Laboratory evaluation of crude oil biodegradation with commercial or ...

22 downloads 0 Views 186KB Size Report
Abstract: Experiments have been performed to screen eight microbial commercial products that, according to the manufacturers, are able to degrade crude oil.
106

Laboratory evaluation of crude oil biodegradation with commercial or natural microbial inocula1 G. Thouand, P. Bauda, J. Oudot, G. Kirsch, C. Sutton, and J.F. Vidalie

Abstract: Experiments have been performed to screen eight microbial commercial products that, according to the manufacturers, are able to degrade crude oil. This study compared the crude oil biodegradation activity of commercial inocula with that of natural inocula (activated sludge and tropical aquarium water). Some of the latter were previously adapted to the crude oil as the only carbon source. Nutrients and sorbents in the commercial formulations were eliminated, and each inoculum was precultured on marine yeast extract medium. Crude oil biodegradability tests were conducted with close initial substrate concentration to initial bacterial concentration ratios (S0/X0) of 0.94 g of crude oil/109 CFU, which allowed a comparison of biodegradation activity. The inocula oxidized the crude oil after a short lag time of less than 3–18 days. After that time, the rate of oxidation varied between 45 and 244 mg O2/(L·day). Crude oil biodegradation after a 28-day test was effective only for 10 out of 12 inocula (from 0.1 to 25% in weight). Biodegradation mainly corresponded to the saturated fraction of the crude oil; the asphaltene fraction was never significantly biodegraded. Our results led to the conclusion that natural inocula, either adapted or not adapted to crude oil, were the most active (from 16 to 25% of loss in crude oil weight) and only one commercial inoculum was able to degrade 18% of the crude oil. Other inocula had a biodegradation activity ranging from 0.1 to 14%. Key words: biodegradability tests, microbial inoculum, crude oil, seeding. Résumé : Une étude a été menée afin de comparer l’activité de biodégradation de huit inocula commerciaux aptes, selon leur fournisseur, à dégrader un pétrole brut. Leur activité de biodégradation a été comparée à celle d’inocula naturels (boue activée de station d’épuration d’effluent urbain, eau d’aquarium tropical), préalablement adaptés ou non au pétrole brut. Les substances nutritives ainsi que l’adsorbant présents dans les formulations commerciales ont été éliminés, puis chaque inoculum a été pré-cultivé dans un milieu marin contenant de l’extrait de levure. Les essais de biodégradation du pétrole brut ont été conduits avec des rapports S0/X0 voisins de 0,94 g de pétrole brut pour 109 unités formant des colonies, autorisant la comparaison des activités de biodégradation. Les 12 inocula ont oxydé le pétrole brut après un temps de latence de moins de 3 jours à 18 jours. Passé ce temps, le taux d’oxydation du pétrole était compris entre 45 et 244 mg O2/(L·jour). Cependant, la biodégradation du pétrole brut après 28 jours d’essai a été effective pour seulement 10 inocula parmi les 12 (de 0,1 à 25% de perte en masse). La biodégradation correspond principalement à l’utilisation de la fraction saturée du pétrole alors que la fraction asphaltène n’est jamais significativement utilisée. Nos résultats démontrent que les inocula naturels, adaptés ou non au pétrole brut, ont été les plus actifs (de 16 à 25% de perte en masse de pétrole brut). Un seul inoculum commercial a été capable de biodégrader 18% du pétrole brut alors que les autres inocula commerciaux biodégradent entre 0,1 et 14% de pétrole. Mots clés : essais de biodégradabilité, inoculum microbien, pétrole brut, amendement (seeding). Thouand et al. 115 Received June 18, 1998. Revision received September 25, 1998. Accepted October 24, 1998. G. Thouand2 and P. Bauda. Laboratoire de génie génétique et microbiologique, Centre des sciences de l’environnement, 1 rue des Récollets, 57000, Metz-France. J. Oudot. Muséum national d’histoire naturelle, Laboratoire de cryptogamie, 12, rue Buffon, Paris 75005, France. G. Kirsch. Laboratoire de chimie organique, Université de Metz, Faculté des sciences, Ile du Saulcy, 57000, MetzFrance. C. Sutton and J.F. Vidalie. Société TOTAL, Tour TOTAL, Service TEP/SEV, 24, cours Michelet, Cedex 47, Paris 92069, France. 1 2

To the memory of Pr. Bernard Capdeville. Author to whom all correspondence should be addressed (e-mail: [email protected]).

