Strong impact of micropollutants on prokaryotic communities at the horizontal but not vertical scales in a subtropical reservoir, China

Highlights

• Physico-chemicals and micropollutants had joint effects on prokaryotic communities.

• Micropollutants had stronger effects on the surface water than vertical communities.

• Micropollutants exerted stronger influence on the core than satellite groups.

• Lifestyle-dependent responses to micropollutants were observed.

• Eight micropollutants had strong associations with abundant core taxa.

Abstract

Micropollutants have become of great concern, because of their disrupting effects on the structure and function of microbial communities. However, little is known about the relative importance of trace micropollutants on the aquatic prokaryotic communities as compared to the traditional physico-chemical characteristics, especially at different spatial dimensions. Here, we investigated free-living (FL) and particle-associated (PA) prokaryotic communities in a subtropical water reservoir, China, across seasons at horizontal (surface water) and vertical (depth-profile) scales by using 16S rRNA gene amplicon sequencing. Our results showed that the shared variances of physico-chemicals and micropollutants explained majority of the spatial variations in prokaryotic communities, suggesting a strong joint effect of the two abiotic categories on reservoir prokaryotic communities. Micropollutants appeared to exert strong independent influence on the core sub-communities (i.e., abundant and wide-spread taxa) than on the satellite (i.e., less abundant and narrow-range taxa) counterparts. The pure effect of micropollutants on both core and satellite sub-communities from FL and PA fractions was ~1.5 folds greater than that of physico-chemical factors at the horizontal scale, whereas an opposite effect was observed at the vertical scale. Moreover, eight micropollutants including anti-fungal agents, antibiotics, bisphenol analogues, stimulant and UV-filter were identified as the major disrupting compounds with strong associations with core taxa of typical freshwater prokaryotes. Altogether, we concluded that the ecological disrupting effects of micropollutants on prokaryotic communities may vary along horizontal and vertical dimensions in freshwater ecosystems.

Introduction

Rivers are an important freshwater resources on Earth that serve the society in a variety of ways (Grill et al., 2019). Hundreds of thousands of dams have been constructed worldwide because of an increasing demand of water for human activities. Besides providing benefits to the society, dams also cause numerous harm to rivers (Fang and Min, 2011), via irregular hydrological conditions (e.g., water flow rate and water residence time) (Nilsson and Berggren, 2006), lowering the water quality (e.g., eutrophication and hypoxia) (Elçi, 2008) and chemical pollution (Fang and Min, 2011). Consequently, the microorganisms inhabiting the rivers are likely to be influenced by such disturbances in the riverine environments. Previous works suggested that prokaryotic communities serve as a good bio-indicator for evaluating the health of aquatic ecosystems (Sun et al., 2012; Zeglin, 2015), provided that they readily respond to the shifting environmental conditions (Hu et al., 2017b; Isabwe et al., 2018; Shen et al., 2018; Sun et al., 2017a; Sun et al., 2017b). In the events of river disturbances, the changes in the prokaryotic communities are not only linked to the changes of water physico-chemical properties such as temperature, salinity, dissolved oxygen (DO) and nutrients etc. (Hu et al., 2017a; Isabwe et al., 2018; Sun et al., 2017a; Sun et al., 2017b) but also to various chemical pollutants (Jeanbille et al., 2016; Sabater et al., 2016; Sun et al., 2012). Several studies have assessed the effects of either physico-chemical factors (Hu et al., 2017b; Isabwe et al., 2018; Sun et al., 2017b; Sun et al., 2017a) or chemical pollutants (Blunt et al., 2018; Jeanbille et al., 2016; Sun et al., 2012) on the diversity and composition of prokaryotic communities. However, little is known about the relative impacts of both of these river disturbances on the assembly of prokaryotic communities, especially, in case of freshwater ecosystems, where the disrupting effects of micropollutants on ecological disturbances have been taken very seriously in the recent years (Richmond et al., 2017).

Micropollutants, such as pharmaceuticals, personal care products, endocrine disrupting compounds and industrial chemicals have raised public concerns due to their pseudo-persistence in environment and potential threats to ecosystem and human health (Liu and Wong, 2013; Yang et al., 2012). Although they commonly occur at low concentrations in the natural environments (Liu and Wong, 2013; Richmond et al., 2017), yet they have implications at sub-lethal concentrations. Some microorganisms are capable of utilizing certain micropollutants as carbon sources, such as triclosan, bisphenol A, ibuprofen (Zhou et al., 2013) and ciprofloxacin (Liao et al., 2016), however, the growth and metabolism of the others may be influenced by the sub-lethal effects of the micropollutants (Richmond et al., 2017). Blunt et al. (2018) reported a relatively higher degradation capabilities of microorganisms in a micropollutant contaminated part of a river than that in its adjacent pristine part (Blunt et al., 2018). Our previous work also suggested that trace micropollutants levels exert relatively greater influence on the co-occurrence of central microbial species than the physico-chemical factors (Hu et al., 2017a). Hence, exploring the relationship among physico-chemical factors, micropollutants and aquatic prokaryotic communities will provide a better understanding of the extent to which these abotic factors affect the prokaryotic community.

