Soil samples were taken in Mendoza, Argentina, from a deposit of soils previously treated by landfarming for 1 year and kept in outdoor piles during 10 years under extreme climatic conditions (i.e., dry arid climate, average rainfall below 200 mm, with almost 3,000 h of sunshine throughout the year, summer average temperature of 28 and 7°C in winter, with daily temperature fluctuations of +12 and +42°C in summer and 2 and 15°C in winter). The soil was air dried and sieved with a 6 mm grid previous to microcosm experiments and 2 mm grid for characterization analysis. Soil samples were stored aerobically at 4°C in the dark until use.

The soil samples were characterized according to the following standard physicochemical parameters and methodologies (APHA et al., 2005): texture; pH; electric conductivity; total nitrogen; available phosphorous and moisture. Elemental analysis of C (carbon), H (hydrogen), N (nitrogen), and S (sulfur) was performed with a LECO Truspec CHNS (LECO Corporation, United States). Based on the chemical characteristics of the soil hydrocarbon contamination, the TNRCC Method 1005 (Texas Natural Resource Commission, 2001) was used for extraction and quantitative analysis of Total Petroleum Hydrocarbons (TPH), as well as for the evaluation of the relative distribution of each TPH in the sample extracts. This method used n-pentane (chromatographic grade, Sintorgan, Buenos Aires, Argentina) as solvent, which was then analyzed by gas chromatography/flame ionization detection (GC/FID) for identifying and quantifying hydrocarbons between nC6 and nC35 by comparing chromatographic profiles with those from reference n-alkane references (C8-C40 standard kit, AccuStandard Inc., New Haven, CT, United States). PAHs were identified and quantified by gas chromatography/mass spectrometry (GC/MS) under the selected ion monitoring mode (SIM), by measuring m/z area signal of 16 EPA priority PAHs according to the EPA Method 8270D. Before the n-pentane extraction, 20 μg g^–1 of o-terphenyl and α-androstene (AccuStandar Inc., New Haven, CT, United States) were added in an acetone solution (1 mg mL^–1) to each sample as surrogate internal standards. Acetone was allowed to evaporate and 10 g of soil were suspended in 10 mL of n-pentane in a sealed screw PTFE cap glass vial, shaken by vortex during 5 min and allowed to settle overnight. Thereafter, 2 mL of sample extracts were transferred to auto sampler vials for further chromatographic analysis.

The biodegradation of saturated compounds (alkanes) was verified using a Clarus 500 Perkin Elmer GC/FID. Compounds were separated on a HP-5 capillary column [25 m by 0.32 mm (i.d.), 0.25-μm film thickness (Hewlett-Packard)]. Column temperature was held at 42°C for 5 min and then programmed to reach 300°C at a rate of 10°C min^–1. This final temperature was held for 80 min. Detector and inlet temperatures were set at 320 and 300°C, respectively. The helium flow was 1.1 mL min^–1 and the injection volume was 1 μL. Concentration ranges of aliphatic hydrocarbons were calculated from the total areas of the peaks in the chromatograms corrected with internal standards areas and standards of aliphatic hydrocarbons (C12-C40 range) of known concentration.

A HP-5 capillary column [30 m by 0.32 mm (i.d.) Hewlett-Packard] with 0.25-μm film thickness and helium as a carrier gas (10 psi) were used. Column temperature was held at 50°C for 5 min and then programmed to 250°C at a rate of 10°C min^–1. This final temperature was held for 35 min. Injector, transfer line and analyzer temperatures were set at 320°C. Injection was in splitless mode keeping the split valve closed for 30 s. The targets for this analysis were the 16 priority EPA PAHs. The obtained reconstructed ion chromatograms were numerically compared with internal standards (16 EPA PAHs standard, AccuStandar Inc., New Haven, CT, United States) for the biodegradation estimation of each analyte in soil samples.

The total TPH concentrations estimated according to the EPA Method 418.1, referred hereafter as TPH-IR, were considered as kinetic biodegradation parameter.

