Published March, 2000
Bioremediation and Biodegradation Biodegradation of Fuel Oil Hydrocarbonsin the Rhizosphere of Maize C. H. Cha~neau, J. L. Morel, and J. Oudot* ABSTRACT Plant roots provide suitable habitats for the growthof microorganisms. Particularly, the release of exudates by living roots enhances the microbial activity in the rhizosphereo This workwas undertaken in the laboratory to study the biodegradationof fuel oil hydrocarbons (HC) in the presence of growingplants and to assess the effects of root exudates on biodegradation. Maize (Zea mays L.) plants were grownin soil microcosmsfor 120 d andin liquid (hydroponic) cultures for 40 d in presence and in absence of fuel oil added at slightly phytotoxic concentrations (soil: 3300 mgkg-l; hydroponiccultures: 850 mg L-~). Controls without plants permitted the quantification of the rhizosphere effects on HCremoval. Concentration and chemical composition of residual HCwere periodically determined. Higher degradation rates of saturated and aromatic HCwere observed in soil in the presenceof plants in the early stages of biodegradationwhen maize growth was exponential. No significant change was observed in the polar fraction. After 120 d, the biodegradationrates wereidentical in the presence or absence of plants. In hydroponic conditions, a stimulation in the degradation of saturates was observed until Day 40. Nosignificant difference wasdetected in the aromaticfraction. The faster biodegradation of HCin the presence of plants was attributed to the changes in environmentalconditions in the rhizosphere; e.g., increase of HCbioavailability, stimulation of bacterial populations due to plant exudation and effects on physical properties of soil. Competition for mineral nutrients betweenplants and microorganisms was shown.
yt
kIE INCREASING USE of petroleum products leads to he contamination of the environment. Hydrocarbons can reach the soil from many sources, including pipeline blowouts, road accidents, leaking of underground storage tanks, landfarming fields and noncontrolled landfilling. The release of oily wastes containing toxic HChas been shown to cause serious damage to natural ecosystems. The vegetation can be totally eliminated for a long period of time in the case of high inputs of HC (Amakiri and Onofeghara, 1983; Bossert and Bartha, 1985; Terje, 1984). In the soil, HCare subjected to physical processes (evaporation, leaching, adsorption) and biological processes (biodegradation) (Atlas, 1991; Bossert and Bartha, 1984; Leahy and Colwell, 1990; Oudot et al., 1989). Residual HCmaypersist for a long time (Atlas, 1991; Bossert and Bartha, 1985; ChaTneau et al., 1995, 1996; Oudot, 1984), and both HC and their metabolites may contaminate the groundwater
C.H. Chalneau and J. Oudot, MusEumNational d’Histoire Naturelle, Laboratoire de Cryptogamie, 12 rue Buffon, 75005 Paris, France; J.L. Morel, Ecole Nationale Sup6rieure d’Agronomie et des Industries Alimentaires de Nancy, Laboratoire Sols et Environnement, INRA, 2 Avenue de la For~t de Haye, B.P. 172, F-54505 Vandoeuvre-LesNancy, C6dex, France. Received 25 Jan. 1999. *Corresponding author (
[email protected]).
(Cozzarelli et al., 1995). Physicochemicaland biologica| treatments involving microorganisms can be used in the decontamination of HC-polluted soils. Phytoremediation, i.e., the use of plants for the remediation of contaminated environments, has been proposed as an alternative or complementary technique to treat soils polluted by mineral and organic compounds [heavy metals, pesticides, halogenated molecules, polycyclic aromatic hydrocarbons (PAH)] (Boyle Shann, 1998; Schnoor et al., 1995; Walton et al. 1995). The action of plants may include the direct uptake of pollutants (Simonich and Hites, 1995). Plant roots also release exudates that stimulate the activity of microorganisms in the rhizosphere (Barber and Lynch, 1977; Benizri et al., 1995), thus increasing the rates of microbial biodegradation in the rhizosphere (Walton et al., 1995). The use of phytoremediation has been recently reviewed (Schnoor et al., 1995). The role of plant roots in the elimination of HCis still being debated (Wackett and Allan, 1995). In the field, the determination of biodegradation rates of HCnecessitates a very high number of analyses to obtain statistically significant results (Chaineauet al., 1996). Radiolabeled substrates (Banks et al., 1999), which can be realized only with pure compounds, can also be used. The estimation of microbial populations is another indicator of microbial activity (Jordahl et al., 1997), but this should be correlated with biodegradation rates determined by chemical analyses (Cha~neauet al., 1995, 1996). Studies on the potential effects of plants on biodegradation of xenobiotics should take into account the agronomic, microbiological, and chemical parameters for a complete evaluation of the efficiency of the technique. Laboratory experiments allow a partial simulation of the full-scale process. Previous works showed that the biodegradation of fuel oil HCcould be accurately modeled in small-scale soil microcosms (Chalneau et al., 1995, 1996). In another work, a reasonably good growth of maize was obtained in small containers placed in climatic chambers (Chaineau et al., 1997). Even though some limitations inevitably exist in small-scale laboratory experiments, the results gained from such studies are valuable in the global evaluation of a process. In a survey of the influence of fuel oil on the development of cultivated plants, it was shownthat at fuel concentrations below 1%, the germination and growth of Abbreviations: ANOVA, analysis of variance; CE, chloroform extract; CFU, colony-forming units; GC, gas chromatography; HAB,hydrocarbon-adapted bacteria; HC, hydrocarbons; MPN,most probable number; PAH, polycyclic aromatic hydrocarbons; R/S, rhizospheric microorganisms/nonrhizospheric microorganisms; THB, total heterotrophic bacteria; UCM,unresolved complex mixture.
