earthworm mortality was determined according to the standard normalized procedure (NFX 31–250). It consisted in the determination of surviving compost worms (Eisenia fetidia) after 15 days of incubation in soils. Perturbation of the photosynthetic process of higher plants was determined by the Hill reaction (Dick, 1974). It was tested on a soil-water extract (1/10 w/v) from T0 and T3 soils at days 0 and 480. A suspension of chloroplasts was added to a reaction mixture containing 2,6-dichloro-phenol-indo-phenol (DPIP) and the water extract. The reaction was followed by illuminating the tubes and recording the decrease in absorption of red light. The standardized Microtox test (NF EN ISO 11348-3) was performed after 0, 90, 180 and 480 d. It allowed the qualification of the toxicity of watersoluble molecules towards the bacteria Vibrio fisheri. The photoluminescence of serial dilutions (1/20 to 1/1300) in distilled water of the water extract of the soil from T0 , T1 , T2 , T3 was determined.


425

3. Results 3.1. HC REMOVAL IN SOIL 3.1.1. Total Crude Oil Average initial content of HC in the soils was 14424±1746 mg HC kg−1 dry soil (IR) and 15853±4095 mg HC kg−1 dry soil (GR) indicating that the controlled spreading of oil was evenly performed (Figure 2). A higher concentration of EOM was recorded by gravimetry in T4 as a result of the extraction of biogenic organic compounds from the vegetal organic matter (Chaîneau et al., 1997). About 20% of volatilization of the light fraction of crude oil below nC14 occurred in the field and during analytical procedures (Chaîneau et al., 1995). HC concentrations decreased logarithmically in all soils during the first 180 d except for the T1 where it was linear indicating a constant rate of removal (38 mg HC kg−1 dry soil d−1 , Figure 2). The influence of fertilization, bulking agent and temperature on HC biodegradation was highly significant (P < 0.01) at day 90. The increase of biodegradation was 39% with nutrients, 45% with nutrients and straw, 56% with nutrients, straw and organic matter and 69% with nutrients, straw and a greenhouse. From day 180 to day 480, no statistically significant difference was observed between all fertilized treatments. The residual HC content in the unfertilized windrow was about 6420 mg HC kg−1 dry soil (IR, GR) whereas the mean HC content in fertilized soils was 4381±524 mg HC kg−1 dry soil (GR), 2645±899 mg HC kg−1 dry soil (IR). The final removal rates for total crude oil were: T1 : 56% (IR), 60% (GR), T2 : 80% (IR), 78% (GR), T3 : 73% (IR), 75% (GR), T4 : 84% (IR), 78% (GR) and T5 : 87% (IR), 75% (GR). The rates of HC removal were slower in Bp than in the windrows (Figure 2). At the end of the experiment, 40% (IR), 61% (GR) of HC were removed. Standard deviations of HC concentrations were as high as 50% of the mean concentration indicating a strong heterogeneity of HC repartition in the biopile and less than 25% for the windrows, indicating a good homogeneity of HC caused by periodical tilling. The curves of HC concentrations determined by GR and by IR were almost parallel indicating a good correlation between the two methods but with differences up to 40% (Oudot and Dutrieux, 1989). GC analyses showed that linear, branched and cyclo-alkanes were only partially biodegraded in the unfertilized soil whereas in fertilized soils, GC-resolved molecules were removed in 480 d (Figure 3). In the biopile, the soil compaction caused a decrease in HC biodegradation as shown by GC analyses (Figure 3). Linear and branched alkanes contained in the 0–60 cm soil layer were more degraded than in deeper soil layers at day 90. The final biodegradation stage was identical for all fertilized soils. The results of the Pollut-Eval analyses showed that the percentage repartition of the three fractions (S1 , S2 , S3 ) was 35/33/32 in the polluted soil before any treatment (Table III). A prevailing ‘heavy’ fraction (S3 ) became predominant (>60%) at day 180. The percentage of the ‘light’ (S1 ) and of the ‘medium’ fraction (S2 ) decreased

426


TABLE III Hydrocarbons analyses determined by Pollut-Eval pyrolysis in soils: S1 : distillable hydrocarbons with a maximum carbon number of 20, S2 : other distillable hydrocarbons (C20 -C40 ) and S3 : non-distillable hydrocarbons with a carbon number over 40. In the control soil, the amount of S1 , S2 , S3 was negligible Time (d) 0