Dry soil (10 g) was added to 100 mL of a sterile NaCl 0.09% (w/v) solution in a 200 mL Erlenmeyer flask and shaken for 20 min on a shaker. After solid particles were allowed to settle for 30 min, serial dilutions were prepared and 100 μL of each dilution were spread homogeneously on the surface of three replicate malt yeast extract agar plates (MYEA: 2% malt extract, 0.2% yeast extract, 1.5% agar, pH 7.0) and on mineral agar amended with phenantrene, as standard PAH, (20 mg PHE crystals L^–1, PHEN medium) as the sole carbon and energy source [PHEN composition per L: 4.5 g KH₂PO₄; 0.5 g K₂HPO₄; 2.0 g NH₄C; 0.1 mg MgSO⋅7H₂O; 120 mg FeCl₃; 50 mg H₃BO₃; 10 mg CuSO₄⋅5H₂O; 10 mg KI; 45 mg MnSO₄⋅H₂O; 20 mg Na₂MoO₄⋅H₂O; 75 mg ZnSO₄⋅H₂O; 50 mg CoCl₂⋅6H₂O; 20 mg AlK(SO₄)₂⋅12H₂O; 13.25 g CaCl₂⋅H₂O; 10 g NaCl]. PHEN agar was also supplemented with a solution of sterilized chloramphenicol (0.02%) and was autoclaved at 120°C for 20 min before dispensing into Petri dishes. PHEN was selected according to previous studies being the most common PAH source to distinguish viable PAH-degrading bacteria and fungi on agar plates (Viñas et al., 2005b; Hesham et al., 2016; Agrawal et al., 2018) and in base of previous analyses of the industrial soil where it was one of the PAH with the highest concentration and prevalence.

Inoculated agar plates were incubated at 25°C during 15 days (MYEA) and 21 days (PHEN) days sheltered from light. The observation of morphological characteristics of several colonies (hyphal tips, mycelium shape, color, surface, elevation, form, and grown velocity) that appeared to belong to different representative isolates were transferred to new MYEA plates.

Colonies were subcultured in MYEA until pure cultures were obtained. Twelve strains were morphologically described as being different on MYEA slants and only one of them was able to grow on PHEN agar. Strains grown only on MYEA agar were labeled correlatively from A1 to A11, and B1 corresponded to the one growing on PHEN agar. All isolates were maintained on MYEA slants incubated at 20°C.

The inoculum consisted of a suspension of mycelia from a defined fungal consortium of twelve selected fungal isolates, prepared as described by Potin et al. (2004). Every one-week-old MYEA and PHEN Petri dish cultures of the twelve isolates was washed with 4 mL of sterile deionized water. Mycelium fragments were removed from the spore suspension by filtration through sterile glass wool. 100 μL of every spore suspension were mixed to form a single one which was estimated using a Thoma chamber. Before inoculation into de soil, the spore suspension was mixed with a solution of glucose and sucrose, each at 5 mg g^–1 soil, as carbon source and kept in agitation for 36 h at 25°C in order to induce spore germination and hyphal elongation. The germinated spores were microscopically controlled and added to the soil in calculated volumes to give an equal contribution of each strain (1:12) and a final total spore concentration of 10⁴ spores g soil^–1. None lignocellulosic amendment was added to the soil.

