Potential quality indicators detected in all three domains of life
The potential of microbial community members as indicators of ecosystem quality has previously been shown in contaminated sediments and marine environments (Acosta-González & Marqués, 2016; Urakawa et al., 2012), as well as rivers (Li et al., 2018). The strongest correlation to ecological status among the individual domains was observed in the ZOTUs from the bacterial community. Furthermore, correlations of interest fromArchaea and Eukaryotic ZOTUs were also observed (Figure 3, S6 and S7)), but these were weaker compared to the Bacteria . The latter correlations may be affected by their relatively low abundance in the bulk samples, and may be more significant when implemented in a network based approach targeting cross-domain interactions as indicators of quality, due to the complex nature of stream ecosystems (Battin et al., 2016).
Bacterial communities have previously been shown to reflect environmental changes relating to land use (Lear et al., 2013), as well as (anthropogenic) pollution (Acosta-González & Marqués, 2016; Mlejnková & Sovová, 2010; Urakawa et al., 2012). Organisms previously reported in relation to ecosystem health were found among the potential indicator OTUs of the prokaryotes. The genus Flavobacterium was observed among the ZOTUs associated to lower quality streams, and has previously been linked to anthropogenic activity in surface waters (Acosta-González & Marqués, 2016; Mlejnková & Sovová, 2010). A positive correlation was observed between the ecological status and a representative of the family Nitrosomonadaceae and the genusNitrospira . This correlation has previously been characterized in a marine ecosystem in which pollution was shown to induce significant changes in the nitrifying communities (Urakawa et al., 2012). Similar observations have described for increased abundances of the familyRhodobacteraceae exposed to oil pollution in sediments (Acosta-González & Marqués, 2016), however the indicator analysis in the present study yielded multiple ZOTUs representing this family, but with opposing tendencies. This highlights the importance of taxonomic resolution, as a family of bacteria can contain numerous of species with a wide range of functions and sensitivity to their environment.
Analysis of the microbial communities, and its vast diversity, present in stream bulk samples could provide an interesting supplement to existing metabarcoding approaches targeting biomarkers for eukaryotes such as the COI gene (Kuntke et al., 2020). Now, when applying a molecular approach for water quality assessment we are no longer limited to the composition of visible BQE´s, such as macroinvertebrates. The microbial community composition in the streams may often capture a better and more comprehensive and sensitive snapshot of the diversity in stream waters.