90

180

270

480

HC composition (%) S1

S2

S3

S1

S2

S3

S1

S2

S3

S1

S2

S3

S1

S2

S3

35 35 35 35 35

33 35 35 35 35

35 33 33 33 33

20 20 14 5 23

42 26 16 15 21

38 54 70 80 73

15 15 12 9 9

47 31 19 25 20

38 54 69 66 71

8 8 7 10 7

44 30 15 37 22

48 62 78 52 71

4 4 6 6 10

42 15 15 11 28

54 81 79 83 62

Treatments T1 T2 T3 T4 T5

respectively to 6 and 15%. Comparatively, in the unfertilized soil, the decrease of S1 and S2 was more limited. 3.1.2. Crude oil fractions The initial concentration of saturated HC was about 6600 mg kg−1 dry soil (Figure 4). In all treated soils, it decreased logarithmically with time. In absence of fertilization, the final biodegradation of saturated HC reached 64%. The influence of soil aeration and bulking agents (straw and vegetal organic matter) was only demonstrated in the first 180 d. The final biodegradation rate was about 79% in fertilized soils. At day 0, the aromatic hydrocarbons soil content was 3500 mg HC kg −1 dry soil. About 40% of the initial PAH remained in soil at day 480 (Figure 4). No significant difference between the treatments was observed. A decrease of the resins was recorded in the specific treatments T1 , T2 , and T3 . The total removal ranged between 15 and 30%. No significant decrease was observed in the other treatments. By integrating over all the different treatments, the final removal rate of resins was about 19%. The hexane-insoluble fraction was composed of asphaltenes and unidentified organics. A general decrease in the concentration was observed in all treated soils much more pronounced with a fertilization. 3.2. M ICROBIAL COUNTS Table IV shows the changes in microbial populations in the untreated and treated soils. In the control, soil treatment (tilling, fertilization, amendment) slightly mod-


427

Figure 2. Hydrocarbon concentrations in treated soils determined by gravimetry () and IR spectrometry () over time.

428


Figure 3. Chromatograms of HC in soils show the biodegradation of crude oil HC in the soil. (A) Initial oil; (B) HC in the unfertilized soil at day 480; (C) HC in fertilized soils at day 480; (D) HC profile in Bp at day 90. The numbers represent the carbon number of n-alkanes; intermediate peaks are branched alkanes and/or GC-resolved PAHs; UCM, unresolved complex mixture.

ified the number of bacteria. The ratio HAB/THB varied from 1 to 20%. The addition of HC increased the number of viable microorganisms. Microbial counts indicated that in T1 , THB and HAB were respectively 10, 1000 times, higher than in T0 . The addition of fertilizers and bulking agents strongly stimulated the growth of soil microorganisms. A 1800 fold increase in the THB population and a 3000 fold increase in the HAB population were recorded in T3 at day 180. The ratio HAB/THB increased up to 100% then, both HUB and THB returned to levels similar to the unoiled soil. 3.3. D RAINAGE WATER During the first 270 d of experiment, the concentration of EOM progressively increased up to 4.5 mg L−1 (0.8 mg kg−1 dry soil) then decreased regularly (Fig-


429

Figure 4. Concentration of (A) saturated hydrocarbons, (B) aromatic hydrocarbons, (C) the hexane insoluble fraction and (D) resins in oil-treated soils over time in T1 (), T2 (), T3 (), T4 (), T5 ().

ure 5). EOM, in T3 , mostly consisted in a unresolved complex mixture (UCM, Figure 6) composed of 17% saturates, 19% aromatics and 64% polars. By contrast, in the control, linear alkanes in the nC25–nC35 range and unidentified biogenic molecules without UCM were detected. GC of EOM extracts were similar at day 480. The concentration of DOC remained quite constant in the first 90 d (15 mg L−1 , 0.75 mg kg−1 dry soil) and increased sharply up to 560 mg L−1 (28 mg kg−1 dry soil) at day 180 (Figure 5). Then, it decreased down to levels similar to the control. The difference between the treated and the control soil was about 4.1 µg kg−1 dry soil. 3.4. L EACHING POTENTIAL The initial concentration of EOM in the leachates was about 43 mg L−1 (430 mg kg−1 dry soil). It decreased steadily and stabilized to 1 mg L−1 (10 mg kg−1 ) (Figure 5). GC analyses were quite similar to those found in drainage waters at day 480. Biogenic alkanes in the nC23–nC37 range and few lipids were detected indicating that the residual petrogenic HC remained trapped on the OMM. The concentration of DOC remained constant during the first 90 d (15 mg kg−1 ), increased sharply

430


TABLE IV Microbial counts (colony forming units g−1 soil) in the control and in the different treated soils. THB: Total Heterotrophic Bacteria; HAB: Hydrocarbon Adapted Bacteria Time

Treatments

(d)