Can. J. Microbiol. 45: 106–115 (1999)

Introduction Crude oil can be accidentally or deliberately released to the sea, leading to the pollution of several sites, and can eventually reach the coasts (Atlas 1995a; Oudot 1994). The restoration of these polluted sites, especially the coasts, is the responsibility of oil companies or governments that have crude oil remediation methods based on chemical or physical treatments (e.g. use of dispersants, burning, photooxidation agents; Prince 1993). These methods can also be complemented or replaced with so-called bioremediation techniques that are less expensive (Atlas 1995b). Bioremediation may be defined as any activity encouraging the natural process of crude oil biodegradation (Prince 1993). Natural biodegradation occurs because most natural environments contain a number of bacteria and fungi © 1999 NRC Canada

Thouand et al.

(autochthonous microorganisms) that are able to transform part of the crude oil (Atlas 1995b; Kämpfer et al. 1991, 1993; Oudot et al. 1987, 1993). Biotreatment techniques are based on two methods. First, biodegradation by autochthonous microorganisms can be activated by nutrient additions (N and P), which are often limiting factors in the environment. The success of this method has been demonstrated in several pollution cases such as the Exxon–Valdez wreck (Bragg et al. 1994). The second method consists of adding allochthonous microbial consortia (seeding), some of which are commercially available, selected for their ability to biodegrade crude oil. They are sold to petroleum companies with doubtful scientific data that do not allow comparison between consortia. Comparative experiments were conducted a few years ago in the laboratory and in situ (Atlas 1995a, 1995b; Dott et al. 1989; Grundmann and Rehm 1991; Venosa et al. 1992a, 1992b). These studies demonstrated that commercial products did not degrade significantly more crude oil than natural microbial communities, with or without nutrient supplies. In these controlled experiments, commercial formulations (microorganisms fixed on a support, with nutrients and different microbial concentrations) were used. Under these conditions, two opposite hypotheses were formulated: (i) that the negative results could be explained by the poor biodegradation capacity of commercial formulations, or (ii) that the allochthonous microorganisms are intrinsically able to biodegrade crude oil, but their activity is impaired either by autochthonous microorganisms or by the added sorbents and nutrients of commercial formulations. Our study differed from the others because it compared the crude oil biodegradation activity of commercial inocula with that of natural inocula (activated sludge and tropical aquarium water) without any sorbents, nutrients, or autochthonous microorganisms, and with close initial bacterial concentrations. For comparison purposes, the biodegradation activity of the eight commercial inocula were compared with that of the two natural inocula, one preadapted to crude oil as the only carbon source and one not preadapted. Nutrients and sorbents in the commercial formulations were eliminated because Rittmann et al. (1994) showed that different supports and nutrients added in variable proportions to a microbial inoculum led to random activity. Each inoculum was then precultured on marine yeast extract medium to produce bacterial biomass. The crude oil biodegradation tests were conducted with the same initial bacterial concentration (X0) to allow a comparison between free-living microorganisms. Simkins and Alexander (1984) and Chudoba et al. (1992) demonstrated for one initial substrate concentration (S0) that the kinetics order and the rate of biodegradation depend on the X0. This key parameter significantly influences the lag time and the success of laboratory biodegradability tests (Thouand et al. 1995).

Materials and methods Culture media Five different media were used. The compositions below were for 1 L of MilliQ water (Millipore). Medium 1 (basic medium) contained 380 mg NH4Cl, 44 mg KH2PO4, and 33 g of mixed salts, to reach a final salinity of 26% w/v (Reef Crystal, Aquarium Sys-