Published in J. Environ. Qual. 29:569-578 (2000).
569
570
J. ENVIRON.QUAL., VOL. 29, MARCH-APRIL 2000
maize were only slightly affected, whereas the development of other crops was drastically reduced (Chaineau et al., 1997; Amakiri and Onofeghara, 1983). Also, maize produces root exudates, which are known to have a positive influence on microbial populations (Benizri et al., 1995; Morelet al., 1991; Schnooret al., 1995; Walton et al., 1995). Theseconsiderations led us to select maize as a model plant for the study of the influence of the rhizosphere on HCbiodegradation. In this work, the influence of the maize rhizosphere on fuel oil biodegradation was evaluated in laboratory soil experiments and in hydroponic cultures. MATERIALS
AND METHODS
Soil Microcosms Previous experimentsin soil microcosmscontaminatedwith fuel oil showedthat the entire samplehad to be analyzed to obtain reliable results of HCconcentrations(Cha~neau et al., 1995).Partial samplingled to uncertainties causedby the heterogeneity of the medium.Onthe other hand, the amountof soil in the microcosmshad to be large enough to allow a maizegrowthsufficient for the establishmentof a rhizosphere (Benizri et al., 1995;Andersonet al., 1997;Chaineauet al., 1997). This was achieved by growingmaize plants in 700-mL beakerscontaining200g of soil, as justified in a previousstudy (Chalneauet al., 1997). The Ap horizon of an eluviated brownsoil (Mollic eutrochrept), already described (Chaineauet al., 1996), was airdried and sieved (2-ram-diameter openings). Twohundred gramsof dry soil werepackedinto a 700-mLglass beaker (7.5 cmdiameter). Fuel oil was addedto the soil at a rate of 800
Maize
Aluminium foil Cotton wool
Open-endglass tube Erlenmeyer flask Oil layer Nutrient solution
Roots
Fig. 1. Experimentaldevice for hydroponiccultures.
tzL per pot whichresulted in a concentration of 3300mgHC kg-~ soil (990 g HCm-Z). Soil and fuel werethoroughlymixed. This concentration was shownto be only slightly phytotoxic (Chaineauet al., 1996, 1997). The fuel oil was composed 60%saturated HC(nC12-nC26), 30% aromatics and 10% resins (Chaineauet al., 1995).The soil wassupplemented with a nutrient solution of 1 g N kg-1 soil (NO3NH4), 1 g P -~ soil (Na2HPO4 and KH2PO4), and 0.4 g K -1 soil (K H2PO4), whichis higher than the amountsnecessaryfor biodegradation of the fuel and sufficient for plant development(see discussion). A pregerminatedmaizeseedling 2 cmhigh was planted in each beaker. Unplantedpots were prepared similarly with and withoutfertilization. Controlswere(a) soil with HC,poisoned every 30 d with 250 mgkg-~ of HgC12in solution to quantify abiotic loss of HC,and (b) soil without HC,planted with maizeto quantify phytotoxicity of fuel oil. Unplanted controls without HCwere not set up since the biogenic HC contentof the soil is lowand doesnot varysignificantly (Chaineau et al., 1995). Each treatment was repeated 15 times, except on the Hg-treated soil, where the treatment was repeated 10 times. Cultureswere conductedfor 120d in climatic chambers[hygrometry65 %, temperature 22°C (16-h day) and 18°C(8-h night)]. Soil moisture was maintainedat 70%of the water-holdingcapacity with distilled water. Immediatelyafter fuel addition, one replicate of each treatment was collected to measurethe initial hydrocarbon concentrationin soil. Then, after 10, 20, 40, 80, and 120d, three replicates of each treatment, except the Hg-treated soil (one replicate only) were sampledand analyzed. The efficiency for HCrecoveryin sterile soils was 99_+4.1% (Cha~neau,1995). At each time of the experiment, the statistical difference betweengroups corresponding to HC-treated unvegetated soils and HC-treated soils with plants was evaluatedby one-wayanalysis of variance (ANOVA, F test). Probability was studied at the 0.01 and 0.05 levels accordingto the values of F(1,4), i.e, by comparison withFo.o5(7.7) andFo.ol(21.2). Hydroponic Cultures Maizeplants were grown under hydroponic conditions in Erlenmeyerflasks according to an adapted version of experimentaldevices described in Andersonet al. (1997). Eachflask contained 200 mLof a nutrient solution modified from the Hoagland (1950) medium(mg L-~: KH2PO4,340; Na2HPO4, 895; MgSO4-7H20, 123; NO3NH4, 500; Ca(NO3)z’4H20, 472; KC1,7.45; Fe EDTA,37; FeSO4, 