1. Introduction
Hydrocarbon pollution is still one of the most important issues affecting both natural and human-transformed environments [
1,
2,
3]. The last huge oil spill on Brazilian beaches is another example of this significant problem [
4]. Consequently, several techniques are employed for the restoration of impacted ecosystems. Bioremediation is a method which is chosen very often, because it is relatively inexpensive and is generally believed to have minimal unintended side effects [
5]. Bioremediation of soils contaminated with petroleum hydrocarbons is usually based upon two approaches: biostimulation (an addition of the appropriate nutrients and/or electron acceptors to stimulate the degradation capacity of the indigenous soil microorganisms) and bioaugmentation (an inoculation of soil with high numbers of autochthonous or allochthonous hydrocarbon-degrading microorganisms) [
6]. The results obtained in different studies advocate the application of either the former [
7,
8,
9] or the latter method [
10,
11,
12]. Numerous microorganisms possess an ability to use hydrocarbons as the sole source of carbon and energy, but only representatives of a few genera, notably
Mycolicibacterium (formerly included in
Mycobacterium genus),
Pseudomonas, and
Rhodococcus, are capable of degrading both aliphatic and aromatic hydrocarbons [
13]. Since crude oil is a complex mixture of various hydrocarbons and their derivatives, its biodegradation clearly requires the action of different microorganisms [
14]. Nutrient amendments may lead to an unspecific stimulation of various soil microorganisms [
15], not only hydrocarbon degraders, which in turn can possibly slow down remediation processes. On the other hand, bioaugmentation based on isolating autochthonous (indigenous) hydrocarbon-degrading strains, and then growing them to high densities and finally inoculating the same soil environment they were derived from, yielded very good results [
16,
17,
18]. This approach, though effective, is time- and labor-consuming. Thus, bioaugmentation is performed much more frequently with an application of allochthonous microorganisms [
19,
20,
21,
22,
23,
24,
25]. However, in this case, additional factors make the whole process much more complicated. First, not all hydrocarbon degraders are equally suitable to be included in degradative inoculants. For example, some are considered opportunistic pathogens (like
Burkholderia cepacia [
26] or
Pseudomonas aeruginosa [
27,
28]) and belong to biosafety level classification 2. Therefore, their field-scale use would pose a serious risk to environmental and human health [
29]. The second issue to be addressed is the proper selection of strains. Bacosa et al. [
30] showed that some bacterial members of the hydrocarbon-degrading consortium were initially inhibited by the presence of aromatic hydrocarbons and seemed not to be active in hydrocarbon degradation but utilized the metabolic products. In oxygen-limiting conditions, bioaugmentation with the strain
Rhodococcus erythropolis T902.1 gave better results than the biostimulation treatment [
31], while under harsh environmental conditions (high hydrocarbon load and low moisture content), the action of
Mycolicibacterium frederiksbergense IN53 (a K-strategist) was superior to that of
Acinetobacter sp. IN47 (an r-strategist) [
32]. Finally, one should consider what the fate will be for the introduced non-indigenous microbes and how they will affect indigenous microbiota. This also seems to be dependent on the selected organism. Some studies suggest that augmented strains persist in their new environments [
23,
33,
34], whereas others report the inability of non-native bacteria to compete with indigenous microbiota for a longer period of time [
12,
31].
In this study, we prepared two hydrocarbon-degrading microbial inoculants (an undefined community C1 and a defined mixed culture C2), tested how they perform in soil polluted with petroleum hydrocarbons with unusually high polycyclic aromatic hydrocarbon (PAH) content, and compared their influence with the action of indigenous microbiota. Apart from examining their biodegradative efficiency by chromatographic analyses, we also used a set of toxicity tests (biotests) to confirm that the remediation process did not leave toxic intermediates. We also checked whether the addition of various allochthonous microorganisms present in the C1 and C2 can change the native microbial community.