T0

T1

T2

T3

T4

T5

Bp (soil layer in cm 0–60

60–120

120–180

THB (× 106 ) 0 30 90 180 480

45 45 25 25 2.5

45 950 250 9.50 –

45 950 95 200 –

45 950 450 4500 95

45 950 950 150 –

45 950 200 250 –

45 – 250 1500 9.50

45 – 45 95 0.45

45 – 45 95 45

HAB (× 104 ) 0 30 90 180 480

950 950 25 2.5 2.5

950 9500 1500 950 –

950 9500 2500 4500 –

950 9500 4000 7500 95

950 9500 9500 1500 –

950 950 9500 4500 –

950 – 9500 4500 2.50

950 – 150 9500 25

950 – 450 9500 450

up to 560 mg kg−1 at day 180 and decreased down to levels similar to the control (Figure 5). 3.5. T OXICITY The initial concentration of crude oil in soil was toxic to seeds and to worms (Table IV). The resistance of seeds to oil contamination followed the decreasing order: wheat > barley > maize > pea > lettuce. Ecotoxicity tests demonstrated that the toxicity regularly decreased. The mortality of worms was not affected by residual HC in the soil. At the same time, the inhibition of seed germination was not observed except for the lettuce; however, a significant reduction of plant growth was observed (Table V). Maximal inhibition of the bacterial luminescence for Vibrio fisheri was observed immediately after the initial pollution (Table VI). The toxicity observed in the control was attributed to the high nutrients content in the soil. The toxicity decreased as bioremediation proceeded but was still higher in the treated soils than in the control. Similarly, the Hill reaction was totally inhibited at day 0, and 25% of inhibition was observed at day 480 (Table VI).


431

Figure 5. Concentration of solvent-extractable organic matter (EOM, ) and dissolved organic carbon (DOC, ) over time in drainage waters (A) and in leachates from leaching potential tests (B).

4. Discussion 4.1. R EMOVAL OF OIL IN SOILS In this study, about 76% of the initial HC were removed through biodegradation under optimal conditions. This rate is close to the maximum value that can be reached by microbial degradation of crude oil (Bossert and Bartha, 1984; Morgan

432


Figure 6. Gas-chromatographic analyses of solvent-extractable organic matter of drainage waters: (A) oil-treated soil T3 ; (B) control T0 .

433


TABLE V Soil toxicity (phytotoxicity and worms mortality tests) over time Tests

Organisms

Inhibition of seed germination (%)

Hordeum vulgare Triticum aestivum Pisum sativum Zea mays Lactuca sativa

Inhibition of plant growth (%)

Hordeum vulgare Triticum aestivum Pisum sativum Zea mays

Worms mortality (%)

Eisenia fetidia

Time (d) 0

30

90

180

270

330

480

100 100 100 100 100

70 66 100 88 100

60 62 100 80 100

20 20 12 30 67

10 4 0 25 30

8 3 0 0 21

0 0 0 3 10 8 21 4 0

100

100

90

50

10

0

0

TABLE VI Soil toxicity (Microtox and Hill reaction) over time Tests

Treatments

Time (d) 0

Microtox: inhibition of luminescence (%)

T0 T1 T2 T3

42 31 75 48

Hill reaction: inhibition of the photosynthesis (%)

T0 T3

5 100

90

480

10 9 56 27

8 12 27 21 5 25

and Watkinson, 1989; Atlas and Bartha, 1992; Suguira et al., 1997; Salanitro et al., 1997). By integrating over all the soil treatments, the mean biodegradation rate (expressed in HC amounts) throughout the experiment was about 1100 mg kg−1 mo−1 and is consistent with literature data (Oudot et al., 1989; Morgan and Watkinson, 1989; Atlas and Bartha, 1992. Pollard et al., 1999). In all treated soils, linear and branched alkanes were assimilated first. Isoprenoids (farnesane, norpristane, pristane, phytane) were slowly assimilated (Atlas and Bartha, 1992; Chaîneau et al., 1995; Suguira et al., 1999) but a persistence of part of the UCM was observed. These results are in agreement with previous experiments (Chaîneau et al.,