400; CuSOn’5H20, 50; H3BO4, 618; MnSO4.H20, 54.5; ZnSO4,80.5). An open-ended glass tube was introducedinto the Erlenmeyerflask and fitted with cotton wool(Fig. 1). The device was sterilized at 120°Cfor 20 min. Fuel oil (200 IxL) was deposited at the surface the nutrient solution outside the tube. Onemilliliter of the supernatant of a soil/distilled water suspension (10%w/w) was addedto the nutrient solution as inoculum.Onepregerminated maizeseedling was placed into the tube with its roots dipped in the nutrient solution, thus avoidingdirect contact with the oil layer. Similar experimentaldevices wereprepared with oil but no plants, and plants withoutoil. Sterile controls without plants were kept abiotic with 500 mgL-1 HgCI2.Erlenmeyer flasks were entirely covered with aluminumfoil to protect roots fromlight. Periodic additionof sterile distilled water kept the level of the solution in the flasks constant. The culture medium was not periodically replaced, so as to be able to draw a nutrient balance at the end of the experiment. Cultures were placed for 40 d in the climatic chambersalong with the soil microcosms.After 0, 10, 20, and 40 d, three replicates of each treatment, and one of the sterile treatment, were sampled.Efficiency for HCrecoveryin sterile flasks was
571
CHA~NEAUET AL.: BIODEGRADATIONOF FUEL OIL HYDROCARBONS IN MAIZE
98_+3.6%(Chaineau, 1995). ANOVA statistical performedas for soil microcosms.
analyses were
Plant Biomass At each time of sampling, the height and dry biomassof the aerial parts of maizewere measuredin soil microcosms and hydroponiccultures. Plant root biomasswas also measuredin the case of hydroponiccultures after extraction of HC.
Nutrient Content in Soil and Culture Medium After 120 d of growth, Nto,al (Kjeldahl), exchangeable+, and P (P205Olsen) were determinedin both bare and planted fertilized soils accordingto standard methods.In hydroponic cultures, NO.~,NH2,and K+ were analyzed at 0, 5, 10, 15, 20, 30, and 40 d in each flask by capillary electrophoresis (Waters Capillary Ion Analyzer). RESULTS
Hydrocarbons For soil experiments, the methodfor HCanalyses (Cha~neauet al., 1995, 1996) consisted of drying the soil samples at 60°C for 12 h and performing Soxhlet extraction of the entire sample, including roots, with chloroformfor 8 h. In liquid cultures, the residual HCwere recoveredby chloroform liquid-liquid extraction, the roots being extracted together with the medium.The chloroformextract (CE) contained total residual HCand part of the biogeniclipids. In both cases, the residues were weighedafter evaporation of the solvent and separatedin saturated, aromaticand polar fractions by successive elution with 60 mLeach of hexane,benzene,and methanol on a 15- by 1-cmchromatographiccolumnfilled with 100- to 200-mesh activated silica gel (120°Cfor 12h). The polar fraction containedthe resins of the oil and somepolar lipid compoundsof the soil or culture medium.After gravimetricdeterminationof the fractions of the three replicates, the saturated and aromatic fractions of the sample closest to the meanof the three replicates wereanalyzed by computerizedgas chromatography(GC) with a Delsi DI 200 chromatographfitted with a direct injection port and a FIDdetector, both set at 340°C;carrier gas was Heunder 0.8 bar; columnwas a CPSil 5 CB(Chrompack)capillary column (50 m × 0.32 mm,film thickness 0.25 ~m); temperature programmingwas 100 to 320°C, 3°C min-1. Acquisition and numerical treatments of data were performed using customized computer programs (Oudot, 1984). Microbiological Counts The numberof colony-formingunits (CFU)of total heterotrophic bacteria (THB)and hydrocarbon-adapted bacteria (HAB)were determined in the soils and in the hydroponic mediumaccording to the most probable number (MPN) method,using three tubes per dilution. The methodpreviously described(Cha/neauet al., 1995)and used by different authors (Nichols et al., 1997, Wrennand Venosa, 1996) consisted the inoculation of appropriate media(trypcase-soy 30 g -1 for THB,mineral mediumfor HAB),with 10-fold dilutions of soil or culture medium.Incubation was run for 3 d for THBand 21 d for HABat 24-+1°C. The number of viable microorganismswas obtained from standard MacCradytables after examinationof growthpositive tubes.