107 tems, Sarrebourg, France; Courtes 1993). After sterilization, 1000 mg·L–1 of crude oil (BAL 250) was added (100 µL of BAL mass is 92.24 mg at 25°C). Medium 2 was used to test the bacterial activity because it was recommended for standard-ready biodegradability tests (EEC standard methods C4A–C4D for easy biodegradability tests; EEC 1992). This medium contained 300 mg (201 mg C) sodium benzoate (Merck), 78 mg NH4Cl, 15 mg KH2PO4, and 33 g of mixed salts (Reef Crystal). Medium 3 contained 1500 mg (540 mg C) yeast extract (Pasteur), 210 mg NH4Cl, 24 mg KH2PO4, and 33 g of mixed salts (Reef Crystal). Medium 4 contained 55.1 g Bacto marine agar (Difco). Medium 5 contained 380 mg NH4Cl, 44 mg KH2PO4, 33 g of mixed salts (Reef Crystal), and 15 g of agar–agar (Difco). The composition of the media (except that of medium 4) gave a C/N/P ratio of approximately 100/10/1, by mass. Each medium was sterilized by autoclaving 20 min at 120°C. The pH was initially adjusted to 7 as is usual in standard biodegradability tests. All reagents were over 99% pure.

Microbial inocula Eight different commercial products were tested in this study. Their identities cannot be revealed (as required by the manufacturers), and coded names were used as follows (information from the manufacturers are in parentheses): inoculum A (petroleum, oil, and grease); inoculum B (hydrocarbons, halogenated hydrocarbons, and aromatic hydrocarbons); inoculum C, supplied in liquid form, but usually in powder form (degradation of light crude oil); inoculum D (biodegradation of organic, mineral, plant, and animal wastes); inoculum E (crude oil and refinery products in marine environments); inoculum F (hydrocarbons, oil, grease); inoculum G (petroleum products, grease), and inoculum H (light and heavy fractions of hydrocarbons). Commercial microbial inocula were stored at 4°C and used for preculture before 30 days, as recommended in the manufacturers’ specifications. One activated sludge inoculum (AS1) from an urban municipal waste-water treatment plant (Metz, France) and one inoculum from a marine tropical aquarium (AW1) were used in this study (Nancy, France).

Preparation of the activated sludge The treatment consisted of taking 200 mL of activated sludge, which was sonicated for 20 s with a work volume of 40 mL (power output, 50 W; soniprobe, 19-mm diameter; frequency, 20 kHz), and subsequently filtering through a bolting cloth (5-µm mesh size) to remove mineral particle clumps and protozoa. The filtrate was centrifuged for 15 min at 5000 × g and the pellet was suspended in 100 mL of sterile Reef Crystal. Ten millilitres of suspension was then used for adaptation experiments and the remaining part was centrifuged a second time as above and finally suspended in 9 mL of Reef Crystal with 1 mL of a 40% sterile glycerol solution before freezing at –20°C.

Preparation of the tropical aquarium water inoculum The treatment consisted of taking 500 mL of synthetic marine water containing marine bacteria that were obtained from a manually wrung polyurethane foam stick from the aquarium. It was subsequently filtered through a bolting cloth as described above. The dissolved organic carbon contamination was limited by centrifuging each inoculum for 15 min at 5000 × g (Courtes 1993). The pellet was suspended in 100 mL of sterile synthetic seawater (Reef Crystal) and a 10-mL aliquot was used for adaptation. The remaining part was stored as indicated above. The inocula were frozen to be used later in biodegradability tests. © 1999 NRC Canada

108

Adaptation of natural inocula to crude oil (BAL 250) Natural inocula (AS1 and AW1) were adapted over 43 days and three steps in 100 mL of medium 1 in 250-mL sterile glass flasks. The transfer of microorganisms from one step to another was based upon macroscopic observations (observation of turbidity, disrupting of the crude oil in small pellets) and microscopic observations. The first step (12 days) was done in duplicate with the addition of the natural inocula (10% v/v). During the second step (16 days), 10% of the inoculum from step 1 was transferred into two new flasks. Adaptation was achieved with a third step lasting 15 days as above but with only 1% of the inoculum from step 2. The inoculum was aerated daily with 100 mL of sterile air (Millipore, 0.22-µm membrane Millex GS) injected for 30 s into the flasks with a compressor (Ets Piot et Tiroufflet). The flasks were incubated at 25°C without shaking. The resulting adapted inocula from step 3 were then frozen as indicated above and were called AS2 for the adapted activated sludge and AW2 for the adapted tropical aquarium water.