Aquatic prokaroytic community composition varies both at seasonal and vertical scales in freshwater ecosystems (Garcia et al., 2013; Salcher et al., 2010; Yu et al., 2015). The vertical variation pattern was put forward as variation in the diversity and composition of the prokaryotic communities due to changes in physico-chemical factors, such as light, temperature, DO, nutrients and chlorophyll-a at different water depths (Salcher et al., 2011; Yang et al., 2015; Yu et al., 2014). The vertical and seasonal variations in the physico-chemical profiles are well documented (Garcia et al., 2013; Morrison et al., 2017; Yang et al., 2015; Yu et al., 2015), however, such variations in the micropollutant distribution are scarcely reported (Aristizabal-Ciro et al., 2017; Deyerling et al., 2016). Likewise, a knowledge gap also exists on the relative importance of micropollutants on the assembly of prokaryotic communities through the water column.

Prokaryotic communities in the aquatic environments can be categorized as particle-associated (PA) and free-living (FL) on the basis of cell size and lifestyle (Riemann and Winding, 2001). Since PA taxa specialize in occupying the unique niches provided by suspended particles (Dang and Lovell, 2016), PA and FL communities usually possess distinct compositions (Yung et al., 2016; Zhao et al., 2017), physiological capabilities and metabolic activities (Grossart et al., 2007). Moreover, due to the biofilm structure, PA communities may be more resistant to environmental stressors, such as disinfectants, biocides and others, than the FL communities (Dang and Lovell, 2016). A recent study has found that antibiotics can enhance the aggregation of specific bacterial taxa with the suspended particles in the drinking water distribution systems (Wang et al., 2018a). These findings raise the possibility that the micropollutants are likely to influence the compositions of PA and FL prokaryotic communities, differently, to some extent. This phenomenon may be addressed in spatial dimensions (e.g., horizontal and vertical scales), provided that the hydrological dynamics may influence the distribution of micropollutants in water bodies. Furthermore, prokaryotic species can be further partitioned into core and satellite groups according to species' occupancy and abundance (Hanski, 1982). In order to model the random distribution (i.e., Poisson distribution) of species, the core-satellite approach uses the dispersion index (i.e., the ratio of variance to mean abundance), rather than setting an artificial cut-off for partitioning of species into ecologically meaningful groups (van der Gast et al., 2011). The core group is composed of non-randomly distributed species, which are abundant and widespread across time and space, while the satellite species are generally randomly distributed and have low abundance and narrow occupancy (Hanski, 1982; van der Gast et al., 2011). The core and satellite groups may have different responses to environmental conditions (Hou et al., 2019; Hu et al., 2017b; Jeanbille et al., 2016). Since abundant core and rare species are both important in term of ecosystem function and responding to environmental disturbances (Campbell et al., 2011; Jacquiod et al., 2018), particularly the former one (Hu et al., 2017b), an in-depth understanding of the micropollutant effects on these ecological meaningful groups is of great significance for predicting the response of prokaryotic communities to future environmental changes.

Here, 16S rRNA gene amplicon sequencing was used to characterize the structure of core/satellite sub-communities within PA and FL size-fractions, and their relationships with the micropollutants and physio-chemical parameters in a subtropical water reservoir were explored. Samples were collected horizontally (surface water) and vertically (depth-profile) from the water body. This sampling strategy allowed us to capture a relatively wide gradient of micropollutants and physico-chemical parameters in different seasons and depths under natural conditions. Specifically, following hypotheses were tested: (i) the micropollutants can significantly predict the variations in the composition of PA and FL prokaryotic communities at both horizontal and vertical scales in the reservoir; (ii) the core groups are more sensitive to physico-chemical factors and micropollutants than their satellite counterparts within PA or FL communities, as the non-randomly distributed taxa are more sensitive to the environmental drivers than the randomly distributed taxa (Barnes et al., 2016; Hou et al., 2019; Hu et al., 2017b); and (iii) prokaryotic taxa from PA and FL size-fractions are expected to have different types of associations with physico-chemical factors and micropollutants, as it is well-known that even for the same taxon, PA and FL taxa have different physiological and metabolic properties (Dang and Lovell, 2016).