HCWeathering in the Soil-Plant
System
Maize Growth The soil was entirely colonized by roots within 20 d, and was considered as a rhizospheric soil. Maize growth was exponential from Day 20 to Day 80. The plants reached maximumdevelopment after 80 d of cultivation, the final height of all plants was around 135 cm. The height and dry aerial biomass of plants were slightly lower in the presence of HCthan in the control (Table 1). At final harvest (120 d), the plant biomasswas equivalent with and without HC, possibly owingto limitations in the development of the control plants due to the small-scale design of the experimental system (see Discussion). Biodegradation of Hydrocarbons As measured in sterile controls, the evaporation (18%) of light compoundsreduced the initial concentration of fuel oil from 3300 to 2720 mg HCkg-I soil (Fig. 2). Evaporation was due to volatilization at room temperature during incubation and to the drying of the samples before extraction. All HC compounds below nC14 were lost by evaporation, as occurs naturally in the field in temperate environments in a few days even without drying (Chaineau, 1995). In the poisoned soils, the concentration of the CEremained constant, indicating that no more abiotic loss of HCoccurred during the experiment. In the contaminated and unfertilized bare soil, the concentration of the CEdecreased with time from 2720 to 2270 mgkg-I soil. The addition of mineral nutrients strongly increased the degradation of HC: After 120 d, the residual mean CE concentration was 1220 mg kg-~ soil. Hence biodegradation of HCwas three times higher in fertilized than in unfertilized soil (55 and 16.5%, respectively). The presence of maize fertilized soils increased the biodegradation of HCfrom Day 20 to Day 80. The statistical difference between bare and vegetated soils was highly significant (P
Table 1. Growthof maize plants in the presence and absence of fuel oil HCin soil (mean _+ SD). Time, d 10
20
40
80
120
Height, cm/plant Unpolloted Polluted Unpolluted Polluted
37.9 ± 3.3 28.4 -+ 2.2 0.2 ± 0.05 0.1 -+ 0.01
69.4 -+ 6.1 54.1 ± 4.9 0.7 ± 0.15 0.3 ± 0.08
76.3 ± 5.5 75.8 _+ 6.8 Shoot, g dry wt/plant 3.7 -+ 0.54 2.6 -+ 0.62
134.9 -+ 2.6 123.4 -+ 2.2 19.5 ± 1.20 17.7 ± 1.53
133 _+ 3 123.5 ± 1 19.5 ± 1.30 19.5 -+ 1.26
572
J. ENVIRON.QUAL., VOL29, MARCH-APRIL 2000 3000 2500"
A
2000 1500 1000 500’
o
2’0 ;o "0’o .’o i~oi~o
2000’ 1500~ 1000 500 .
0 2’0
4’0
6’0
~. . . 80 1 (~0 120
1000 800
~
13
600 4001 20O 0 ’ " 0 2’0
" 4’0
6’0
80
1(~0 120
1000 900 800 700 6OO 500 400 300 200
Fig. 2. soil; soil; ized
20
40
60
80
100 1~0
TIME(days) Hydrocarbonconcentration in soil microcosms. ([~), Sterile (I), oil-treated nonfertilized soil; (O), oil-treated fertilized (O), oil-treated fertilized andplantedsoil; (A), nonoiledfertilplanted soil. Bars represent -+1 SD.
0.01) at Day40. It reached 600 mgkg-1 soil at Day40, i.e., removalof HCwas 20%faster in the presence of plants than in the nonvegetatedsoils. After 80 d, the difference wasno longer significant. Thefinal biodegradationrates for total fuel oil wereidentical in all fertilized soils, whetherplanted or not. The degradation of saturated HCwas activated by addition of mineral nutrients (Fig. 2). In the absence
Fig. 3. Gas chromatographicanalyses of saturated HCof fuel oil in soil microcosms. The numbers represent the carbon number of n-alkanes. No difference was observed between planted and bare soil after 120d. (A), initial fuel oil; (B), nonfertilized soil after d; (C), fertilized soil after 120 d. I.S, internal standard(n-eicosene); UCM,unresolved complex mixture.