434


1996, 1999; Angehrn, 1998; Pollard et al., 1999; Oudot, 2000). UCM is mainly composed of polycyclic alkylated saturated HC, polycyclic alkylated aromatic hydrocarbons (Oudot et al., 1987; Killops and Al-Juboori, 1990; Angehrn, 1998), T-shaped molecules (Gough and Rowland, 1990) and are viewed as resistant to microbial metabolism. The biodegradation of alkanes was in the order n > iso > cyclo alkanes as already observed in field and laboratory experiments (Oudot et al., 1987, 1989; Chaîneau et al., 1995, 1999; Pollard et al., 1999). In the absence of added mineral nutrients, 56% reduction occurred. GC analyses showed that alkanes were only partially degraded whereas biodegradation was quite complete in fertilized soil. The initial content of N in soil was 960 mg kg−1 , which could support a theoretical biodegradation of about 12 000 mg HC kg−1 of soil (C/N = 10; Bossert and Bartha, 1984). This should have led to a theoretical residual concentration of about 3500 mg HC kg−1 . In fact, 6500 mg HC kg−1 persisted in the soil at day 480. Nutrient limitation occurred in the non-fertilized soil indicating an uncompleted microbial assimilation of biogenic nutrients (Chaîneau et al., 2000b). Fertilization strongly stimulated biodegradation (Bossert et Bartha, 1984; Morgan and Watkinson, 1989; Atlas and Bartha, 1992) and the final biodegradation rate was over 75%. The positive influence of bulking agents, greenhouse and tilling on HC biodegradation was restricted to the early stages of the process, the final biodegradation stage was identical in all treated soils. Previous studies showed that the addition of bulking agents such as compost, wheat or sawdust in HC-polluted soils may increase in HC biodegradation rates (Joergensen et al., 1997; Rhykerd et al., 1999), by improving the stabilization of the soil structure and the soil aeration. The addition of vegetal organic matter to soil could enhance oil degradation as a result of a priming effect on microbial populations (Joergensen et al., 1997). This effect was not significant during the experiment. The positive effect of tillage on hydrocarbon disappearance is attributed to an increase in aeration and bioavailability of oil by redistributing the oil in the soil which increased the area of exposure of oil to microorganisms (Rhykerd et al., 1999). In the biopile, the compaction of the soil reduced air diffusion causing a decrease in bioremediation rates. During the experiment, the mean air temperature was 10 ◦ C. It was typical of temperate zones but was lower than the optimal temperature for biodegradation of HC (Morgan and Watkinson, 1989; Atlas and Bartha, 1992). The low temperatures recorded in the first three months did not affect biodegradation rate and microorganisms growth. It has been previously observed that HC can be biodegraded in cold environments, since the site specific conditions lead to the selection of a specific soil population with a lower optimal temperature (Morgan and Watkinson, 1989; Atlas and Bartha, 1992; Margesin and Schinner, 1997). The results of biodegradation activity of the main fractions of crude oil were comparable to those found in previous studies (Oudot et al., 1987; Morgan and Watkinson, 1989; Suguira et al., 1997; Chaîneau et al., 1995, 1996, 2000b). The saturated HC were 79% degraded meanwhile the rates of aromatic removal was about 60% in optimal conditions. A decrease in the left part of the UCM revealed


435

the preferential assimilation of low molecular hydrocarbon corresponding to mono and dicyclic naphthenes and to mono and diaromatic PAH. Adding nutrients caused a 15% increase of the saturates biodegradation whereas aromatic HC were degraded at the same extent in all soils, fertilized or not. The non-effect of fertilization on PAH biodegradation has scarcely been observed (Fedorak and Westlake, 1981; Carmichael and Pfaender, 1997). Hutchins et al. (1991) found an overall inhibitory effect on hydrocarbon degradation by nitrate. In most cases, a positive influence of nutrients on the biodegradation of aromatic hydrocarbons rates is observed (Bossert and Bartha, 1984; Morgan and Watkinson, 1989; Atlas and Bartha, 1992; Chaîneau et al., 2000b). The lack of effect of nutrients on PAH assimilation can be attributed to the non-stimulation of the aromatic-degrading microorganisms (Carmichael and Pfaender, 1997). In most field studies, resins and asphaltenes are degraded to a low extent only (Chaîneau et al., 1995; Pollard et al., 1994). A temporary or permanent increase in the polar fraction is usually observed in batch liquid cultures as a result of the accumulation of metabolic intermediates (Oudot et al., 1987; Thouand et al., 1999). In the field, the removal of resins was limited to 20% in agreement with other experiments (Oudot et al., 1989; Chaîneau et al., 1999). The hexane insoluble fraction containing asphaltenes and polar molecules was shown to be partially or completely resistant to microbial assimilation (Pollard et al., 1999; Thouand et al., 1999). It was reported that in optimal cultures, a maximum of 20% of biodegradation can occur (Oudot et al., 1987). By contrast, in this field experiment, the hexane insoluble fraction was assimilated to a 75% extent in fertilized soils vs 60% in non-fertilized soils. The exact composition of the HIF in oils is unknown but it contains aggregates of extended polyaromatics, naphthenic acids, sulfides, polyhydric phenols, fatty acids and metalloporhyrins (Sugiura et al., 1997). Our results suggest that the oil used in this experiment contained easily degradable polar molecules that are pooled with asphaltenes during hexane precipitation. The composition of the residual oil was saturated HC: 34,5%, aromatic HC: 33%, polar compounds: 32,5% as observed in other field experiments (Oudot et al., 1989; Chaîneau et al., 1996). 4.2. L EACHING OF HYDROCARBONS Analyses of drainage water showed that in the clayey polluted soil a low amount of HC eluted in the percolating water. The percolation was progressive and maximal at day 180. By integrating over the two-year field experiment, the total input was in the order of 0.02% of the initial HC load. Similarly, the leaching potential test, that extract much more compounds than natural percolation showed that the release of HC in water was maximal during the first 90 d. It remained constant at a concentration of 20 mg kg−1 of soil. When released onto the soil, HC adsorb rapidly on the soil OMM which determines their migration in the soil profile (Kowalska et al., 1994; Li and Gupta, 1994; Fine et al., 1997; Chaîneau et al., 2000a). The sorption of hydrophobic compounds on the OMM is correlated with the soil organic