3. Discussion
In this study, the effects of biostimulation and biostimulation combined with the bioaugmentation of soil contaminated with aliphatic hydrocarbons and a high load of PAHs were compared. Bioaugmentation is regarded as a strategy to enhance bioremediation via the introduction of a microbial agent pre-adapted to the environmental conditions of the contaminated site. There are two main criteria that should be considered for the selection of a microbial formula—the hydrocarbon-tailored catabolic potential of microorganisms, their functional activity, and their persistency under the given conditions. Therefore, the choice of microorganisms for inoculation should not be indiscriminate. This is quite easy to achieve when indigenous microorganisms are chosen for bioremediation purposes [
16,
17,
18,
35]. However, it should be stressed that under the term “bioaugmentation”, one should consider only bioaugmentation combined with parallel nutrient amendments [
36], as we did here and in our earlier studies [
16,
17]. We applied two different approaches to obtain allochthonous inoculants, namely a hydrocarbon-degrading community as well as a mixed culture. They both were constructed with bacteria originating from long-term contaminated soils with crude oil or diesel fuel, which are composed of aliphatic and aromatic hydrocarbons. The mixed culture C2 was developed via the careful selection of bacterial strains with identified metabolic capabilities towards both
n-alkanes and monocyclic and polycyclic aromatic compounds. This approach (that is, constructing a defined mixed culture with well-characterized metabolic activities) is a widely accepted solution [
19,
20,
37,
38]. In turn, the community C1 was obtained via enrichment from soils with a long history of contamination with crude oil and a random selection of hydrocarbon degraders. As determined by Miseq sequencing, the C1 contained a myriad of different bacteria but was dominated by few genera. Among these taxa,
Alcaligenes,
Pseudochrobactrum, Aquamicrobium,
Enterococcus,
Brevundimonas, and
Alkaliphilus presented the highest abundance. Representatives of the genera
Alcaligenes [
39,
40],
Aquamicrobium [
41],
Brevundimonas [
42,
43], and
Enterococcus [
44,
45] were reported to have hydrocarbon-degrading abilities, while others were not. Thus, this is evidence that both the enrichment method and randomly picking colonies grown on crude-oil-coated agar plates, apart from yielding hydrocarbon degraders, may produce some unwanted “companions” (e.g.,
Clostridium sensu stricto 16 and
Clostridioides). In this case, cooperation in the metabolic degradation of pollution should have naturally selected the microbial population, as shown recently [
46,
47]. On the other hand, some of these members might have been considered not necessarily beneficial since some representatives of the
Brevundimonas [
48] and
Enterococcus [
49] genera are also known as opportunistic pathogens. The composition of the community C1 was analyzed but the detailed catabolic activities of the C1 and its individual members were not examined. However, the observed removal of alkanes and PAHs in bioaugmented microcosms (BA-C1, BA-C2) indicated that both microbial inoculants were composed of organisms capable of degrading various structurally diversified hydrocarbons, even though a considerable increase in the relative abundance of the indigenous potential PAH-degrader
Proteiniphilum [
50] was noted in the former microcosms (BA-C1). Furthermore, in our study, the numbers of introduced allochthonous microbes were relatively low (not exceeding 10
6 colony forming units/g d.w. soil (cfu/g d.w. soil)) and were only slightly higher, if at all, than the numbers of indigenous microbiota (8.7 ± 4.3 × 10
5 cfu/g d.w. soil). These observations fully support the idea of bioaugmentation as a bioremediation strategy and are contrary to some other reports [
22,
25].
The presented chromatographic results clearly indicate that the introduction of the mixed culture was a more efficient bioremediation approach than the other studied bioaugmentation variant, as well as biostimulation sensu stricto. Additionally, only inoculation with the C2 treatment resulted in almost complete removal of nearly the total range of the analyzed compounds. In turn, long-chain
n-alkanes, as well as four- and five-ring PAHs, were poorly degraded in the biostimulated soil and less efficiently in the one bioaugmented with the community C1. Furthermore, the BA-C2 treatment seemed to be a suitable solution for the restoration of hydrocarbon-contaminated soil with a high load of PAHs since it led to the most effective elimination of both toxic and genotoxic effects. The members of the mixed culture were selected according to their hydrocarbon-degrading capabilities. Therefore, the mentioned bacteria might have predominantly participated in the conversion of toxicants to less or non-toxic intermediates or by-products. In turn, the biodegradation efficiency of the other bioaugmentation option was limited. Thus, the presence of residual pollutants might have posed a threat to living organisms, as revealed by the bioassay results. It was shown that a decrease of hydrocarbon content in bioremediation treatments does not always infer a reduction in toxicity [
51,
52]. Mineralization of hydrocarbons in aerobic conditions results in formation of non-toxic substances, such as CO
2 and water. Some transformations that do not lead to complete metabolism of the parent compound are inevitable in a complex system such as polluted soil [
52,
53]. In the case of PAHs, incomplete metabolism (especially in the oxidation step) of an aromatic structure may lead to formation of more polar and mobile intermediates, which may be as toxic as a parent compound or even more so [
54]. The relaxed substrate specificity of aromatic ring-hydroxylating dioxygenases is the reason for oxygenated PAHs (oxy-PAHs) generation. These compounds are regarded as intermediates but also “dead-end” metabolites [
55]. Moreover, identification of the novel product of pyrene degradation in post-remediated soil [
53] indicates that our current knowledge on hydrocarbon biotransformation is still limited since it is mainly based on the information gathered from studies which were performed using a single, model compound. In turn, degradation of a complex mixture, such as a PAH mixture, may also involve antagonistic effects associated with the use of an unusual pathway that is not normally used by the organism in the degradation of a single compound [
56]. Thus, bioremediation of soil contaminated with hydrocarbons may lead to generation of unknown compounds with unknown toxicity. It was demonstrated that products of PAH transformation were accumulated in several soil remediation studies [
52,
54,
57] and they may be responsible, at least in part, for an increase in toxicity observed during remedial treatments of contaminated soils ([
52,
54] and references cited therein) since these compounds are not degraded concomitantly with PAHs. Despite recent progress in analysis of compounds generated during PAH biodegradation [
53], it is challenging to monitor the presence and concentration of these substances. Therefore, bioassays are an alternative to evaluate hazards before, during, and after remediation. However, it should be highlighted that the toxicity/mutagenicity response is an effect of the total load of soil toxicants and potential interferences among them. The abilities of the microorganisms to (1) mineralize the target substances or (2) generate products to such low levels that they are safe from the perspective of risk, will determine the efficacy of bioremediation. In this study, potential soil toxicity after treatments was assessed by a set of bioassays. In each bioremediation variant, removal of hydrocarbon content was associated with a reduction in toxicity and these observations are consistent with those of Płaza et al. [
58].