Activity of microbial inocula for crude oil oxidation and biodegradation For each inoculum, two steps were necessary. All experiments were performed under sterile conditions.

Rehydration and preculture Microbial product (5 g or 5 mL of liquid inocula) was rehydrated in 5 mL of nutrient broth (bioMérieux) diluted in Reef Crystal for 15 min at room temperature. Reef Crystal (15 mL) was then added. The inoculum was sonicated for 30 s and centrifuged for 1 min at 2500 × g to settle the sorbent (wood particles, stone). Preculture started with 100 mL of medium 3 in a 1-L Erlenmeyer flask inoculated with the above supernatant to reach a final A620 of 0.1–0.2. The flasks were incubated at 25°C and shaken at 150 rpm. The microbial growth was monitored up to the exponential phase with absorbance at 620 nm. At this stage of growth, each inoculum was centrifuged at 6500 × g for 15 min and the pellet was suspended in 50 mL of Reef Crystal solution. The preculture was then characterized by the total bacterial number (epifluorescence), the total cultivable bacterial number, and the cultivable crude oil biodegrader number. The free microorganisms obtained were then called commercial inocula.

Biodegradability test of crude oil BAL 250 crude oil was supplied by TOTAL and corresponded to an Arabian light crude oil topped by distillation at 250°C. It was composed of 75.6% saturated hydrocarbons, 13.5% aromatic hydrocarbons, and 7.6% polar compounds (6.6% of asphaltenes). The choice of a heavy fraction took into account the elimination of the most volatile fraction that could vanish without any biological action (evaporation of compounds up to C14; J. Oudot, personal communication). Each test comprised three 250-mL biological oxygen demand flasks with 100 mL of medium 1 and twenty 14-mL tubes (Pierce) with 4 mL of medium 1 sealed with a Teflon cap. Each flask and tube were inoculated with each precultured inoculum to reach 1010 total cells·L–1 (X0, initial concentration in the test). The total bacteria number was then checked in the flasks by epifluorescence count before starting the test. The cell concentration was chosen according to Thouand et al. (1996). The crude oil (100 mg) was then added to the surface of the medium. Preliminary experiments showed that heterogeneity of the BAL 250 addition was less than 18%. For aeration, each Pierce tube was fitted with two sterile needles pierced through the Teflon cap (one inlet, one outlet). Sterilized air (about 10 mL) obtained after filtration through a 0.22-µm filter (Gelmann and Millex GS) was injected daily into the air space of each tube for 10 s with a compressor (Ets Piot et Tiroufflet). The 250-mL flasks were simply opened un-

Can. J. Microbiol. Vol. 45, 1999 der sterile conditions for 5 min. This aeration was as efficient as bubbling (unpublished results). Biodegradability tests were incubated for 28 days at 25°C (± 1°C) and were not shaken to avoid projection of oil onto the flask walls. Crude oil oxidation was monitored every 3 days by chemical oxygen demand (COD) duplicate determination of the Pierce tube contents. After 28 days of incubation, the growth of microorganisms was evaluated by the total bacteria, the total cultivable bacteria, and the cultivable crude oil biodegrader numbers. The flasks were then pooled for chemical analysis and the crude oil biodegradation was monitored with gravimetric determination of crude oil fractions. Four controls were set up to confirm the biodegradation activity of the inocula, and they were conducted under the same conditions as the assays: (i) a control without crude oil (Ci) took into account carbon contamination of the inoculum, and COD and cellular density were determined; (ii) a control of microbial activity (Ca) in medium 2, and biodegradation was followed by COD determination (the yield of sodium benzoate oxidation was 100%); (iii) a negative control with dead cells (Cd) was performed in 10 Pierce tubes (five duplicates) containing the same mass of crude oil, the same number of bacteria as in the assays, and 1% of mercuric chloride (50 g·L–1), and the action of mercuric chloride was measured by cultivable bacteria determination in medium 4 at days 0 and 28; (iv) an abiotic control (Cabio) was added to the above and run in duplicate to measure abiotic loss of crude oil. No cultivable bacteria were detected after 28 days incubation in medium 4 (results not shown).

