Predicting environmental stressor levels with machine learning: a comparison between amplicon sequencing, metagenomics, and total RNA sequencing based on taxonomically assigned data.

Introduction: Microbes are increasingly (re)considered for environmental assessments because they are powerful indicators for the health of ecosystems. The complexity of microbial communities necessitates powerful novel tools to derive conclusions for environmental decision-makers, and machine learning is a promising option in that context. While amplicon sequencing is typically applied to assess microbial communities, metagenomics and total RNA sequencing (herein summarized as omics-based methods) can provide a more holistic picture of microbial biodiversity at sufficient sequencing depths. Despite this advantage, amplicon sequencing and omics-based methods have not yet been compared for taxonomy-based environmental assessments with machine learning. Methods: In this study, we applied 16S and ITS-2 sequencing, metagenomics, and total RNA sequencing to samples from a stream mesocosm experiment that investigated the impacts of two aquatic stressors, insecticide and increased fine sediment deposition, on stream biodiversity. We processed the data using similarity clustering and denoising (only applicable to amplicon sequencing) as well as multiple taxonomic levels, data types, feature selection, and machine learning algorithms and evaluated the stressor prediction performance of each generated model for a total of 1,536 evaluated combinations of taxonomic datasets and data-processing methods. Results: Sequencing and data-processing methods had a substantial impact on stressor prediction. While omics-based methods detected a higher diversity of taxa than amplicon sequencing, 16S sequencing outperformed all other sequencing methods in terms of stressor prediction based on the Matthews Correlation Coefficient. However, even the highest observed performance for 16S sequencing was still only moderate. Omics-based methods performed poorly overall, but this was likely due to insufficient sequencing depth. Data types had no impact on performance while feature selection significantly improved performance for omics-based methods but not for amplicon sequencing. Discussion: We conclude that amplicon sequencing might be a better candidate for machine-learning-based environmental stressor prediction than omics-based methods, but the latter require further research at higher sequencing depths to confirm this conclusion. More sampling could improve stressor prediction performance, and while this was not possible in the context of our study, thousands of sampling sites are monitored for routine environmental assessments, providing an ideal framework to further refine the approach for possible implementation in environmental diagnostics.

Perceived multiple stressor effects depend on sample size and stressor gradient length.

Multiple stressors are continuously deteriorating surface waters worldwide, posing many challenges for their conservation and restoration. Combined effect types of multiple stressors range from single-stressor dominance to complex interactions. Identifying prevalent combined effect types is critical for environmental management, as it helps to prioritise key stressors for mitigation. However, it remains unclear whether observed single and combined stressor effects reflect true ecological processes unbiased by sample size and length of stressor gradients. Therefore, we examined the role of sample size and stressor gradient lengths in 158 paired-stressor response cases with over 120,000 samples from rivers, lakes, transitional and marine ecosystems around the world. For each case, we split the overall stressor gradient into two partial gradients (lower and upper) and investigated associated changes in single and combined stressor effects. Sample size influenced the identified combined effect types, and stressor interactions were less likely for cases with fewer samples. After splitting gradients, 40 % of cases showed a change in combined effect type, 30 % no change, and 31 % showed a loss in stressor effects. These findings suggest that identified combined effect types may often be statistical artefacts rather than representing ecological processes. In 58 % of cases, we observed changes in stressor effect directions after the gradient split, suggesting unimodal stressor effects. In general, such non-linear responses were more pronounced for organisms at higher trophic levels. We conclude that observed multiple stressor effects are not solely determined by ecological processes, but also strongly depend on sampling design. Observed effects are likely to change when sample size and/or gradient length are modified. Our study highlights the need for improved monitoring programmes with sufficient sample size and stressor gradient coverage. Our findings emphasize the importance of adaptive management, as stress reduction measures or further ecosystem degradation may change multiple stressor-effect relationships, which will then require associated changes in management strategies.

Gaps in current Baltic Sea environmental monitoring - Science versus management perspectives.

Legislations and commitments regulate Baltic Sea status assessments and monitoring. These assessments suffer from monitoring gaps that need prioritization. We used three sources of information; scientific articles, project reports and a stakeholder survey to identify gaps in relation to requirements set by the HELCOM's Baltic Sea Action Plan, the Marine Strategy Framework Directive and the Water Framework Directive. The most frequently mentioned gap was that key requirements are not sufficiently monitored in space and time. Biodiversity monitoring was the category containing most gaps. However, whereas more than half of the gaps in reports related to biodiversity, scientific articles pointed out many gaps in the monitoring of pollution and water quality. An important finding was that the three sources differed notably with respect to which gaps were mentioned most often. Thus, conclusions about gap prioritization for management should be drawn after carefully considering the different viewpoints of scientists and stakeholders.

Polar Arctic Circle biomass enhances performance and stability of aerobic granular sludge systems operated under different temperatures.

Three bioreactors were inoculated with Polar Arctic Circle-activated sludge, started-up and operated for 150 days at 8, 15 and 26 °C. Removal performances and granular conformation were similar at steady-state, but higher stability from start-up was found when operating at 8 °C. Important changes in the eukaryotic and prokaryotic populations caused by operational temperature were observed, being fungi dominant at 8 °C and 15 °C, while that ciliated organisms were found at 26 °C. The qPCR results showed higher copies of bacteria, and nitrifiers and denitrifying bacteria at cold temperature. The emission of nitrous oxide was linked directly with temperature and the involved microorganisms. This study represents a proof of concept in performance, greenhouse gas emission, granular formation and the role of the Polar Arctic Circle microbial population in AGS technology under different temperatures with the aim to understand the effect of seasonal o daily changes for implementation of AGS at full-scale.

The future depends on what we do today - Projecting Europe's surface water quality into three different future scenarios.

There are infinite possible future scenarios reflecting the impacts of anthropogenic multiple stress on our planet. These impacts include changes in climate and land cover, to which aquatic ecosystems are especially vulnerable. To assess plausible developments of the future state of European surface waters, we considered two climate scenarios and three storylines describing land use, management and anthropogenic development ('Consensus', 'Techno' and 'Fragmented', which in terms of environmental protection represent best-, intermediate- and worst-case, respectively). Three lake and four river basins were selected, representing a spectrum of European conditions through a range of different human impacts and climatic, geographical and biological characteristics. Using process-based and empirical models, freshwater total nitrogen, total phosphorus and chlorophyll-a concentrations were projected for 2030 and 2060. Under current conditions, the water bodies mostly fail good ecological status. In future predictions for the Techno and Fragmented World, concentrations further increased, while concentrations generally declined for the Consensus World. Furthermore, impacts were more severe for rivers than for lakes. Main pressures identified were nutrient inputs from agriculture, land use change, inadequately managed water abstractions and climate change effects. While the basins in the Continental and Atlantic regions were primarily affected by land use changes, in the Mediterranean/Anatolian the main driver was climate change. The Boreal basins showed combined impacts of land use and climate change and clearly reflected the climate-induced future trend of agricultural activities shifting northward. The storylines showed positive effects on ecological status by classical mitigation measures in the Consensus World (e.g. riparian shading), technical improvements in the Techno World (e.g. increasing wastewater treatment efficiency) and agricultural extensification in the Fragmented World. Results emphasize the need for implementing targeted measures to reduce anthropogenic impacts and the importance of having differing levels of ambition for improving the future status of water bodies depending on the societal future to be expected.