CHA~NEAUET AL.:
573
BIODEGRADATIONOF FUEL OIL HYDROCARBONS IN MAIZE
of fertilization, the biodegradation reached 43%. GC analyses showed that linear, branched, and cycloalkanes were only partially biodegraded (Fig. 3). In fertilized soils, linear and branched alkanes were totally removedin 120 d (Fig. 3, Table 2). In rhizospheric soils, the rates of biodegradation were similar to fertilized bare soils during the first 20 d. From 20 to 80 d, the biodegradation of saturates was higher in the presence of plant roots (A = 250 mg kg-I, P < 0.01 at Day 40). GCanalyses showed a higher extent of biodegradation of branched alkanes and naphthenes in the root zone (Table 2). The final saturates biodegradation rate 120 d was 75%in fertilized soils in the presence and absence of plants. The general weathering of aromatic HCwas similar to that of saturated HC(Fig. 2). At Day 0, the aromatic HCsoil content was 900 mgHCkg-1 soil. The fertilization stimulated the assimilation of aromatics, which reached 70%in the fertilized soils and 44%in the nonfertilized soils. Maizeincreased the biodegradation rate from Day 20 to Day 80 (A = 110 mg kg-~, P < 0.01 at Day 40). From Day 80 to Day 120, the degradation of aromatics slowed down and ANOVAshowed no significant difference between rhizospheric and bare soils. At Day 120, 70% of aromatic HCwere biodegraded in all fertilized soils, whether planted or not. A slight increase in the polar fraction was recorded in the nonfertilized bare soil. The root development caused a significant increase of this fraction, probably due to the production of polar metabolites (Fig. 2). significant degradation of the resins was observed in any case.
~ ~
1012¯ 1011¯ 1010. 1°9
~ 10 107 ~ 10 1° ~ 104 0
20
40
60
80
100 120
TIME (days) Fig. 4. Bacterial populations (CFUg-1 soil) in the soil microcosms. Oil-polluted soils: total heterotrophic bacteria (THB)in the rhizosphere (O) and in the bare soil (©); hydrocarbon-adapted bacteria (HAB)in the rhizosphere (I) and in the bare soil ([~). Unpolluted soils: (A) THBand (K) HABin the rhizosphere.
Assimilation of Nutrients After 120 d, total N and exchangeable K+ concentrations in unplanted contaminated soils showed a slight decrease (Table 3). By contrast, in the presence plants, the remaining concentration of total N and exchangeable K+ in soil was very low, indicating that most of the supplied nutrients had been assimilated by the growing plants. However, in the unpolluted planted microcosms, the plants did not use all of the N supplied, whereas all K added was assimilated. The uptake of P was slightly different in the presence of maize in the polluted and unpolluted soils. The concentration of P wasstill high at 120 d in soils. Hydroponic Cultures
Microbial Counts
Maize Growth
The number of THBand HABin Hg-treated soil was always lower than 10 CFUg-i soil. In the unpolluted soil, fertilization and plant growth did not modify the number of THB, but the number of HABincreased 100 times in the first :20 d (Fig. 4). Then, both THBand HAB remained constant until the end of the experiment. The addition of fuel oil sharply increased the number of THBand HABin the first :20 d of incubation, especially in the presence of plants. The populations reached a maximumbetween :20 and 40 d, then decreased regularly. At Day 120, both THBand HABin polluted bare soils had returned to levels similar to the initial soil. At the same time, in the rhizospheric soils, microbial populations were 10 times higher than in bare soils.
Maize plants were grownfor 40 d in nutrient solution amended or not with fuel oil and inoculated with a complexsoil microflora (Table 4). Plant aerial and root biomass were not significantly affected by the presence of HC. Biodegradation of Hydrocarbons After evaporation of the light fraction of the fuel (21%), the mass of HCin sterile conditions remained Table 3. Total N, K, and P content (g kg-1) in (A) polluted bare soils, (B) vegetated polluted soils, and (C) vegetated unpolluted soils. Treatments
Table 2. Biodegradation rates of n-alkanes, branched alkanes, and cyclic HC(naphthenes and aromatic HC) in fertilized soils (rhizosphere and bare soil) as determined by GC. Baresoil
Rhizosphere
Before fertilization
% 86 92 94 97 100
53 62 68 87 89
2 18 33 42 55
78 94 97 98 100
49 70 85 91 92
18 33 48 53 65
120 d
Total N A B C
1.6 1.6 1.6
A B C
0.15 0.15 0.15
B C
0.06 0.06
2.6 2.6 2.6
2.4 1.6 1.8
0.55 0.55 0.55
0.46 0.3 0.08
1.08 1.08
0.28 0.36
K
Timen-alkanes iso-alkanes Cyclic HCn-alkanes iso-alkanes Cyclic HC d 10 20 40 80 120
After fertilization
P
574
J. ENVIRON.QUAL., VOL. 29, MARCH-APRIL 2000
Table 4. Growth of maize plants in hydroponic cultures in the presence and absence of fuel oil HC.