All the microcosm treatments were conducted in triplicate with 1,200 g of 6 mm sieved soil inside 5 L glass recipients with plastic lids and was incubated for 120 days at room temperature. Three different microcosm treatments were designed: (i) bioaugmentation (B), consisting of soil inoculated with fungi up to an initial concentration of 10⁴ spores per gram of soil. Soil moisture was maintained regularly at 20% w/w by gravimetry and K₂NO₃ and K₂HPO₄ were added on days 30 and 90 to prevent the C:N:P ratio descended below 100:10:1; (ii) biostimulation (BS), in which soil moisture was maintained regularly at 20% w/w by gravimetry, and K₂NO₃ and K₂HPO₄ were added on days 30 and 90 to prevent the C:N:P ratio descend below 100:10:1, the optimal for bacterial biostimulation (Bossert and Bartha, 1984; Morgan and Watkinson, 1989; Mills and Frankenberger, 1994; Chaineau et al., 2005); and (iii) a control (C) formed by air-dried untreated contaminated soil (13% humidity), with the aim to evaluate abiotic losses of hydrocarbons in the absence of enough water content. In each microcosm, air renewal occurred weekly through manual removal by using cleaned and sterilized stainless-steel spatulas under sterile flow cabinet. As microcosm B includes biostimulation by the addition of nutrients plus bioaugmentation, BS also represented a control treatment for B. Soil samples from each microcosm (120 g) were collected manually with clean and sterilized (ethanol 70%) stainless steel spatulas on days 0, 30, 60, and 120 for microbiological and physicochemical analysis.

Acute hydrocarbon toxicity in soil microcosms was assessed on soil samples collected on days 0, 30, 60, and 120 by quantifying the EC50 with the Microtox^TM Basic Solid-Phase Test (Microtox BSPT, AZUR Environmental, CA, United States). This Vibrio fischeri based bioluminiscence inhibition assay has been previously described as a versatile and suitable test to detect polar compounds that could be accumulated in soil during biodegradation processes (Parvez et al., 2006; Sabaté et al., 2006). Soil samples were diluted at an initial and final concentration of 200 and 0.781 g soil L^–1, respectively, and were osmotically corrected. Cells were resuspended with Microtox resuspension agent and analyzed spectrophotometrically according to the instructions of the manufacturer. Statistical regressions between Microtox readings and TPH concentrations were obtained by log-log plots using the Statistical software (IBM, United States). The resulting EC50 values were defined as the antilog of the regression line‘s intersection point.

Soil microbial counts were performed using a miniaturized most probable number (MPN) method in 96-well microtiter plates, with eight replicate wells per dilution (Wrenn and Venosa, 1996). Total heterotrophs were counted in tryptone soy broth and aromatic hydrocarbon-degraders were counted in liquid mineral medium (BMTM) containing a mixture of phenanthrene (0.5 g L^–1), with fluorene, anthracene, and dibenzothiophene (each at a final concentration of 0.05 g L^–1) according to Viñas et al. (2005b).

The PAHs were administered in a pentane solution that was allowed to evaporate in each well, where the BMTM mineral medium was subsequently added. Aged soil was used as the starting point (day 0). MPN plates were incubated at room temperature (25 ± 2°C) for 30 days. Positive grown wells were detected by turbidity (heterotrophs) and by the presence of coloration (brownish/yellow) for PAH degraders. Microbial population in soil microcosms were quantified on days 0, and after 30, 60, and 120 days of incubation. Since no antifungal agent was added to the medium, the total heterotrophs quantification includes both bacteria and fungi.