Our findings also prove that inoculation using pre-selected isolates with the ability to metabolize both
n-alkanes and PAHs is a better option than the enrichment approach for bioremediation, although solutions like the immobilization of bacterial cells may increase the degradation rate of hydrocarbons [
59,
60]. Most PAHs, as well as
n-alkanes, were degraded most efficiently in the BA-C2 microcosms. Additionally, both the toxicity and mutagenicity assessment revealed that BA-C2 microcosms presented the highest levels of soil reclamation among all treatments, although some level of toxicity was detected due to residual PAH content. The mixed culture C2 produced a significant increase in
n-alkane and PAH biodegradation efficiency and outperformed the expectations placed upon it. The proper selection of bacterial strains is always a crucial step, and well-known hydrocarbon degraders may succeed or fail depending on the environment they are introduced to [
11,
32]. In favorable conditions, bioaugmentation with even a single strain resulted in the effective removal of hydrocarbons [
61] and even within one species, one could find a difference among individual strains in this regard [
62].
There are many reports showing the success of bioaugmentation to remediate hydrocarbon-polluted soils [
17,
23,
32,
62,
63,
64]. However, its positive effects may be limited to the early stage of treatments [
12,
24]. On the other hand, in some cases this strategy is not the best remedial option [
65]. In this study, we showed the advantage of bioaugmentation over biostimulation as well as the advantage of the mixed culture C2 (defined) over the community C1 (undefined). The latter observations are consistent with the findings of other authors [
66,
67]. Microbial strains growing and evolving together, as in the C1, should be more productive as a group since some adapt to utilizing the by-products of other members. However, in our study the total number of hydrocarbon degraders present in the C2 was apparently higher than in the C1, because each strain of the C2 exhibited wide hydrocarbon-degrading capabilities, while presumably only some members of the community C1 could have been regarded as true hydrocarbon degraders. Moreover, the antagonism between bacterial hydrocarbon-transforming populations in the C1 during degradation should not be excluded. Festa et al. [
67] proved that negative interactions existed within undefined communities even if a single compound (phenanthrene) was degraded and these interactions influenced the final outcome. The other possible advantage of the mixed-culture C2 was that it included both K- (for example
Mycolicibacterium frederiksbergense IN53) and r-strategists (for example
Pseudomonas sp. IN132), that is, fast- and slow-growing organisms in roughly similar numbers (within the same order of magnitude), after growing in a nutrient-rich medium. In the undefined community C1 the K-strategists were virtually outgrown by r-strategists as judged by the low (1%) abundance of slow-growing
Leucobacter species. Our results are contrary to other studies in which an excellent efficiency of hydrocarbon-degrading communities/consortia [
30,
68] was clearly presented. However, those communities/consortia were grown exclusively on mineral media supplemented with hydrocarbons and therefore consisted only of hydrocarbon degraders. Sydow et al. [
68] demonstrated that members of the hydrocarbon-degrading consortium drastically change their abundance upon exposure to different carbon sources (various hydrocarbons and biodiesel), whereas the overall structural and functional integrity of the consortium was maintained, but was apparently not retained in our undefined community C1. On the other hand, field-scale use usually requires large volumes (hectoliters or even cubic meters) of microbial inoculants and this is usually achieved by growing cells in a nutrient-rich medium or in a mineral medium supplemented with easily accessible water-soluble industrial by-products such as molasses [
69]. Furthermore, from this point of view, mixed cultures could be also regarded as effective inoculants for bioremediation purposes [
66,
70].
