Abstract
Humanity's reliance on clean water and the ecosystem services provided makes identifying efficient and effective ways to assess the ecological condition of streams ever more important. We used high throughput sequencing of the 16S rRNA region to explore relationships between stream microbial communities, environmental attributes, and assessments of stream ecological condition. Bacteria and archaea in microbial community samples collected from the water column and from stream sediments during spring and summer were used to replicate standard assessments of ecological condition performed with benthic macroinvertebrate collections via the Benthic Index of Biotic Integrity (BIBI). Microbe-based condition assessments were generated at different levels of taxonomic resolution from phylum to OTU (Operational Taxonomic Units) in order to understand appropriate levels of taxonomic aggregation. Stream sediment microbial communities from both spring and summer were much better than the water column at replicating BIBI condition assessment results. Accuracies were as high as 100% on training data used to build the models and up to 80% on validation data used to assess predictions. Assessments using all OTUs usually had the highest accuracy on training data, but were lower on validation data due to overfitting. In contrast, assessments at the order-level had similar performance accuracy for validation data, and a reduced subset of orders also performed well, suggesting the method could be generalized to other watersheds. Subsets of the important orders responded similarly to environmental gradients compared to the entire community, where strong shifts in community structure occurred for known aquatic stressors such as pH, dissolved organic carbon, and nitrate nitrogen. The results suggest the stream microbes may be useful for assessing the ecological condition of streams and especially useful for stream restorations where many eukaryotic taxa have been eliminated due to prior degradation and are unable to recolonize.
Highlights
Microbes are crucial to biogeochemical processes and cycles at all spatial scales, from intra-individual to global
The classification accuracies of the Discriminant Analysis of Principal Components (DAPC) model compared favorably to accuracies of Benthic Index of Biotic Integrity (BIBI) scores calculated on replicate benthic macroinvertebrate samples collected by the MBSS program
For field replicate samples collected in 55 randomly selected reaches by the MBSS program, we applied the same number of categories and cutoffs to the BIBI as we did in assessing the DAPC model accuracies
Summary
Microbes are crucial to biogeochemical processes and cycles at all spatial scales, from intra-individual to global. They are key to the maintenance of biodiversity and ecosystem function, and identifying factors influencing microbe distributions and diversity may be important for understanding ecosystems across natural and human-influenced gradients. The diversity and composition of the microbial community may be useful as direct or indirect proxies for assessing ecosystem condition and health and have recently been used to predict hydrologic function in large Arctic rivers [2]. Microbe-based applications for ecosystem monitoring and assessment are exciting for classifying the condition of freshwater streams because a small, collected sample could augment or replace the substantial efforts required for traditional methods based on eukaryotes [e.g., 3]. Analogs exist for the human body and other animals, where the overall status of an individual can be inferred by the composition of its microbiome as indicators of age, disease, or other stressors [4,5,6,7]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