Bacterial analyses The total bacteria number was determined by direct count with epifluorescence microscopy adapted from King and Parker (1988) and Porter and Feig (1980). Samples stored in 5% formaldehyde (Merck, Cat. No. 4003) were diluted in 0.1 M Tris–HCl buffer, pH 7.1 (sterilized by filtration through an analypore 0.22-µm filter). A stock solution (1 mL) of 4′,6-diaminido-2-phenylindol (DAPI, Sigma, Cat. No. D1388) at a final concentration of 25 µg·mL–1 was filtered through a 0.22-µm filter before adding to the sample. After stirring for 15 min, the mixture was filtered under low pressure through a premoistened black polycarbonate filter (Millipore, Cat. No. GTBP04700; 47-mm diameter, 0.22-µm porosity). Slides were read by epifluorescence microscopy (λexcitation = 365 nm; λemission > 390 nm). Thirty microscopic fields were randomly counted. The number of total cultivable bacteria was determined in medium 4 after spreading 0.1 mL inoculum, either diluted or not, in Reef Crystal solution. The petri dishes were incubated for 5 days at 25°C. After incubation, the colonies were counted and expressed as colony-forming units (CFUs) per litre of inoculum. The number of cultivable crude oil biodegraders was determined after spreading 0.1 mL of inoculum (diluted or not in Reef Crystal solution) on the surface of medium 5. When all the liquid was absorbed, 25 µL of nonsterile BAL 250 was evenly spread on the surface of the agar medium. The plates were incubated for 28 days at 25°C. After incubation, colonies were white with a dark halo. They were expressed as CFUs·L–1 of inoculum after correcting for CFUs grown on medium 5 without BAL 250. The crude oil did not initially contain any cultivable crude oil biodegrader.

Chemical analyses The COD was used to monitor the oxidation of a part of the crude oil by the inocula (the oxidation yield was 47%). Experiments were done with a COD microkit Merck spectroquant (ref. 14540 or 14541). The chloride ions of synthetic seawater (Reef Crystal) were complexed with the Macherey Nagel reactant (Nagel, NANOCOLOR, 918.911) with the same yield as that with AgNO3 (data not shown). The COD of the crude oil medium was © 1999 NRC Canada

Thouand et al.

109

Table 1. Bacterial characteristics of the eight commercial and four natural inocula after preculture in medium 3.

Inocula

Growth rate (h–1)a

Total bacteria (×109 cells·L–1)b

Cultivable crude oil biodegraders (×109 CFU·L–1)c

Proportion of crude oil biodegraders (%)d

S0/X0 (g crude oil/109 CFU)e

A B C D E F G H AS1 AS2 AW1 AW2

0.45 0.49 0.36 0.40 0.34 0.27 0.44 0.40 ND ND 0.46 0.31

8.30 6.5 10 2.9 6.9 5.3 5.8 3.9 8 7.8 3 8.2

2 0.8 1.7 0.014 0.66 1.3 1.8 1.1 1.4 1.7 0.47 1.25

24 12 16 0.5 10 24 32 27 17 22 16 15

0.5 1.25 0.59 71.4 1.5 0.77 0.56 0.91 0.71 0.59 2.13 0.8

Notes: Concentrations are those in the biodegradability tests at day 0. Inocula A–H were the commercial inocula compared with the natural inocula from an urban treatment plant activated sludge (AS) and tropical aquarium water (AW) either unadapted to crude oil (AS1 and AW1) or adapted (AS2 and AW2). ND, not determined. a Growth rates on medium 3. Preculture was performed in batch after rehydration of the product in medium 3 at 25°C, 150 rpm. Growth was followed by the absorbance at 620 nm. b Direct count with epifluorescence. c CFUs on medium 5 with crude oil after deduction of the control (CFU on medium 5 without crude oil). d Specific cultivable degraders/total cells number (epifluorescence). e S0, initial crude oil mass for 1 L of medium; X0, initial cultivable crude oil biodegraders for 1 L of medium.