436


content (Means et al., 1980; Chaîneau et al., 2000a) and is influenced by the soil moisture (Chiou and Shoup, 1985). Sorption binds contaminants and removes them from the dissolved state (Providenti et al., 1993; Kowalska et al., 1994). HC can elute with percolating water as soon as water leached in the soil column (Chaîneau et al., 1996, 2000a). Anghern et al. (1998) showed that the mobility of residual hydrocarbons was very low as a result of a strong adsorption on the soil OMM. By contrast, polar compounds resulting from HC biodegradation are less adsorbed on the mineral matter than HC as a result of their polarity. This resulted in a temporary increase of DOC when active biodegradation proceeded (Chaîneau et al., 1996). The nature of these metabolites, e.g., alkools, alkanals, organic acids and aromatic ketones is partially known (Langbehn and Steinhart, 1995; Cozzarelli et al., 1995; Anghern et al., 1998). 4.3. T OXICITY OF OIL AND ASSOCIATE METABOLIC BY- PRODUCTS The vegetation development can totally be eliminated or seriously reduced in the case of inputs of oil in the soil (Amakiri and Onofeghara, 1983; Wetzel and Werner, 1995; Chaîneau et al., 1997). With an initial HC content in soil of 1.5% (w/w), the germination was totally inhibited. It has been previously demonstrated that the phytotoxicity was highly correlated with the presence of aromatic hydrocarbons and of light HC i.e. the naphthas in oil (Chaîneau et al., 1997). Our results suggested that as biodegradation proceeded, the potential phytotoxicity reduced (Wang and Bartha, 1990; Xu and Johnson, 1995; Chaîneau et al., 1996; Salanitro et al., 1997). Substantial toxicity remained as HC persisted in soil; the response of plants varying with the plant species and with the concentration of oil (Hund and Traunspurger, 1994; Chaîneau et al., 1996, 1997). This residual toxicity may be attributed to the inherent toxicity of non-biodegraded HC and to the perturbation of the soil-water exchange (Chaîneau et al., 1997; Li et al., 1997). Infiltration of water in HC-contaminated soils has been shown to be lower than in oil-free soils due to the hydrophobic properties of oil (Sawatsky and Li, 1997). As a result of poor soil water capacity and residual HC toxicity, root growth is strongly impaired which resulted in the decrease of dry matter yields (Xu and Johnson, 1995; Chaîneau et al., 1996; Li et al., 1997). The direct contact of the contaminated soil with organisms was also experimented with worms. As for seeds, the initial toxicity may be attributed to the solvent properties of naphthas. When HC concentration decreased, the mortality reduced to levels similar to the control as already observed in other experiments (Hund and Traunspurger, 1994; Salanitro et al., 1997). The apparent reduction of toxicity may result from the irreversible adsorption of toxic HC on the soil OMM (Wetzel and Werner, 1995). Initial oil acute toxicity in Microtox assays was caused by the presence of light HC that could impair bacterial growth. The initial toxicity in the control soil suggested that fertilization was toxic to bacteria in the early stages. In most cases, the toxicity reduced when concentrations of HC in soil decreased (Wang and Bartha,


