By applying the metabarcoding technique, we evaluated the effect of heavy metals soil smooth contamination on prokaryotic soil communities in areas surrounding cement plants, using the structural community metrics (α- and β-diversity metrics) and functional community metrics (PICRUSt2 approach). The comparative diversity analyses concerning the soil bacterial communities accessed in this study revealed that communities in areas surrounding CPSs showed higher richness and diversity of bacterial groups (Shannon index) and equitability than those in CAs. Evidence from previous studies shows that there is no general trend indicating that bacterial diversity of soil communities responds positively or negatively to heavy metal soil contamination, whose results were similar to those we found (Mohamed & Adbelmajid 2017; Hong et al. 2015; Lazzaro et al. 2008; Ellis et al. 2001), as well as studies with opposite results (Pan et al 2020; Zeng et al 2020; Lin et al. 2019; Fajardo et al. 2019; Hemmat-Jou et al. 2018). These results reinforce the recent view that variation in alpha-diversity indices alone are insufficient as an indicator of environmental disturbance and that the idea that “the more diversity the better” is not necessarily a rule of thumb (Chase et al 2019; Shade, 2017). On the other hand when we consider phylogenetic diversity, which reflects the diversity of distinct evolutionary branches present in each community, we see no significant differences between CPSs and CAs. While intuitively we might assume that higher levels of phylogenetic diversity could harbor more functionally diverse ecosystems with greater resilience, in prokaryotic communities not all functional features necessarily present a phylogenetic signal given the capacity for genomic mobility exhibited by these organisms (Schloter et al. 2018; Srivastava et al 2012). Thus, even though we did not find significant differences between the CPS and CA communities, this is not clear evidence that the communities in the soil surrounding the CPSs were not functionally impacted by the deposition of heavy metals.
We found that the composition of the communities found in the three CPSs were similar to each other, but significantly different from the communities accessed in the control areas, which are agricultural production areas. Previous studies that have characterized soil bacterial communities from agricultural production areas have shown that prokaryotic communities in this type of environment are naturally diverse and their structure is directly affected by several conditioning factors, including pH, MOC, water content (Wang et al. 2020; Liu et al. 2019; Delgado-Baquerizo et al. 2018). However, several evidences have already revealed that when the soil of environments managed for agricultural production is impacted by long-term heavy metal contamination, these factors become the main determinants of the taxonomic composition of prokaryotic communities, as changes in soil chemical conditions alter the selective regimes on the bacterial groups present can lead to significant changes in community diversity and structure (Gong et al. 2021; Zhao et al. 2019; Abdu et al. 2017; Giller et al. 2009). Most recent studies on the effect of heavy metal contamination on community composition have observed high concentrations of heavy metals. Differently, we focused in this study on the effects of moderate heavy metal contamination caused by human action (cement production) and revealed that the composition of soil bacterial communities are sensitive to the presence of heavy metals even under smooth concentrations (Li et al. 2021; Pan et al. 2020; Zeng et al. 2020; Lin et al. 2019; Jiang et al. 2019). Although they conducted experiments under laboratory conditions, Song et al. (2021) found similar results to ours, which come from a field experiment. All these pieces of evidence, added to our findings, reinforce the potentiality of using variations in beta-diversity metrics as an analysis parameter in “structural community metrics” approaches to verify the effect of human activities on the quality of soil ecosystems under both high and moderate heavy metal contamination.
