Description of Microbial Communities of Phosphate Mine Wastes in Morocco, a Semi-Arid Climate, Using High-Throughput Sequencing and Functional Prediction

Abstract

Soil microbiota are vital for successful revegetation, as they play a critical role in nutrient cycles, soil functions, and plant growth and health. A rehabilitation scenario of the abandoned Kettara mine (Morocco) includes covering acidic tailings with alkaline phosphate mine wastes to limit water infiltration and hence acid mine drainage. Revegetation of phosphate wastes is the final step to this rehabilitation plan. However, revegetation is hard on this type of waste in semi-arid areas and only a few plants managed to grow naturally after 5 years on the store-and-release cover. As we know that belowground biodiversity is a key component for aboveground functioning, we sought to know if any structural problem in phosphate waste communities could explain the almost absence of plants. To test this hypothesis, bacterial and archaeal communities present in these wastes were assessed by 16S rRNA metabarcoding. Exploration of taxonomic composition revealed a quite diversified community assigned to 19 Bacterial and two Archaeal phyla, similar to other studies, that do not appear to raise any particular issues of structural problems. The dominant sequences belonged to Proteobacteria, Bacteroidetes, Actinobacteria, and Gemmatimonadetes and to the genera Massilia, Sphingomonas, and Adhaeribacter. LEfSe analysis identified 19 key genera, and metagenomic functional prediction revealed a broader phylogenetic range of taxa than expected, with all identified genera possessing at least one plant growth-promoting trait. Around 47% of the sequences were also related to genera possessing strains that facilitate plant development under biotic and environmental stress conditions, such as drought and heat.

Keywords: PICRUSt prediction; biodiversity; metabarcoding; microbial communities; phosphate mine wastes.

FIGURE 1

Boxplot displaying the concentrations of different chemical parameters. Organic carbon (OC) and nitrogen (N) are presented by percent. Major oxides such as silica (SiO₂), phosphorus pentoxide (P₂O₅), ferric oxide (Fe₂O₃), lime (CaO), magnesia (MgO), potash (K₂O), titanium oxide (TiO₂), and alumina (Al₂O₃) are in milligrams per kilogram. The letters present the compact display letters (CDL). Error bars represent standard deviation of the mean value (analyses performed in triplicates). Values in each column followed by the same letter are not significantly different at 0.05% (Tukey’s).

FIGURE 2

Microbial diversity present in the alkaline phosphate mining wastes. At the phylum level (A), all sequences are presented, while at the genus level (B), only the most abundant sequences are presented.

FIGURE 3

Non-metric multidimensional scaling (NMDS) presenting the differences of the microbial communities between the different replicates of phosphate mining waste samples. The stress values were < 0.05, indicating that these data were very well represented by the two-dimensional representation. Arrows are the projections of possible explanation variables obtained by vector fitting. Only correlations with a false discovery rate (fdr) corrected p < 0.05 were indicated. The angle and length of the vector indicate the direction and strength of the variable. The r² correlation coefficient and the p-values are presented in Supplementary Table 3.

FIGURE 4

Heatmap of the statistically significant coefficients of Pearson’s correlations (false discovery rate (fdr) corrected; p < 0.05) between the relative abundance of taxa and concentrations of environmental parameters across all sites. Only threshold higher than 0.7 was plotted.

FIGURE 5

Linear discriminant analysis effect size (LEfSe), (A) chart and (B) cladogram, on microbial diversity in different store-and-release phosphate mine waste layer, overburden waste rock (cells 1 and 4), and phosphate mine wastes (cell 3) produced during the concentration processes.

FIGURE 6

Correlogram representing the Pearson’s correlation coefficient r² between the 19 key bacterial genera ordered by the “hclust” method.

FIGURE 7

Binary heatmap highlighting presence () absence () of the interested enzymes in identified genera. Enzymes implicated in: cellulose degradation by producing the cellulase enzyme (EC:3.2.1.4); sulfur metabolism by oxidizing thiosulfate (EC:1.8.2.2); siderophores production by catalyzing aerobactin biosynthesis pathway (EC:1.14.13.59, EC:2.3.1.102, EC:6.3.2.38 and EC:6.3.2.39); phosphate solubilization by producing either one or multiple of the enzymes C-P lyases (EC:4.7.1.1), phytase 1 (EC:3.1.3.8), phytase 2 (EC:3.1.3.26), alkaline phosphatase (EC:3.1.3.1), acid phosphatase (EC:3.1.3.2), and phosphonatase (EC:3.11.1.1); nitrogen metabolism by catalyzing at least one of the following pathways: nitrogen fixation I (ferredoxin) (EC:1.18.6.1), ammonia assimilation cycle I (EC:6.3.1.2 and EC:1.4.1.14); ammonia assimilation cycle II (EC:6.3.1.2 and EC:1.4.7.1), ammonia assimilation cycle III (EC:6.3.1.2 and EC:1.4.1.13), ammonia oxidation I and IV nitrite producing (EC:1.14.99.39 and EC:1.7.2.6), nitrate reduction VI L-glutamine forming (EC:1.7.7.2, EC:1.7.7.1, and EC:6.3.1.2), and nitrate reduction VI L-glutamate forming (EC:1.7.7.2, EC:1.7.7.1, and EC:1.4.1.4), and finally in auxin biosynthesis by producing enzymes implicated in indole-3-acetate biosynthesis III (EC:1.13.12.3 and EC:3.5.1.4, or in indole-3-acetate biosynthesis IV (EC:4.2.1.84 and EC:3.5.1.4) or in indole-3-acetate biosynthesis V (EC:3.5.5.1).