437

1990; Belkin et al., 1994; Hund and Traunspurger, 1994). Without fertilization, the toxicity reduced rapidly to a level similar to the control. In fertilized soils, substantial toxicity remained even with HC concentrations lower than in unfertilized soils as a result of the production of metabolites (Belkin et al., 1994; Hund and Traunspurger, 1994). In batch liquid cultures, an increase of the polar fraction is frequently observed (Oudot et al., 1987; Chaîneau et al., 1999; Shelton et al., 1999). The acute toxicity of the water-soluble metabolic byproducts was demonstrated (Mahaney, 1994; Belkin et al., 1994; Ramirez et al., 1996; Shelton et al., 1999). In soil, metabolites can be subjected to adsorption/desorption on the OMM and/or may also be assimilated by microorganisms (Chaîneau et al., 1995). The leaching potential tests pointed out that the release of water-soluble compounds was not permanent and decreased when degradation ceased. The Hill reaction showed that a residual toxicity persisted in the soil at the end of the experiment. In presence of oil, the photosynthetic process was strongly modified due to structural damages to membranes and chloroplasts (Tukaj and Szurkowski, 1993). Our results suggest that at low rates, the residual hydrocarbons and/or metabolic by-products may cause serious perturbations to the cellular metabolisms as a result of structural damages in case of uptake by plants. The results of toxicity studies are strongly dependent on the sensitivity of the test used. Some tests (worms, seed germination) can highlight very high acute toxicity only, whereas Microtox and Hill reaction are much more sensitive and indicative. 5. Conclusions When released onto the soil, HC cause serious damages to the natural environment. The persistent toxic effect may last over a long time. When treated appropriately by optimizing the biodegradation potential of natural-occurring hydrocarbon degraders, the main part of an HC-pollution can be eliminated more rapidly. Adequate fertilization and periodical tillage were efficient as compared to untreated soil. Attention should be paid to the release of polar toxic metabolic by-products during HC microbial assimilation. The endpoint of bioremediation treatments should be determined in the light of the combinaison of chemical analyses of residual pollutant to assess the final biodegradation stage and appropriate toxicity tests to determine the impact of treated soil on the environment. Acknowledgements We thank Fabrice Molin and COPAREX for their support, without which this study would not have been possible. Part of this work was supported by Traitement Valorisation Décontamination under funds by TOTALFINAELF. We also thank Alex Graham who kindly checked the English grammar and spelling.

438


References AFNOR: 1982, Norme NFX 31-201, Essai d’inhibition de germination de semences par une substance. Association Française de Normalisation, Paris. AFNOR: 1984, Norme NFX 31–250, Détermination de la toxicité d’une substance vis à vis des Lombriciens (Eisenia fetida). Association Française de Normalisation, Paris. AFNOR: 1986, Norme NFX 31–202, Essai d’inhibition de la croissance des végétaux par une substance. Association Française de Normalisation, Paris. AFNOR: 1999, Norme NF EN ISO 11348–3, Détermination de l’effet inhibiteur d’échantillon d’eau sur la luminescence de Vibrio fisheri. Association Française de Normalisation, Paris. Amakiri, J. O. and Onofeghara, F. A.: 1983, ‘Effect of crude oil pollution on the growth of Zea mays, Abelmoschus esculentus and Capsicum frutescens’, Oil Petrochem. Pollut. 1, 199–205. Angehrn, D., Gälli, R. and Zeyer, J.: 1998, ‘Physicochemical characterization of residual mineral oil contaminants in bioremediated soil’, Environ. Toxicol. Chem. 17, 2168–2175. Atlas, R. M. and Bartha, R.: 1992, ‘Hydrocarbon biodegradation and oil spill bioremediation’, Advan. Microb. Ecol. 12, 287–338. Belkin, S., Stieber, M., Tiehm, A., Frimmel, F. H., Abeliovich, A., Werner, P. and Ulitzur, S.: 1994, ‘Toxicity and genotoxicity monitoring stimulated hydrocarbon biodegradation processes’, Environ. Toxicol. Water Qual. 9, 223–230. Bossert, I. and Bartha, R.: 1984, ‘The Fate of Petroleum in Soil Ecosystem’, in R. M. Atlas (ed.), Petroleum Microbiology, Macmillan Co., New York, pp. 435–476. Carmichael, L. M. and Pfaender, F. K.: 1997, ‘The effect of inorganic and organic supplements on the microbial degradation of phenantrene and pyrene in soils’, Biodegradation 8, 1–13. Chaîneau, C. H., Morel, J. L. and Oudot, J.: 1995, ‘Microbial degradation in soil microcosms of fuel oil hydrocarbons from drilling cuttings’, Environ. Sci. Technol. 29, 1615–1621. Chaîneau, C. H., Morel, J. L. and Oudot, J.: 1996, ‘Land treatment of oil-based drill cuttings in an agricultural soil’, J. Environ. Qual. 4, 858–867. Chaîneau, C. H., Morel, J. L. and Oudot, J.: 1997, ‘Phytotoxicity and plant uptake of fuel oil hydrocarbons’, J. Environ. Qual. 26, 1478–1483. Chaîneau, C. H., Morel, J. L., Dupont, J., Bury, E. and Oudot, J.: 1999, ‘Comparison of the fuel oil biodegradation potential of hydrocarbon-assimilating microorganisms isolated from a temperate agricultural soil’, Sci. Total Environ. 227, 237–247. Chaîneau, C. H., Morel, J. L. and Oudot, J.: 2000a, ‘Vertical infiltration of fuel oil hydrocarbons in an agricultural soil’, Toxicol. Environ. Chem. 77, 111–124. Chaîneau, C. H., Morel, J. L., Oudot, J.: 2000b, ‘Biodegradation of fuel oil hydrocarbons in the rhizosphere of Maize (Zea mays L.)’, J. Environ. Qual. 29, 569–578. Chiou, C. T. and Shoup, T. D.: 1985, ‘Soil sorption of organic vapors and effects of humidity on sorptive mechanism and capacity’, Environ. Sci. Technol. 19, 1196–1200. Cozzarelli, I. M., Herman, J. S. and Baedecker, M.: 1995, ‘Fate of microbial metabolites of hydrocarbons in a coastal plain aquifer: The role of electron acceptors’, Environ. Sci. Technol. 29, 458–469. De Jonge, H., Freijer, J. I., Verstraten, J. M. and Westerveld, J.: 1997, ‘Relation between bioavailability and fuel oil hydrocarbon composition in contaminated soils’, Environ. Sci. Technol. 31, 771–775. Dicks, J. W.: 1974, ‘An experiment on herbicides: Inhibition of the Hill reaction’, Biochem. Edu. 2, 69–71. Ducreux, J., Lafargue, E., Bocard, C., Marquis, F. and Pillot, D.: 1997, ‘Use of the Rock-Eval Method for Assessment of Oil-contaminated Soils: Pollut-Eval Analyzer’, in J. Gottlieb, H. Hötzl, K. Huck and R. Niessner (eds), Proceedings of Europe, First International Conference on Strategies and Techniques for the Investigation and Monitoring of Contaminated Sites, Karlsruhe, 29 September–1 October, pp. 251–257.