determined after sonication for 1 min with a 25-W sonotrode. The suspension (1 mL) was immediately added to the COD tube containing 2 mL of Macherey Nagel complexing agent. After vortexing, the tubes were placed in a thermoreactor (TR205, Merck) for 2 h at 150°C. The COD (in mg O2·L–1) was obtained after reading a photometer (SQ 118, Merck). Gravimetric determination of crude oil fractions was performed according to Alula et al. (1989). All three flasks of each biodegradability experiment were pooled and extracted three times with chloroform (3 × 20 mL). The combined organic fractions were dried with sodium sulfate, filtered, and concentrated. All solvent evaporations were done under gentle vacuum with water bath temperatures not exceeding 40°C to avoid any loss of the light fractions. The resulting degraded crude oil was fractionated with hexane into two main fractions: a hexane-soluble maltene fraction (HS fraction including saturated hydrocarbons, aromatic hydrocarbons, and resin fraction) and a hexane-insoluble fraction (HI fraction). The latter contained an asphaltene fraction (AS) and some polar compounds produced by the microbial metabolism. HI and HS were separated by filtration through a Millipore SC 8-µm filter, followed by rinsing with hexane. After drying, the HI fraction was weighed and the HS fraction was concentrated until a weight was obtained. After weighing, the HS fraction was dissolved in 200 mL of chloroform and 7 g of silica gel (Geduran Si 60, Merck) was added. After stirring for 15 min, the chloroform was distilled off, leaving a dry silica gel with oil absorbed onto it. The silica gel was used to make a column (0.5-cm diameter), which was successively eluted with three 20-mL fractions of cyclohexane, benzene, and methanol. The three fractions were weighed after evaporation of the solvent and were named as follows: S fraction (elution with cyclohexane: saturated hydrocarbons), A fraction (elution with benzene: aromatic hydrocarbons), and R fraction (elution with methanol: resins). Gas chromatographic analysis was carried out in some cases to confirm the biodegradation activity of the inoculum. The analysis was achieved with a Carlo Erba gas chromatograph (GC 6000 VEG A). The apparatus was equipped with a direct injection port and a flame ionization detector. The glass capillary column was a CP Sil 5 CB (25-m length, 0.32-mm diameter). Temperature was

programmed from 60°C (for 3 min) to 300°C (for 30 min), at 4°C·min–1. The carrier gas was He (1.1 bar flow rate).

Results The comparison of crude oil biodegradation using different inocula was achieved in two steps. The first one was an initial preculture of microbial inocula on medium 3 to adjust the cellular density to the same value. The second step was the crude oil biodegradation assay followed by COD measurements (oxidation rate), gravimetric analysis of the residual oil (biodegradation rate), and bacterial enumerations. Preculture and characteristics of the inocula at the beginning of the assay All the inocula grew rapidly on medium 3 with similar growth rates (growth rate between µ = 0.27 and 0.49·h–1 at 25°C, Table 1). When the exponential phase was reached, the inocula were characterized with the total bacteria number (epifluorescence), the total cultivable bacteria number, and the cultivable crude oil biodegrader number. The precultures produced between 0.26 × 1012 and 3.82 × 1012 total cells·L–1. Inocula were then diluted to adjust the total cell number in the crude oil biodegradability test. Table 1 shows characteristics of inocula at the beginning of each crude oil test. The total bacteria number was between 2.9 × 109 and 10 × 109 cells·L–1. The ratio of cultivable crude oil biodegraders to total bacteria after precultures and at the beginning of each crude oil biodegradability test was the same, with an average of 19.5% (SD = 6.8%), except for that of inoculum D (see upcoming section for more details). At the beginning of each test, commercial and natural inocula did not differ, even if the latter was adapted to crude oil as the only carbon source. The average total bacteria number was 6.2 × 109 © 1999 NRC Canada

110

Can. J. Microbiol. Vol. 45, 1999 Table 2. Crude oil biodegradation activity and bacterial characteristics after 28-day experiments for the 12 inocula tested. Oxidation rate (mg O2/(L·day))b

Crude oil degradation (% weight)c

Crude oil oxidation (% COD)d

Production of cells

Inocula

Lag phase (days)a

Total cellse

Degradersf

A B C D E F G H AS1 AS2 AW1 AW2

3 0 3 ND 3 18 0 ND