Pure strains of previously isolated fungi were cultivated on mineral agar amended with ABTS [2,2′-azine-bis (3-etilbenztiazoline-6-sulfonic acid)], tannic acid, and Red Orange (RO) to screen their laccase, phenol oxidase, and manganese peroxidase secretion. These chromogenic agar methods were characterized by developing a pigmentation in the oxidized state that was green for ABTS, dark orange for tannic acid, and an orange red halo for RO. The culture mineral medium was prepared as described by Prenafeta-Boldú et al. (2001). Per liter of dematerialized water: 4.5 g KH₂PO₄; 0.5 g K₂HPO₄; 2.0 g NH₄C; 0.1 g MgSO₄⋅7H₂0; 120 mg FeCl₃; 50 mg H₃B0₃; 10 mg CuSO₄⋅5H₂O; 10 mg KI; 45 mg MnSO₄⋅H₂O; 20 mg Na₂MoO₄⋅H₂O; 75 mg ZnSO₄⋅H₂O; 50 mg CoCl₂⋅6H₂O; 20 mg AlK(SO₄)⋅12H₂O; 13.25 g CaCl₂⋅H₂O; 10 g NaCl; 2 g Glucose; 100 mg nicotinic acid; 200 mg calcium pantothenate; 25 mg cyanocobalamin; 100 mg inositol; 20 mg p-amino benzoate; 50 mg thiamin⋅HCl; 25 mg pyridoxin⋅HCl; 10 mg biotin; 10 mg riboflavin; 10 mg folic acid;10 mg thioctic acid and 16 g agar). This medium was supplemented with ABTS (500 mg), tannic acid or orange red chromogenic substrate (100 mg each) according to Pointing (1999). A collection strain of Trametes versicolor (ATCC 20869^TM ) was used as a positive control for ligninolytic activity (laccase, Mn peroxidase and polyphenol oxidase).

A 250 mg sample of biomass from each fungal isolate and of soil from each microcosm was placed in a sterile tube and stored at −20°C prior to analysis. DNA was extracted by a bead beating protocol using the PowerSoil^TM DNA extraction kit (MoBio Laboratories, Inc., Carlsbad, CA, United States), following the manufacturer’s instructions. A further purification of soil DNA extracts was needed using the Clean DNA Wizard Kit (Promega, WI, United States) to avoid PCR inhibition.

Total genomic DNA from the original soil and from microcosm test samples were characterized by PCR-DGGE (Polymerase Chain Reaction and Denaturing Gradient Gel Electrophoresis) using specific primers for amplifying the hypervariable V3–V5 region of the bacterial 16S rRNA gene (F341GC/R907: Yu and Morrison, 2004). For the identification of isolated fungal strains, the ITS1-ITS2 rRNA region was amplified, which also contains the 5.8S coding region (ITS1F-ITS4R Esteve-Zarzoso et al., 1999). All the PCR reactions were performed with a Mastercycler (Eppendorff, Hamburg, Germany) and each reaction mix (25 μL mix/reaction) contained 1.25 U of Ex Taq DNA polymerase (Takara Bio, Otsu, Shiga, Japan), 12.5 mM dNTPs, 0.25 μM of each primer and 100 ng of DNA. The obtained amplicons from the 16S rRNA were loaded in an 8% (w/v) polyacrylamide gel with a chemical denaturing gradient ranging from 30 to 70% [100% denaturant contains 7 M urea and 40% formamide (w/v)], and electrophoretically resolved in a DGGE-4001 equipment (CBS Scientific Company, Del Mar, CA, United States). The electrophoresis was carried out at 60°C and at 100 V for 16 h in a 1X TAE buffer solution (40 mM Tris, 20 mM sodium acetate, 1 mM EDTA, pH 7.4).

The DGGE gel was stained for 45 min in 1 × TAE buffer solution containing SybrGold^TM (Molecular Probes, Inc., Eugene, OR, United States) and then scanned under blue light by means of a blue converter plate (UV Products Ltd., Cambridge, United Kingdom). Predominant bacterial DGGE bands were excised with a sterile filter tip, resuspended in 50 μL sterilized Milli-Q water and stored at 4°C overnight. A 1:50 dilution of the supernatant was subsequently reamplified by PCR as described previously and sequenced by using the R907 primer.

Sanger sequencing of amplicons of isolated fungal strains were performed with the ITS4 primer. Sequencing was accomplished using the ABI Prism Big Dye Terminator Cycle-Sequencing Reaction Kit v. 3.1 (Perkin–Elmer Applied Biosystems, Waltham, MA, United States) and an ABI 3700 DNA sequencer (Perkin–Elmer Applied Biosystems, Waltham, MA, United States), according to the manufacturer’s instructions. Sequences were edited using the BioEdit software package v. 7.0.9 (Ibis Biosciences, Carlsbad, CA, United States). The sequences were aligned with the NCBI genomic database using the BLAST search alignment tool (Altschul et al., 1990) and were related to the phylogenetic groups using the RDP Naive Bayesian Classifier (Wang et al., 2007). The 16S rRNA bacterial and ITS1 rRNA fungal gene nucleotide sequences determined in this study were deposited into the Genbank database under accession numbers F779671-JF779675 and JF729186-JF729196.