439

Fedorak, P. M. and Westlake, D. W. S.: 1981, ‘Degradation of aromatics and saturates in crude oil by soil enrichments’, Water, Air and Soil Pollut. 16, 367–375. Fine, P., Graber, E. R. and Yaron, B.: 1997, ‘Soil interactions with petroleum hydrocarbons: Abiotic processes’, Soil Technol. 10, 133–153. Gough, M. A. and Rowland, S. J.: 1990, ‘Characterisation of the unresolved complex mixtures of hydrocarbons in petroleum’, Nature 344, 648–650. Haeseler, F., Blanchet, D., Druelle, V., Werner, P. and Vandecasteele, J. P.: 1999, ‘Analytical characterization of contaminated soils from former manufactured gas plants’, Environ. Sci. Technol. 33, 825–930. Hund, K. and Traunspurger, W.: 1994, ‘Ecotox-evaluation strategy for soil bioremediation exemplified for a PAH-contaminated site’, Chemosphere 29, 371–390. Hutchins, S. R., Sewell, G. W., Kovacs, D. A. and Smith, G. A.: 1991, ‘Biodegradation of aromatic hydrocarbons by aquifer micro-organisms under denitrifying conditions’, Environ. Sci. Technol. 25, 68–76. Joergensen, R. G., Figge, R. M. and Kupsch, L.: 1997, ‘Microbial decomposition of fuel oil after compost addition to soil’, Z. Pflanzenernähr. Bodenk 160, 21–24. Killops, S. D. and Al-Juboori, M. A. H. A.: 1990, ‘Characterisation of the unresolved complex mixture (UCM) in the gas chromatograms of biodegraded petroleum’, Org. Geochem. 15, 147–160. Kowalska, M., Güler, H. and Cocke, D. L.: 1994, ‘Interactions of clay minerals with organic pollutants’, Sci. Total Environ. 141, 223–240. Langbehn, A. and Steinhart, H.: 1995, ‘Biodegradation studies of hydrocarbons in soils by analysing metabolites formed’, Chemosphere 30, 855–868. Li, X., Feng, Y. and Sawatsky, N.: 1997, ‘Importance of soil-water relations in assessing the endpoint of bioremediated soils. I. Plant growth’, Plant and Soil 192, 219–226. Li, Y. and Gupta, G.: 1994, ‘Adsorption/desorption of hydrocarbons on clay minerals’, Chemosphere 28, 627–638. Mahaney, P. A.: 1994, ‘Effects of freshwater petroleum contamination on amphibian hatching and metamorphosis’, Environ. Toxicol. Chem. 13, 259–265. Margesin, R. and Schinner, F.: 1997, ‘Bioremediation of diesel-oil contaminated alpine soils at low temperatures’, Appl. Microbiol. Biotechnol. 47, 462–468. Means, J. C., Wood, S. G., Hassett, J. J. and Banwart, W. L.: 1980 ‘Sorption of polynuclear hydrocarbons by sediments and soils’, Environ. Sci. Technol. 14, 1524–1528. Morgan, P. and Watkinson, R. J.: 1989, ‘Hydrocarbon degradation in soils and methods for soil biotreatment’, CRC Crit. Rev. Biotech. 4, 305–333. Oudot, J., Dupont, J., Haloui, S. and Roquebert, M. F.: 1987, ‘Biodegradation potential of hydrocarbon-assimilating tropical fungi’, Soil Boil. Biochem. 25, 1167–1173. Oudot, J. and Dutrieux, E.: 1989, ‘Hydrocarbon weathering and biodegradation in a tropical estuarine ecosystem’, Mar. Environ. Res. 27, 195–213. Oudot, J., Amblès, A., Bourgeois, S., Gatellier, C. and Sebyera, N.: 1989, ‘Hydrocarbon infiltration and biodegradation in a landfarming experiment’, Environ. Pollut. 59, 17–40. Oudot, J.: 2000, ‘Biodégradabilité du fuel de l’Erika’, CR Acad. Sc.i III-Vie. 323, 945–950. Pollard, S. J. T., Hrudey, S. E. and Fedorak, P. M.: 1994, ‘Bioremediation of petroleum- and creosotecontaminated soils: A review of constraints’, Waste Manage. Res. 12, 173–194. Pollard, S. J. T., Whittaker, M. and Risden, G. C.: 1999, ‘The fate of heavy oil wastes in soil microcosms I: A performance assessment of biotransformation indices’, Sci. Total. Environ. 226, 1–22. Providenti, M. A., Lee, H. and Trevors, J. T.: 1993, ‘Selected factors limiting microbial degradation of recalcitrant compounds’, J. Ind. Microbiol. 12, 379–395.