140’ 120
Time, d 10 Polluted Unpolluted
47 66
Polluted Unpolluted
0.18 0.19
Polluted Unpolluted
20
30
100-
40
80
Height, cm/plant ± 5 72 +- 3.2 79 -+ 3.1 83 ± 3.1 ± 2.5 71 +_ 3.5 68 ± 7.1 78 ± 3.5 Root, g dry wtJplant ± 0.05 0.80 ± 0.05 1.18 _ 0.06 1.38 ± 0.16 ± 0.02 0.71 _+ 0.13 1.23 _+ 0.22 1.43 ± 0.24 Shoot, g dry wtJplant
60 40 20 0
0.23 ± 0.01 1.64 ± 0.1 2.42 _+ 0.13 3.42 ± 0.23 0.32 ± 0.08 1.59 ± 0.13 2.23 _+ 0.26 3.71 ± 0.41
1’0
2’0
3’0
4’0
1’0
2’0
3’0
4’0
0
1 ’0
2’0
3’0
4 0
0
1’0
2’0
3’0
4’0
70"
!6o
constant (Fig. 5), thus indicating that no abiotic losses occurred during the experiment. In the presence of microorganisms, the concentration of HCdecreased steadily during the 40-d growth period. Whenplants were present, degradation of HCwas faster than in controls. After 40 d, total HC were more degraded (P < 0.01) in the presence of maize roots (80% with maize, 64% in the controls without plants). The saturated HCwere more degraded (P < 0.01) in presence of maize (84% in the rhizosphere, 67%in the control). No significant difference was recorded in the aromatic fraction (80% in both cases). The polar fraction concentration decreased during the first 10 d, indicating a microbial assimilation, then increased due to the accumulation of metabolic by-products in the mediumwithout plants. By contrast, no change was detected after 10 d in the presence of plants.
40" 30" 20" 10"
40" 30~ 20’ 10"
Microbial Counts In cultures without plants, the mediumremained clear during the whole experiment. However, microbial population was high (Table 5). In the presence of plants, the medium was turbid. Maize roots induced a 600fold increase in THBand a 160-fold increase in HAB populations. The stimulation of bacterial populations was muchhigher and faster in hydroponic cultures than in the soil experiments.
0 3O
[’-’
20
~ lO Assimilation of Nutrients Concentration of mineral nutrients (N, K) in solution was monitored during the 40-d growth period (Fig. 6). A general decrease in concentrations was observed, but removal of mineral ions from the solution was much higher in the rhizosphere. Without plants, 33% of NH~-, 22% of NO~-, and 23% of K+ were removed by microbial assimilation. In the presence of plants, the concentrations decreased sharply during the first 10 d. From Day 20 to Day 30, almost all N and K was removed from the medium. DISCUSSION This work showed that at HCconcentrations compatible with plant growth, the rhizosphere of maize had an influence on biodegradation rates. Previous experi-
TIME (days) Fig. 5. Weatheringoftotaloilaudfractiousiuthehydtoponi¢cultures. ([~), Sterile medium;cultures (0) with and without plants.
ments showed that when applied at low rates to agricultural soils, HCwere assimilated by the indigenous microbial population (Chaineau et al., 1995, Oudot et al., 1989), even if cover crops were cultivated (Chalneau al., 1996). As already observed, HCbiodegradation was strongly stimulated by the addition of nutrients (Atlas, 1991; Bossert and Bartha, 1984): 16.5% biodegradation (total fuel HC) occurred in the control soil, and 55% in soils fertilized with mineral nutrients. GCanalyses
CHA~NEAUET AL.:
Table 5. Changes in bacterial populations in the culture medium in the presence and absence of maize plants after 40 d.’~ Bacteria THB -~) (CFU mL HAB(CFU niL 1)
A 25 × 106 9.5 × 106
575
BIODEGRADATIONOF FUEL OIL HYDROCARBONS IN MAIZE
B
C
15 × 109 1.5 × 109
25 × 108 1.5 × 108
2oo| 175] 150~
"~ A, oil only; B, oil + plant; C, plantonly.