The normal distribution of the data was determined with the Kolmogorov-Smirnov test and the statistically significance of differences between treatments were evaluated by one-way analysis of variance (ANOVA). The homoscedasticity between the groups was studied using the Levene test. The post hoc tests were carried out using Tukey, when there was homoscedasticity, and by using Games Howel when there was none. In the case of non-normal distributions, the Kruskal–Wallis test was used to determine whether or not the consequences of treatment were significant. ANOVA and post hoc analysis with Tukey’s multiple-range test with a significance level of 0.05, was applied to the results to determine their statistical significance, (P < 0.05) by the use of SSPS V15, IMB SPSS Statistics, United States. DGGE bands of the samples were processed by GeneTools analysis software (Syngene, United Kingdom) and they were studied by multivariate detrended correspondence analysis (DCA) by using the Canoco Software (version 5, Microcomputer Power, NY, United States).

The soil was loamy clay, neutral (pH: 7.6), with a high electric conductivity (1,416 μS cm^–1) and a low content of heavy metals. Chemical analysis showed a total nitrogen and phosphorous content of 300 mg-N kg^–1 and 6 mg-P kg^–1, respectively, a moisture of 12% (w/w), and a water holding capacity 34% (w/w). Elemental analysis yielded a molar ratio C(carbon):H(hydrogen):N(nitrogen):S(sulfur) of 13:3:0.2:0.6, which indicated that the C:N ratio was slightly above the recommended values (25:1:1 to 38:1:1), according to Atagana (2009) and Alexander (1999), for the fungal PAHs bioremediation under an optimum water and oxygen content. The comparative elemental analysis of a pristine soil sample taken near the polluted site under study indicated that the carbon content in the later was increased from 2% up to 13% (Supplementary Table 1). Likewise, the total nitrogen content in the contaminated soil was fivefold higher with respect to the non-polluted soil. Such high values might be explained by the fertilization applied during the previous landfarming activities. Further increases in nitrogen content in bioaugmented and biostimulated microcosms (BS), in relation to the control, were due to the addition of inorganic nutrients.

The content of aliphatic and aromatic hydrocarbon fractions evidenced that readily degradable compounds had already been removed from soil during previous landfarming treatments in the polluted site. Concentrations of all PAHs subjected to regulation exceed the threshold values proposed by the Argentinian and European legislation for agricultural and residential soils (Table 1). Heterotrophic and hydrocarbonoclastic microbial populations quantified by the MPN were relatively high (heterotrophs: 2 × 10⁷ MPN g soil^–1, alkanes and PAHs degraders: 8 × 10⁵ and 1 × 10⁴ MPN g soil^–1 respectively), indicating that there was an abundant indigenous microbial population with hydrocarbon-degrading capabilities.

