Biome analysis and composition of Danish freshwater streams
One aim of the present study was to assess the merit of using high-throughput technologies for stream water quality assessment to increase sensitivity and simultaneously reduce labour intensity and cost. One of the steps in conventional bioassessment protocols is pre-treatment and sorting of the collected samples. Previous studies have developed generalised protocols for sample collection and handling for this purpose (Aylagas et al., 2016). Recently obtained results from a study involving metabarcoding of stream water bodies showed that direct sample homogenisation without pre-treatment was equally effective for molecular analysis compared to samples that had been washed and sorted prior to DNA extraction (Kuntke et al., 2020). The obtained sequencing results in the present study supports the direct use of samples without pre-treatment, as it was possible to obtain high quality sequences to a sufficient depth for detailed biome analysis from all three domains in 50 out of 53 samples. The omission of pre-treatment steps from the methodology reduces the sample preparation time and the extent of the biases associated with this process.
The obtained sequencing data was of high quality and shown to be of sufficient sequencing depth for high resolution analysis of domain-specific community profiles (Figure S1), with the majority of the captured diversity originating from the bacterial communities present in the samples (Figure 1a). This was as expected, as sediment material makes up a large fraction of the sample, and the sediment microbiome is one of the most complex biomes that has been studied (Battin, Besemer, Bengtsson, Romani, & Packmann, 2016). The methodology of analysing multiple domains for ecosystem quality studies has previously been attempted in a study involving the use of multiple biomarkers (16S rRNA , 18S rRNA and COI genes) and sequencing technologies aimed to develop a framework for metasystematic analysis of bulk samples containing arthropods (Gibson et al., 2014). The sequencing data obtained in the present study reveal a measured diversity for all domains that is multiple times higher than previously reported. Furthermore, the novel approach of using a single primer set to capture a comprehensive picture of the stream water biome simultaneously improves data handling and comparability, and highlights the quality and convenience of the chosen methodology.
A number of known sediment associated bacteria were detected abundantly across samples of all qualities (Figure S2-S4), including the generaRhodoferax , Nitrospira and the familyComamonadaceae , as well as representatives of the archaeal familyNitrosospaeraceae (Battin et al., 2016). The composition of plant life in forest ecosystems has previously been used to predict soil microbiome profiles in grasslands, with moderate success (Leff et al., 2018), which highlights the relationship of the soil microbes with the flora and fauna of the ecosystem they are present in. Furthermore, microbial community monitoring has previously been applied in the tracking of bioremediation after oil spills in marine sediment environments (Acosta-González & Marqués, 2016; Urakawa et al., 2012). Moreover, the quality of the soil and its microbes has previously been shown to function as a predictor for ecosystem health based on land usage (Lear et al., 2013), and it can thus be suggested that measuring the microbiome as well as the fauna and flora composition could improve assessment resolution for quality index studies in stream waters. Not surprisingly, a number of taxonomic groups containing organisms previously used as indicator species for bioassessment were also detected (Skriver et al., 2000), including representatives ofDiptera , Caenogastropoda , Mollusca andColeoptera .