440


Ramirez, N. E., Vargas, M. C. and Sanchez, F. N.: 1996, ‘Use of the sediment-chromotest for monitoring stimulated hydrocarbon biodegradation processes’, Environ. Toxicol. and Water Qual. 11, 223–230. Rhykerd, R. L., Crews, B., McInnes, K. J. and Weaver, R. W.: 1999, ‘Impact of bulking agents, forced aeration, and tillage on remediation of oil-contaminated soil’, Bioresource Technol. 67, 279–285. Salanitro, J. P., Dorn, P. B., Hueseman, H., Moore K. O., Rhodes, I. A., Rice Jackson, L. M., Vipond, T. E., Western, M. M. and Wisniewski, H. L.: 1997, ‘Crude oil hydrocarbon bioremediation and soil ecotoxicity assessment’, Environ. Sci. Technol. 31, 1769–1776. Sawatsky, N. and Li, X.: 1997, ‘Importance of soil-water relations in assessing the endpoint of bioremediated soils. II. Water-repellency in hydrocarbon contaminated soils’, Plant and Soil 192, 227–236. Shelton, M. E., Chapman, P. J., Foss, S. S. and Fisher, W. S.: 1999, ‘Degradation of weathered oil by mixed marine bacteria and the toxicity of accumulated water-soluble material to two marine Crustacea’, Ach. Environ. Contam. Toxicol. 36, 13–20. Suguira, K., Ishihara, M., Shimauchi, T. and Harayama, S.: 1997, ‘Physicochemical properties and biodegradability of crude oil’, Environ. Sci. Technol. 31, 45–51. Thouand, G., Bauda, P., Oudot, J., Kirsch, G., Sutton, C. and Vidalie, J. F.: 1999, ‘Laboratory evaluation of crude oil biodegradation with commercial or natural microbial inocula’, Can. J. Microbiol. 45, 106–115. Tukaj, Z. and Szurkowski, J.: 1993, ‘Photoacoustic spectra affected by fuel oil in the chlorococcal alga Scenedesmus armatus’, Acta. Physiol. Plant. 15, 219–226. Wang, X. and Bartha, R.: 1990, ‘Effects of bioremediation on residues, activity and toxicity in soil contaminated by fuel spills’, Soil Biol. Biochem. 22, 501–505. Wetzel, A. and Werner, D.: 1995, ‘Ecotoxicological evaluation of contaminated soil using the legume root nodule symbiosis as effect parameter’, Environ. Toxicol. Water Qual. 10, 127–133. Wrenn, B. A. and Venosa, A. D.: 1996, ‘Selective enumeration of aromatic and aliphatic hydrocarbon degrading bacteria by a most-probable number procedure’, Can. J. Microbiol. 42, 252–258. Xu, J. G. and Johnson, R. L.: 1995, ‘Root growth, microbial activity and phosphatase activity in oilcontaminated, remediated and uncontaminated soils planted to barley and field pea’, Plant and Soil 173, 3–10.