Twelve morphologically different fungal strains were isolated successfully from the aged polluted soil and were maintained as pure cultures in MYA slants. Only one of these strains, labeled as B1, was able to grow on PHEN agar, while the remaining were labeled correlatively from A1 to A11. The identification of these fungi was attempted by sequencing the ITS1 rRNA region. Only five of the 12 isolated fungal strains turned out to be different after sequencing (Table 2). All sequences showed high similarity (99–100%) with other sequences from type material deposited in GenBank. Sequence matches from the isolated fungi corresponded to Penicillium spp., Penicillium crysogenum, Ulocladium spp., Ulocladium atrum, Aspergillus terreus (three strains), Fusarium oxysporum (four strains), and Aspergillus parasiticus (three strains). These fungi are rather fast-growing and highly sporulating ascomycetes and might, therefore, not be representative of the dominant or active fungal hydrocarbonoclastic diversity of the soil. Nevertheless, all these species have commonly been described as HMW-PAHs degraders (D’Annibale et al., 2006; Cerniglia and Sutherland, 2010; Romero et al., 2010). Of particular interest is the genus Fusarium in that several strains from this group have been associated to the biodegradation of petroleum hydrocarbons, as reviewed recently by Prenafeta-Boldú et al. (2019). This extremely diverse and ubiquitous genus is widely distributed in soils but also in plant material, either as harmless saprobe or as a pathogen. The assimilation of benzo(a)pyrene as the sole source of carbon and energy by a F. solani strain has been demonstrated under laboratory conditions (Rafin et al., 2000). The involvement of CYPs, lignin peroxidases, and laccase enzymes was suggested as fundamental for the biodegradation process in this fungus (Rafin et al., 2008). Furthermore, the translocation of the 4-ringed pyrene along mycelial networks of the closely related Fusarium oxysporum has also been demonstrated (Guivernau et al., 2012). The relatively high frequency of isolation of Fusarium strains in this work could be an indication of the predilection of this genus toward PAHs polluted environments.

Five out of the twelve isolates were positive for the appearance of a dark halos around fungal colonies in mineral ABTS agar, which is an indication of extracellular laccase secretion. These positive strains included Fusarium oxysporum (strains A5, A6, A8, and B1), which displayed dark-green halos, and Ulocladium sp. (strain A2) with a dark-purple halo (Supplementary Figure 1A). This later strain A2 was the only one that also displayed a halo in tannic acid agar (Supplementary Figure 1B), pointing to the secretion of polyphenol oxidases, whereas all the other tested strains were negative (Table 2). No indication of Mn peroxidase secretion was observed through coloration of RO agar in any of the tested strains, with the exception of the control ligninolytic fungus Trametes versicolor, which tested positive on all chromogenic agar assays.

Despite limited knowledge at the proteomic level, extracellular laccases from Fusarium oxysporum have been proposed as one of the main factors responsible for its virulence as a plant pathogen. This claim has been substantiated by the finding of three genes that may encode laccases sensu stricto in a comparative genomic study of twelve F. oxysporum strains (Kwiatos et al., 2015). Putative laccase and phenol oxidase production observed in colonies of the darkly pigmented Ulocladium sp. (class Dothideomycetes) is also noteworthy. These two enzymes are involved in the fungal melanization process and laccases have also been found to be secreted by other dematiaceous fungi from the Dothydeomycetes (Tetsch et al., 2005). Fungal laccases have been recognized as industrial enzymes that might also be of interest in the mycoremediation of PAH-polluted soils. For example, Li et al. (2018) revealed that the white-rot fungi Pycnoporus sanguineus transformed phenanthrene via laccase and CYP, and benz[a]anthracene via laccase. Despite the fact that further research is needed, our enzymatic screening provides a first hint toward the involvement of specific enzyme systems in the fungal biodegradation of PAHs. In fact, a previous study in aged-creosote polluted soils by means of 16S rRNA amplicon sequencing, revealed that the main fungal community in a biostimulated soil was dominated by native ascomycetes (i.e., Fusarium spp. and Scedosporium spp.) before and after the biostimulation process. These species were even dominant after bioaugmentation with ligninolytic strains—and supplementation with lignocellulosic material—of the WRF Trametes versicolor and Lentinus tigrinus (Lladó et al., 2015). Further research must also consider other fungal enzymes, such as unspecific peroxygenases (UPOs), that act as the extracellular counterparts of intracellular CYP systems, and might contribute to the biodegradation of PAHs as well (Karich et al., 2017).

