Machine learning techniques to characterize functional traits of plankton from image data.

Plankton imaging systems supported by automated classification and analysis have improved ecologists' ability to observe aquatic ecosystems. Today, we are on the cusp of reliably tracking plankton populations with a suite of lab-based and in situ tools, collecting imaging data at unprecedentedly fine spatial and temporal scales. But these data have potential well beyond examining the abundances of different taxa; the individual images themselves contain a wealth of information on functional traits. Here, we outline traits that could be measured from image data, suggest machine learning and computer vision approaches to extract functional trait information from the images, and discuss promising avenues for novel studies. The approaches we discuss are data agnostic and are broadly applicable to imagery of other aquatic or terrestrial organisms.

Comparison of markers for the monitoring of freshwater benthic biodiversity through DNA metabarcoding.

Metabarcoding of bulk or environmental DNA has great potential for biomonitoring of freshwater environments. However, successful application of metabarcoding to biodiversity monitoring requires universal primers with high taxonomic coverage that amplify highly variable, short metabarcodes with high taxonomic resolution. Moreover, reliable and extensive reference databases are essential to match the outcome of metabarcoding analyses with available taxonomy and biomonitoring indices. Benthic invertebrates, particularly insects, are key taxa for freshwater bioassessment. Nevertheless, few studies have so far assessed markers for metabarcoding of freshwater macrobenthos. Here we combined in silico and laboratory analyses to test the performance of different markers amplifying regions in the 18S rDNA (Euka02), 16S rDNA (Inse01) and COI (BF1_BR2-COI) genes, and developed an extensive database of benthic macroinvertebrates of France and Europe, with a particular focus on key insect orders (Ephemeroptera, Plecoptera and Trichoptera). Analyses on 1,514 individuals representing different taxa of benthic macroinvertebrates showed very different amplification success across primer combinations. The Euka02 marker showed the highest universality, while the Inse01 marker showed excellent performance for the amplification of insects. BF1_BR2-COI showed the highest resolution, while the resolution of Euka02 was often limited. By combining our data with GenBank information, we developed a curated database including sequences representing 822 genera. The heterogeneous performance of the different primers highlights the complexity in identifying the best markers, and advocates for the integration of multiple metabarcodes for a more comprehensive and accurate understanding of ecological impacts on freshwater biodiversity.

Morphological vs. DNA metabarcoding approaches for the evaluation of stream ecological status with benthic invertebrates: Testing different combinations of markers and strategies of data filtering.

Macroinvertebrate assemblages are the most common bioindicators used for stream biomonitoring, yet the standard approach exhibits several time-consuming steps, including the sorting and identification of organisms based on morphological criteria. In this study, we examined if DNA metabarcoding could be used as an efficient molecular-based alternative to the morphology-based monitoring of streams using macroinvertebrates. We compared results achieved with the standard morphological identification of organisms sampled in 18 sites located on 15 French wadeable streams to results obtained with the DNA metabarcoding identification of sorted bulk material of the same macroinvertebrate samples, using read numbers (expressed as relative frequencies) as a proxy for abundances. In particular, we evaluated how combining and filtering metabarcoding data obtained from three different markers (COI: BF1-BR2, 18S: Euka02 and 16S: Inse01) could improve the efficiency of bioassessment. In total, 140 taxa were identified based on morphological criteria, and 127 were identified based on DNA metabarcoding using the three markers, with an overlap of 99 taxa. The threshold values used for sequence filtering based on the "best identity" criterion and the number of reads had an effect on the assessment efficiency of data obtained with each marker. Compared to single marker results, combining data from different markers allowed us to improve the match between biotic index values obtained with the bulk DNA versus morphology-based approaches. Both approaches assigned the same ecological quality class to a majority (86%) of the site sampling events, highlighting both the efficiency of metabarcoding as a biomonitoring tool but also the need for further research to improve this efficiency.

Lessons from linking bio- and ecological traits to stoichiometric traits in stream macroinvertebrates.

Ecologists rely on various functional traits when investigating the functioning of ecological systems and its responses to global changes. Changing nutrient levels, for example, can affect taxa expressing different trait combinations in various ways, e.g., favoring small, fast-growing species under high phosphorus conditions. Stoichiometric traits, describing the elemental composition of organism body tissues, can help in understanding the mechanisms behind such functional shifts. So far, mainly life-history traits have been related to body stoichiometry (e.g., the growth rate hypothesis) on a limited number of taxa, and there is little knowledge of the general link between stoichiometric and other functional traits on a taxonomically large scale. Here, we highlight this link in the freshwater macroinvertebrates, testing predictions from underlying trait-based and Ecological Stoichiometry Theory (EST) in >200 taxa belonging to eight larger taxonomic groups. We applied a series of multivariate analyses on six of their stoichiometric traits (%C, %N, %P, C:N, C:P, and N:P) and 23 biological and ecological traits. We found significant relationships between stoichiometric traits and other types of traits when analyzing single-trait and multi-trait profiles. Patterns found within traits related to organism development or nutrient cycling were in line with our assumptions based on EST, e.g., traits describing predators were associated with high %N; traits suggesting a fast development (small maximum body size and high molting frequency) with high %P. Associations between ecological traits and body stoichiometry could be explained by the longitudinal stream gradient: Taxa preferring headwater habitats (i.e., high altitude, coarse substrate, and cold temperature) exhibited high %N and %P. Demonstrating the link between stoichiometric and both bio- and ecological traits on a large diversity of taxa underlines the potential of integrating stoichiometric traits into ecological analyses to improve our understanding of taxonomic and functional responses of communities-and ecosystems-to changing environmental conditions worldwide.

How diatom-, invertebrate- and fish-based diagnostic tools can support the ecological assessment of rivers in a multi-pressure context: Temporal trends over the past two decades in France.

The degradation of aquatic ecosystems, induced by worldwide intensification in the use of both land and aquatic resources, has highlighted the critical need for innovative methods allowing an objective quantification and ranking of anthropogenic pressure effects on aquatic organisms. Such diagnostic tools have a great potential for defining robust management responses to anthropogenic pressures. Our objective was to explore how the outputs of three diagnostic tools (based on benthic diatoms, macroinvertebrates and fishes) could be combined to (i) disentangle the temporal effects of multiple pressures over two decades and (ii) provide policy-relevant information for stream managers and decision makers. The diagnostic tools estimated, using taxonomy- and trait-based metrics, the impairment probabilities of biotic assemblages over time by different pressure categories, describing the alteration of water quality, hydromorphology and land use related to anthropogenic activities, in French streams (number of sites = 312). The main result shows that a large proportion of the time series exhibited no significant temporal patterns over the two decades (61.5% to 87.8%, depending on the used tests). Among time series exhibiting significant change, positive trends in impairment probabilities (i.e., degradation) were less frequent than negative ones, indicating a modest improvement in water quality at national scale over the study period. However, trends can be substantially different according to hydroecoregion and pressure category. The three biological compartments displayed convergent temporal responses according to the pressure category and regional context (e.g., lowland plains vs. mountains, pristine vs. agricultural regions). Altogether, this study proposes a unifying approach to integrate a vast amount of information in a single ecological diagnosis using an unparalleled database on natural and anthropized environments. Strengthening the synthesis of biological information provided by various biological compartments should be a priority before implementing evidence-based sustainable conservation and restoration actions.