Field scale biodegradation of total petroleum hydrocarbons and soil restoration by Ecopiles: microbiological analysis of the process

Abstract

Ecopiling is a method for biodegradation of hydrocarbons in soils. It derives from Biopiles, but phytoremediation is added to biostimulation with nitrogen fertilization and bioaugmentation with local bacteria. We have constructed seven Ecopiles with soil heavily polluted with hydrocarbons in Carlow (Ireland). The aim of the study was to analyze changes in the microbial community during ecopiling. In the course of 18 months of remediation, total petroleum hydrocarbons values decreased in 99 and 88% on average for aliphatics and aromatics, respectively, indicating a successful biodegradation. Community analysis showed that bacterial alfa diversity (Shannon Index), increased with the degradation of hydrocarbons, starting at an average value of 7.59 and ending at an average value of 9.38. Beta-diversity analysis, was performed using Bray-Curtis distances and PCoA ordination, where the two first principal components (PCs) explain the 17 and 14% of the observed variance, respectively. The results show that samples tend to cluster by sampling time instead of by Ecopile. This pattern is supported by the hierarchical clustering analysis, where most samples from the same timepoint clustered together. We used DSeq2 to determine the differential abundance of bacterial populations in Ecopiles at the beginning and the end of the treatment. While TPHs degraders are more abundant at the start of the experiment, these populations are substituted by bacterial populations typical of clean soils by the end of the biodegradation process. Similar results are found for the fungal community, indicating that the microbial community follows a succession along the process. This succession starts with a TPH degraders or tolerant enriched community, and finish with a microbial community typical of clean soils.

Keywords: Ecopile; bioremediation; hydrocarbon; microbial succession; microbiota.

Figure 1

Boxplot representing the bacterial Shannon index values for each timepoint and Ecopile. Boxplots contain the Shannon index values of the three replicates of each sample. Differences between Shannon index values at the different timepoints were assessed with the Kruskal-Wallis test. Asterisks represent the significance at a p adjusted-value ≤0.05. Colors of boxplots according to the different Ecopiles.

Figure 2

Bacterial relative abundances at the level of class for the different Ecopiles at the three sampling times. Relative bacterial ASVs abundance at the level of class of the Ecopiles at each of the timepoints for the 20 most abundant classes. The remaining classes were grouped under the category Other. The barplots represent the sum of ASVs from the three replicates.

Figure 3

Clustering analysis of the bacterial communities within Ecopiles at the different timepoints. (A) Principal coordinate analysis (PCoA) of Ecopiles using Bray-Curtis distances. Colors according to Ecopile, shapes according to timepoint and size according to Shannon index. Averages for each ecopile and sampling time are represented. (B) Heatmap of the hierarchical clustering of Ecopiles using Bray-Curtis distances and dendrogram. Color scale indicates Bray-Curtis distances.

Figure 4

Differential abundance analysis of bacterial ASVs. The analysis shows the log2FoldChange of individual ASVs from the different bacterial families that significantly changed comparing the T2 and T5 timepoints (December 2019 and November 2020, respectively, p adjusted value <0.01). Colors according to the phylum to which the represented ASVs and families belong.

Figure 5

Non-metric multidimensional scaling (NMDS) analysis using Bray-Curtis distance matrix. NMDS plot where the shapes represent the samples according to sampling times and colors according to the Ecopile. Bacterial classes and contaminants were used as explanatory variables driving the Ecopiles distribution pattern.

Figure 6

Fungal relative abundances at the level of class and family and Shannon indexes of Ecopiles at the three sampling times. (A) Relative ASVs abundance at the level of class of the Ecopiles at each of the timepoints. Low abundance taxa were grouped under the category Other. The barplots represent the average of ASVs from Ecopiles. (B) Evolution of Shannon alpha-diversity indexes within Ecopiles along sampling time.

Figure 7

Clustering analysis of the fungal communities within Ecopiles along the different timepoints. (A) Principal coordinate analysis (PCoA) of Ecopiles using Bray-Curtis distances. Colors according to Ecopile, shapes according to timepoint and size according to Shannon index. Averages are represented for each ecopile and sampling time. (B) Heatmap of the hierarchical clustering of Ecopiles using Bray-Curtis distances and dendrogram. Color scale indicates Bray-Curtis distances.

Figure 8

Differential abundance analysis of fungal ASVs. The analysis shows the log2Fold Change of individual ASVs from the different fungal families that significantly changed comparing July 2019 and November 2020 sampling times (p adjusted value <0.01).