After 60 days of incubation, no significant differences were observed in terms of ecotoxicity between the different soil treatments. However, after 120 days, a remarkable reduction of ecotoxicity (increase of the EC50 value) was observed with fungal bioaugmentation (B), compared to biostimulation (BS) and the control (C) untreated soil (P < 0.05, Figure 4), so that low toxicity values comparable to the clean soil were obtained (Figure 4). TPH biodegradation efficiencies of 34–40% obtained with B also resulted in a significant decrease in toxicity. This phenomenon could be explained by a decrease of the content of secondary metabolites, due to higher mineralization rates of hydrocarbons during biodegradation, thanks to the synergistic interactions between the autochthonous soil microbiota and the bioaugmented fungi. The implemented V. fischeri bioluminescence assay has been reported as the most sensitive ecotoxicological test for a wide range of environmental contaminants, compared to other bacterial assays such as nitrification inhibition, respirometry, ATP luminescence and enzyme inhibition (Cotou et al., 2002). It has also shown good correlations with other standard acute toxicity assays across a wide spectrum of toxicants in soils and sediments (El-Alawi et al., 2002; Abbondanzi et al., 2003). Although mostly applied for aqueous phase samples and organic extracts, the test can also be conducted directly on soil and sediment samples to measure the toxicity due to the bioavailable fraction (Parvez et al., 2006). Nevertheless, the availability of PAHs due to the extractive saline solution of the method rather than their total concentration may affect the results, as reported by Harkey and Young (2000). In our case, given the rather low quantitative contribution of PAHs to TPHs [88.17 mg kg^–1 of PAHs on 45,000 mg kg^–1 of TPH (IR) or 16,114 mg kg^–1 TPH C10-C35], the polar metabolites or partial oxidation products affecting the toxicity test could derive mainly from the hydrocarbons comprised in TPHs and not only from the HMW-PAHs. This justifies further the use of TPHs for the statistical regression analysis and Microtox readings. In this sense, the different degradation of these compounds described previously between B and BS is also coherent with the results of this test.

Bioaugmentation caused an increase of two to three orders of magnitude of the total bacterial heterotrophic population counts, compared to those from the initial soil, up to values of 2.54 × 10⁸ MPN g soil^–1 and 1.84 × 10⁹ MPN g soil^–1 after 30 and 60 days of incubation, respectively (Table 3). In addition, from day 60–120, the alkane-degrading bacterial populations increased by one order of magnitude and stabilized at 7.8 × 10⁷ MPN g soil^–1. Similarly, PAHs-degrading bacteria increased one and three orders of magnitude at days 30 and 60, respectively, but decreased down to a basal level of 7.1 × 10⁴ MPN g soil^–1 after 120 days.

In the biostimulation treatment, the aliphatic-degrading bacterial populations increased in numbers by two orders of magnitude, up to 1.4 × 10⁶ MPN g soil^–1, while the total heterotrophs initially decreased by one order of magnitude with respect to the control soil without treatment, remaining stable until day 120, at 6 × 10⁵ MPN g soil^–1. The addition of water and nutrients did not cause a biostimulation of the PAHs-degrading bacterial populations, which remained stable in the order of 10⁴ MPN g soil^–1 throughout the study time. Interestingly, during fungal biaugmentation the aliphatic-degrading bacteria achieved the highest counts in the late stages (60–120 days), with values above 10⁷ MPN × g^–1 (2.5 magnitude order higher than those observed in the BS treatment). Such high bacterial counts were also coincident with the maximum biodegradation of TPH of late stages in B (Figure 1). These results demonstrate that the inoculation of the autochthonous fungi promoted the establishment of active hydrocarbon-degrading bacterial populations that biodegraded both aliphatic hydrocarbons (late stages) and PAHs (early and late stages) in aged-polluted soils. Similarly, D’Annibale et al. (2006) also described an increase of hydrocarbon-degrading bacterial populations after fungal bioaugmentation, which could be linked to a potential increase of PAHs bioavailability due to the generation of more polar metabolites by hydrocarbon-degrading fungi (Meulenberg et al., 1997; Capotorti et al., 2005).