In our study, the phylum Proteobacteria was found ubiquitously, with a significantly lower abundance in the samples under impact from cement plants. This trend of higher abundance of proteobacteria in areas under less effect of heavy metals was also verified in other studies whose metal concentration was higher than those seen in our study (Li et al. 2021; Fajardo et al. 2019; Sun et al. 2018; Xu et al. 2016). On the other hand, we found the phyla Actinobacteria, Nitrospira, and Rokubacteria in higher abundance in soil samples from CPSs. Recent previous studies reported that the first two phyla are dominant in metal-contaminated areas, indicating they have individuals with suitable metabolic adaptations to survive in such environments (Li et al. 2021, Gong et al. 2021; Zeng et al. 2020; Pan et al. 2020; Hemmat-Jou et al. 2018). Several authors relate the high abundance of Actinobacteria in heavy metal-contaminated areas to the fact that this phylum presents several individuals with robust stress resistance, including members capable of acting as bioremediators (Alvarez et al. 2017; Barka et al. 2016; Mangold et al. 2012; Elangovan et al. 2010). In this study, the phylum Nitrospira was represented by ASV assigned to the family Nitrospiraceae which has members that perform important ecosystem functions in soils impacted by human activities such as nitrite oxidation (nitrogen cycle), iron oxidants in the biolixiviation of metal ores, and degradation of organic compounds (Daims 2014). The fact that we recorded a preferential distribution of some phyla with similar patterns to other studies that have addressed the effect of high concentrations of heavy metals in soil indicates that situations of slight soil contamination by heavy metals are able to favor the colonization of these environments by some phyla, as well as limit the growth of others. In our study we identified only one group of Archaea with a frequency above 1%, represented by a single ASV assigned to the family Nitrososphaeraceae (Phylum Thaumarchaeota). Spang et al. (2012) reported that members of this family possess metal-resistant genes and are found in high abundance in metal-polluted habitats. Evidence of the high prevalence of this family of archaea in heavy metal-contaminated environments has been reported by other studies that also used cultivation-independent molecular methods (Diquattro et al. 2020; Hemmat-Jou et al. 2018; Touceda-Gonzalez et al. 2015).
Although our study was not designed with the aim of identifying marker organisms of human impact or soil ecosystem quality (“taxonomic-based strategy”), we revealed in it that moderate heavy metal contaminations would be able to structure the composition of soil microorganism communities. This happens to the point that some bacterial families identified from the accessed SBAs present a distribution restricted to heavy metal contaminated (HPA) or non-high metal contaminated (HPA) environments. We consider that the use of molecular techniques that allow more refined taxonomic identification of the prokaryotic groups present, such as the shotgun metagenomics approach, coupled with more robust sampling, could be employed to identify the rate of suitable prokaryotes to be used in taxonomic-based approaches (Durazzi et al. 2021; Khachatryan et al. 2020; Lanzén et al. 2017; Quince et al. 2017).
In this study, we used the “structural community metrics” approach in association with the “functional community metrics” approach, on the premise that taxonomic and functional profiles may respond differently under environmental perturbations (Cheaib et al. 2018). Further, we chose to use the metabolic function profiles predicted by PICRUSt2 as parameters for diversity and functional structure analysis which, although they are not considered as direct evidence on functional profiles as metagenomics or metatranscriptomics analyses, this approach has the advantage of being more cost-effective (Cordier et al 2020; Cristecu et al 2018; Langille et al 2013). The comparative analyses of molecular functions predicted by PICRUSt2 (predicted KOs) for the CPS and CA communities, revealed that in general the predicted molecular function profiles (KOs) for the prokaryotic soil communities surrounding cement plants are different from the profiles found for the communities in the control areas. This result indicates that, even if the prokaryotic communities from the soil of the impacted areas (CPS) presented a higher taxonomic diversity, it was not enough to generate ecological equivalents to maintain a functional redundancy capable of structuring functional diversity. This result is therefore another source of evidence that even moderate levels of heavy metal soil contamination are capable of impacting soil ecosystems. Zeng et al. (2020), who also used predictive analyses using PICRUSt, found that sediments with higher concentrations of heavy metals experienced a significant reduction in the number of predicted molecular functions, as well as changes in their composition. Additionally, our evidence for the effect of heavy metals on the functional structure of prokaryotic communities is also supported by previous studies that used different methods than those commonly employed in environmental genomics (Sheik et al. 2012; Chen et al 2014; Kandeler et al. 2000, Kandeler et al. 1996).
Furthermore, comparative analysis of the predicted molecular functions structured by the secondary levels revealed that categories related to the provision of important ecosystem services were impacted by contamination of carbohydrate and lipid metabolism (soil carbon cycle), secondary metabolite production (interactions with rhizosphere and other soil organisms), and xenobiotic degradation (biodegradation and bioremediation). Another metabolic function category with significant change was the cell cycle-related functions category. Abdu et al. (2017) also reported that heavy metals such as Pb, Cd, Hg, and Ni can alter the cell cycle of bacteria from contaminated soils by inhibiting the cell cycle.
Also, we observed significant alterations in the “Signal transduction” category, with the KOs related to Two-component systems showing the greatest differences in abundance between the profiles found in CPS and CA samples. This result was also found by Zeng et al. (2020) who reported higher abundance of KOs associated Two-component systems in sediment samples with higher heavy metal concentrations. Two-component systems play vital signaling roles in bacteria and function in response to diverse environments, including stress situations from heavy metal contamination (Zeng et al. 2020).