were typical of HCthat have undergone microbial degradation: n- and branched alkanes, as well as GC-resolved aromatic, were the most assimilated HC. Part of the cyclo-alkanes, and of the aromatics as well as the resins, were almost refractory to microbial biodegradation; they constitute a stable form of organic matter (Chaineau et al., 1995; Oudot, 1984; Oudot et al., 1998; Phelps et al., 1994). At the end of the experiments, the biodegradation rates of the saturated and aromatic fractions were in the order of 70 to 80% in the microcosms and in the hydroponic cultures. These rates are close to the maximumvalues that can be reached by microbial degradation of fuel oil (Cha~neauet al., 1995, 1996); no significant further decrease of HCis to be expected from biological processes, even on longer time scales. Biodegradation
in the Rhizosphere
Someworks have investigated the biodegradation of PAHin the rhizosphere (Aprill and Sims, 1990; Gtinther et al., 1996; Qiu et al., t994; Reilley et al., 1996). Very few studies have been undertaken to quantify the specific effect of plants on fuel oil hydrocarbon removal. This work demonstrated that dissipation of HC was accelerated in the presence of living plant roots. Direct uptake of HCby the plant itself did not appear to be a significant process in the case of maize (Chaineau et al., 1996, 1997). A 6 to 19%increase in early stages of biodegradation of petrogenic HCwas recorded in the rhizosphere as compared with unplanted soil. Saturated and aromatic HCwere more rapidly degraded in the presence of plants. The difference between control and planted soils appeared as soon as plant growth was exponential, allowing for the developmentof the root system that caused a large soil-root contact. These results are in concordance with the observed biodegradation of organic pollutants such as PAH,alkanes, pesticides, and chloroorganic compoundsin the rhizosphere (Anderson et al., 1993; Andersonet al., 1994; Aprill and Sims, 1990; Boyle and Shann, 1998; Gtinther et al., t996; Qiu et al., 1994; Reilley et al., 1996; Shimpet al. 1993; Wiltse et al., 1998). In our study, when plant growth ceased, no more significant differences were observed between planted and bare soil, indicating that the influence of plants on HCdegradation rates was limited in time. Influence
of Root Exudation
Maize was used as a plant model that produces exudates that are composedof various substances, ranging from low-molecular-weight molecules (sugars, proteins, organic acids, and uronic acids) to high-molecularweight mucilages (Morel et al., 1991). To assess the contribution of exudates to biodegradation, maize
50 25
o 0
10
20
30
40
8O ,.--, 6O
Z
20
0
10
20
30
40
10
20
30
40
80-
40" 20"
TIME(days) Fig. 6. Nutrient concentration in hydroponiccultures. ([]), Sterile medium; (©) microorganisms only and (O) microorganisms + plants.
plants were grown under hydroponic conditions, thus avoiding interference with soil components, i.e., clay minerals and organic matter. Biodegradation of saturates increased (20%) in the presence of maize. difference was recorded for the biodegradation of aromatic HCin hydroponic cultures, whereas soil rhizosphere caused an increase in the biodegradation of this fraction. No decrease of the polar fraction was measured in the soil experiments, thus confirming that resins are highly resistant to biodegradation (Chaineauet al., 1995, 1996; Oudot, 1984; Oudotet al., 1998). However,a slight but temporary decrease of the polar fraction was noted in the hydroponic cultures, as already observed in pure cultures of bacteria and fungi on crude oil (Oudot et al., 1993). Exudates can act as promoters for biodegradation by providing alternative carbon sources or growth factors that stimulate the development of bacterial populations (Benizri et al., 1995). They mayalso have an influence on the bioavailability of HCdue to their surfactant properties (Banat, 1995; Simonich and
576
J.
ENVIRON. QUAL., VOL. 29,
Hites, 1995). This could be beneficial for aged contaminants, which are more firmly bound to the organo-mineral matrix of the soil (Providenti et al., 1993). Microbiological
Aspects
A strong increase in bacterial counts is usually observed after HCcontamination in field experiments (Atlas, 1991, Cha~neauet al., 1996) and in laboratory conditions (Cha~neau et al., 1995). In this study, the response of the populations to HCcontamination was quite immediate, without a time lag associated with microbial adaptation or acclimatation in both planted and nonvegetated soil microcosms (Cha~neau et al., 1995). From 40 to 80 d, when maize growth was exponential, total bacterial counts were about 10 times higher in the rhizosphere than in the root-free systems for THBand HAB,i.e., rhizospheric stimulation of bacterial populations was limited in the soil. The R/S (rhizospheric microorganisms/nonrhizospheric microorganisms) ratio ranged from 4 to 55 in the soil. These values are in agreement with what has already been observed in soils (Andersonet al., 1993; Gtinther et al., 1996; Nichols et al., 1997). R/S varied between 100 and 600 in hydroponic cultures, which indicates that the stimulation was much more important in the liquid cultures. It is knownthat plant roots have a direct effect on microbial populations due to the local environmental conditions they create. Root hairs are also used as physical support for the microbial attachment of nonmotile microorganisms. This may explain the higher counts observed in the hydroponic cultures, where this effect is much more marked than in the soil. Exudates may have