We also analyzed how all three treatments affected the indigenous soil microbial community. Since only pooled samples were analyzed, we did not perform a multivariate statistical analysis. Thus, this analysis was primarily used to evaluate the effects of the treatments on the bacterial taxa that were characterized by the highest abundance, as in [
12]. We focused mostly on bacterial families and genera because the observed differences were significant on those taxonomical levels. The finding that all treatments strongly affected the anaerobic part of the native community is not unusual, as hydrocarbon removal usually indicates a facilitated mass transfer rate of air (oxygen) to soil (this effect can be also achieved by the prior use of a bulking agent [
71]). Moreover, each treatment was performed in aerobic conditions. The native hydrocarbon-degrading population was also impacted since potential hydrocarbon degraders present in the non-amended soil, such as
Immundisolibacter [
72],
Altererythrobacter [
73,
74], and
Pseudoxanthomons [
75], were clearly outcompeted by other microorganisms. However, the differences in dominant genera between all treated microcosms (with the exception of the most abundant,
Luteimonas) and the weak presence of typical hydrocarbon degraders were unexpected. This was also true for the introduced bacteria, among which only
Pseudomonas and
Mycobacterium/
Mycolicibacterium exhibited relative abundances higher than 1.5% in the BA-C2 microcosm, but these genera were also present in the control soil sample, so there is no way to determine whether they were indigenous or non-indigenous. The abundance of the other introduced hydrocarbon degraders was marginal, but the
Gordonia and
Rhodococcus genera were present only in soil inoculated with the C2. Moreover, our metatranscriptomic results showed that the
alkB transcripts assigned to
Mycobacterium/
Mycolicibacterium,
Pseudomonas,
Arthrobacter, and
Gordonia were found in the BA-C2 microcosms. These observations suggest that non-native bacteria present in the defined mixed culture sustained competition from indigenous microbiota for at least the time required to remove most of the contaminants; however, they were unable to establish larger populations in their new habitat.
Although we did not study the influence of natural attenuation on hydrocarbon removal in great details, our results show relatively low effectiveness of this approach compared to other reports [
65,
76,
77,
78]. In turn, other authors also noticed insufficient effects of naturally occurring processes as a remediation strategy [
79,
80]. It was suggested that both soil properties and the indigenous soil microbial population determine the biodegradation effect [
65]. Natural attenuation occurs in environments, but at least one of following conditions must be fulfilled: (1) it should be the adequate concentration of available P and N, and/or (2) the soil texture should allow bacteria to have facilitated access to hydrocarbons and oxygen. In laboratory conditions, the latter could be achieved by passing the soil through a sieve [
76,
77,
78]. However, this practice was not applied in this study. Other parameters such as soil pH and nutrient availability should also be considered. In this study, the pH of non-amended soil was slightly acidic (6.1) and the C:N:P ratio was not optimal for bacterial requirements (approximately 100:2:0.4). Addition of nutrients and a change in pH value led to improved biodegradation as revealed by hydrocarbon removal in biostimulated soil (BS). This finding also suggests that studied native microorganisms may have the potential to metabolize mentioned compounds, but are not as active in non-optimal conditions as those in the control microcosm. This also confirms the results obtained by Adetutu et al. [
79] and Košnář et al. [
80]. In our study, the presence of crude oil led to developed anaerobic conditions, in which oxygen transfer was limited. Even performed manipulations (aeration) did not help to change these conditions as indicated by the structure of the bacterial community in the control soil (natural attenuation) after a 60-day incubation. Although the structure of the bacterial community at the beginning (0 day) of the process was not analyzed, the non-amended soil at the end of experiment was rich in obligate anaerobes (e.g.,
Anaerolinea, Anaeromyxobacter,
Ruminiclostridium,
Longilinea and
Geobacter; together more than 11% of all bacteria). The effectiveness of anaerobic hydrocarbon biodegradation is significantly lower than aerobic processes. Our data also proved that the action of the native population of hydrocarbon-degrading microorganisms in the control soil was limited. Light, relatively easily biodegraded
n-alkanes (
nC
8-
nC
12,
Figure 1A) and low molecular weight PAHs (naphthalene, phenanthrene,
Figure 1B) were still present in the 50-year weathered soil pollution. Assuming high effectiveness of indigenous hydrocarbon-degrading populations, only more recalcitrant compounds should be present in the analyzed soil as reported by Gallego et al. [
81]. It should be noted that the extent of natural attenuation processes depends also on hydrocarbon bioavailability. The slow desorption from the sorbed to aqueous phase means restricted bioavailability and thereby low extent of hydrocarbon degradation.