How hydropower induced changes in temperature, quality and quantity of water influences structure and functioning of riverine biofilms

Abstract

River ecosystems are hot-spots of biodiversity, hosting many unique species and providing key ecosystem services. Nowadays these ecosystems are threatened by multiple stressors and human impact is the paramount reason. Flow alterations due to hydropower development and damming are among the most important sources of impact in river ecosystems, having detrimental effects on their fragile ecology. This thesis will look at a biological component so far neglected by the research addressing the effects of flow alteration on freshwater organisms, namely the prokaryotes. Prokaryotes are key players in the biogeochemical processes of river ecosystems. Most prominently, they have a variety of adaptive mechanisms which makes them capable of colonizing almost every ecological niche. Within river systems, prokaryotes are mainly found in biological matrices colonized by multiple species, also called biofilms. Biofilms were the focus of this thesis (paper I-II-III), both attached to inorganic and organic substrates. In paper I a combination of quantitative (Catalyzed Reported Deposition -Fluorescence in situ Hybridization) and semi-quantitative (16S rRNA metabarcoding) molecular techniques was used to screen the biofilm communities living in Norwegian river systems affected by multiple stressors. Strong environmental gradients were driving the prokaryotic community composition, revealing the importance of features such as pH and nutrient loads for the community dynamic. In this study a new methodology to detect prokaryotic indicator taxa by using metabarcoding data is presented. Taxa showing highest variance and prevalence across the sampling sites were selected as candidate bioindicators. Specific relationships between the candidate bioindicators (at different taxonomic levels) and the main environmental drivers were detected. The impact of hydropower and dams on the prokaryotic community structure and functioning was assessed in paper II by using 16S rRNA gene metabarcoding and litter bags with different mesh sized to analyze the organic matter breakdown in impacted and unimpacted reaches of ten Norwegian river systems. The biofilm community structure showed no clear pattern when looking at upstream downstream gradients, while the organic matter breakdown seemed to be significantly influenced by the presence of the dams, with the downstream reaches showing higher decomposition. To further analyze the effects of flow regime alterations on the prokaryotic communities and the biogeochemical processes they carry out, in paper III, we designed a mesocosm experiment where sixteen stainless steel flumes were used to implement four different flow regimes scenarios typical of managed river systems: homogenized flow regimes; flow regimes affected by agricultural management; drought treatments with water level reduced by 60%; and natural flow regime as control. Major effects of flow alterations on the prokaryotic community structure and functioning were detected. Drought treatments showed the strongest impact on both the prokaryotic assemblages and the decomposition of organic matter, which was significantly higher compared to the other treatments. Significant differences in the beta diversity were found among the prokaryotic communities from the four treatments, however we did not always find differences in the organic matter breakdown. In paper IV we employed machine learning algorithms to detect prokaryotic bioindicators for water quality classification in the Danube river by using an existing dataset of 16S metabarcoding for planktonic microbes. We found that different sequence variants combinations for the microbial community in upstream sites were yielding accurate prediction for water quality classification at a given downstream site. A number of sequence variants were found to be functionally redundant and, as such, mutually exclusive in terms of predictive information. These mutually exclusive sequence variants were at times phylogenetically related, and thus we can conclude that specific taxonomic groups have a generalizable predictive power and would need to be tested in other river basins to evaluate their performance in predicting water quality. The results from this thesis showed significant changes in both structure and functioning of prokaryotic communities in relation to flow alterations due to river management, damming, water abstraction and hydropower developments. However, it was also evident that effects on prokaryotic communities were very context dependent with catchment specific differences sometime overruling effects of flow changes in terms of community composition. The clear response found in the experimental study of flow alterations supports the contention that natural variability is high when undertaking field studies and is masking singular effect of flow. Overall, the findings in this thesis suggest a potential and widely overlooked impact of flow alterations on riverine microbial communities. Shifts in the prokaryotic assemblages might have repercussions on the energy transfer within the freshwater food webs, impairing crucial ecosystem processes such as carbon and nitrogen cycling, which is extremely relevant considering future climate change scenarios. By showing that identification of prokaryotic indicators for water quality classification is possible using various statistical tools, this thesis suggests that prokaryotic indicators can be implemented in current monitoring networks. The usefulness of these indicators is further exemplified providing new insights in the ecological status of river ecosystems impacted by human induced stressors such as hydro-morphological alterations due to hydropower and damming.