played a role as complementary substrate for the microbiota, allowing maintenance of higher bacterial counts in the rhizosphere (Benizri et al., 1995; Maryet al., 1993). Nutrient Assimilation Microorganisms and plants need nutrients for the buildup of their cellular biomass. A competition for nutrient uptake can occur in the case of planted contaminated soils (Xu and Johnson, 1997). In the field using standard agricultural practices, one maize plant is usually fertilized with an amount of 500 mg N per plant (140 kg N ha 1, 10 plants m-Z). The resulting aerial biomass produced is over 200 g per plant, i.e., the ratio between N applied and plant biomass is currently lower than 0.2% w/w. In the soil experiment, maize plants were supplied with 200 mgN per plant, for a resulting aerial biomass of 19 g. The N/biomass ratio was 0.8%, i.e., more than four times higher. This indicates that the growth of maize was certainly not significantly affected by nutrient shortening, but rather by the physical limitations due to the small size of the containers that restricted root development. Similarly, in the hydroponic cultures the N/biomass ratio was 0.7 % w/w. If phytoremediation with cultivated plants is to be used, the N, P, and K fertilization should be adjusted to adequate levels to compensate for the needs of plant nutrition. These results are confirmed in hydroponic cultures (Fig. 6) in which Cdegraded/Nassimilated and Cdegraded/
MARCH-APRIL2000
Kassimilated ratios were respectively 5.6 and 17.7, thus giving a C/N/Kproportion of 100/17/6 in cultures without plants. The favorable nutrient conditions for HC bioremediation are often expressed as C/N/P ratios. The recommendedvalues range from 100/10/1 to 100/50/10 (Bossert and Bartha, 1984; Mills and Frankenberger, 1994; Morgan and Watkinson, 1989). Our results are in agreement with these findings but also show that the role of K has certainly been underestimated. The nutrient analyses of the mineral mediumin hydroponic cultures pointed out that nitrate, ammonium, and potassium were almost completely assimilated by maize after 20 d. However, HCbiodegradation (Fig. 5) and plant growth (Table 4) did not decrease after that time, even though the culture mediumwas deficient in mineral forms of N and K. No plateau was observed in the plant growth. It is likely that the turnover of N and K under organic forms due to microbial activity may have been fast enough to ensure further biodegradation and plant development. From Day 20 to Day 40, the microbial populations were totally developed and did not need any supplementary amount of N and K, since the biomass necessary for biodegradation activity was already elaborated before Day 20. Similarly, the plants had stocked in their tissues a sufficient amountof nutrients during the first 20 d for complete developmentuntil Day 40. Other Effects Besides the positive effects of exudation on microbial communities, plant development has other effects on local environmental soil parameters. Living plant roots cause significant changes in the physical, chemical, and biological properties of the soil by their action on moisture conditions, pH, water content, temperature, oxygen tension, and soil adsorptivity (Gtinther et al., 1996). They also exert important effects on the soil structure through mechanical action, decompaction, and perforation of aggregates, exposing the organic matter to decomposition by microorganisms (Morel et al., 1991). our experimental design, the physical effects of rhizosphere on HCbiodegradation were limited, since all parameters were kept near optimal values even in the absence of plants. In the field, the effect of plant roots on HCremoval could be more marked, especially where physical environmental conditions are unfavorable for biodegradation (Aprill and Sims, 1990; Giinther et al., 1996). Our results suggest that indirect enhancement of HC biodegradation in the root zone may be caused by an increase of HCbioavailability and by the stimulation of microbial populations through plant exudation. In the field, the positive physical effects of plants on soil environmental conditions may also accelerate the rates of biodegradation. However,the effects of plants on HC biodegradation must not be overestimated. Weshowed that the positive effects were statistically significant during the early stages of biodegradation, whereas the final biodegradation stage was identical in planted and bare soils. Part of the cyclic HCand all of the resins were
CHAINEAU ET AL.: BIODEGRADATION OF FUEL OIL HYDROCARBONS IN MAIZE
not degraded, even in the rhizosphere, and persisted in the soil as stable organic matter. The combination of an adequate N/P/K fertilization and the selection of highly tolerant plant species could have overall beneficial effects on HC biodegradation and might also improve the stabilization of contaminated soils. It should be pointed out that these results were obtained on slightly contaminated soils. Maize has been shown to tolerate up to 1.2% as a maximum HC concentration compatible with reasonably good plant growth (Chafneau et al., 1996,1997). In the case of higher inputs, as initially occurs in accidental spillage, the phytotoxicity of HC would not allow correct plant development until HC content and toxicity were reduced by microbial activity or physicochemical processes to levels acceptable for plants. ACKNOWLEDGMENTS The assistance of Ms. Severine Putas is gratefully acknowledged. Part of this work was supported by Traitement Valorisation Decontamination.
577
578
J. ENVIRON. QUAL., VOL. 29, MARCH-APRIL 2000
