A benchmark dataset for machine learning in ecotoxicology.

The use of machine learning for predicting ecotoxicological outcomes is promising, but underutilized. The curation of data with informative features requires both expertise in machine learning as well as a strong biological and ecotoxicological background, which we consider a barrier of entry for this kind of research. Additionally, model performances can only be compared across studies when the same dataset, cleaning, and splittings were used. Therefore, we provide ADORE, an extensive and well-described dataset on acute aquatic toxicity in three relevant taxonomic groups (fish, crustaceans, and algae). The core dataset describes ecotoxicological experiments and is expanded with phylogenetic and species-specific data on the species as well as chemical properties and molecular representations. Apart from challenging other researchers to try and achieve the best model performances across the whole dataset, we propose specific relevant challenges on subsets of the data and include datasets and splittings corresponding to each of these challenge as well as in-depth characterization and discussion of train-test splitting approaches.

Effects of microplastics mixed with natural particles on Daphnia magna populations.

The toxicity of microplastics on Daphnia magna as a key model for freshwater zooplankton is well described. While several studies predict population-level effects based on short-term, individual-level responses, only very few have validated these predictions experimentally. Thus, we exposed D. magna populations to irregular polystyrene microplastics and diatomite as natural particle (both ≤63 µm) over 50 days. We used mixtures of both particle types at fixed particle concentrations (50,000 particles mL-1) and recorded the effects on overall population size and structure, the size of the individual animals, and resting egg production. Particle exposure adversely affected the population size and structure and induced resting egg production. The terminal population size was 28-42 % lower in exposed compared to control populations. Interestingly, mixtures containing diatomite induced stronger effects than microplastics alone, highlighting that natural particles are not per se less toxic than microplastics. Our results demonstrate that an exposure to synthetic and natural particles has negative population-level effects on zooplankton. Understanding the mixture toxicity of microplastics and natural particles is important given that aquatic organisms will experience exposure to both. Just as for chemical pollutants, better knowledge of such joint effects is essential to fully understand the environmental impacts of complex particle mixtures. ENVIRONMENTAL IMPLICATIONS: While microplastics are commonly considered hazardous based on individual-level effects, there is a dearth of information on how they affect populations. Since the latter is key for understanding the environmental impacts of microplastics, we investigated how particle exposures affect the population size and structure of Daphnia magna. In addition, we used mixtures of microplastics and natural particles because neither occurs alone in nature and joint effects can be expected in an environmentally realistic scenario. We show that such mixtures adversely affect daphnid populations and highlight that population-level and mixture-toxicity designs are one important step towards more environmental realism in microplastics research.

Incubation in Wastewater Reduces the Multigenerational Effects of Microplastics in Daphnia magna.

The aging of microplastics in the environment changes their physicochemical properties. While this may affect their toxicity, comparative data on the effects of aged compared to pristine microplastics are scarce. One of those aging processes is the sorption of chemicals, which has mainly been studied for individual pollutants present in marine ecosystems. To investigate how the sorption of a complex mixture of freshwater pollutants affects the toxicity of microplastics, we incubated irregular polystyrene particles (≤63 µm) in either wastewater or ultrapure water. We exposed Daphnia magna to these aged microplastics and their pristine counterparts (80, 400, 2000, and 10,000 particles mL-1) over four generations using food limitation as an additional, environmentally realistic stressor. Both particle types affect the survival, reproduction, adult and neonate body lengths, and growth. An exposure to pristine microplastics results in the extinction of the third generation of daphnids. In contrast, wastewater-incubated particles induced a lower mortality. The incubation with wastewater does not change the microplastics' size, surface charge, and structure. Consistent with the literature, we assume that the adsorption of dissolved organic matter is a key aging process reducing the toxicity of microplastics. Consequently, toxicity testing using pristine microplastics may overestimate the effects of plastic particles in nature.

Microplastics but not natural particles induce multigenerational effects in Daphnia magna.

Several studies have investigated the effects of nano- and microplastics on daphnids as key freshwater species. However, while information is abundant on the acute toxicity of plastic beads, little is known regarding the multigenerational effects of irregular microplastics. In addition, a comparison of microplastics to naturally occurring particles is missing. Therefore, we investigated the effects of irregular, secondary polystyrene microplastics (<63 µm) and kaolin as natural reference particle on the survival, reproduction, and growth of Daphnia magna over four generations under food-limited conditions. Additionally, we tested the sensitivity of the neonates in each generation to a reference compound as a proxy for offspring fitness. Exposure to high concentrations of microplastics (10,000 and 2000 particles mL-1) reduced daphnid survival, resulting in extinction within one and four generations, respectively. Microplastics also affected reproduction and growth. Importantly, an exposure to kaolin at similar concentrations did not induce negative effects. The sensitivity of neonates to potassium dichromate was not affected by maternal exposure to particles. Taken together, our study demonstrates that irregular PS particles are more toxic than natural kaolin in daphnids exposed over multiple generations under food limitation. Thus, our work builds towards more realistic exposure scenarios needed to better understand the impacts of microplastics on zooplankton.

Laboratory-to-field extrapolation: Increase in carbamazepine toxicity in a higher tier, multiple-stress experiment.

The toxicity and environmental risk of chemicals, such as the antiepileptic drug carbamazepine (CBZ), is commonly assessed using standardized laboratory tests and laboratory-to-field extrapolation. To investigate the toxicity of CBZ to aquatic key organisms in a more complex and environmentally relevant scenario, we conducted a 32-day multiple-stress experiment in artificial indoor streams. We exposed the non-biting midge Chironomus riparius, the blackworm Lumbriculus variegatus, and the New Zealand mud snail Potamopyrgus antipodarum to 80 and 400 µg CBZ/L in six artificial indoor streams. In addition to hydraulic stress, species' interaction, and low organic content in the sediment, organisms were co-exposed to the herbicide terbutryn (TBY) as a second chemical stressor at a concentration of 6 µg/L. The exposure to CBZ under multiple stress conditions resulted in a 10- to more than 25-fold higher toxicity in C. riparius and P. antipodarum when compared to a previous, standardized laboratory experiment. The co-exposure to TBY enhanced the adverse effects of CBZ on snails (reduced production of embryos). This effect was additive as the single exposure to TBY also reduced the reproduction of snails, most likely through the reduction of biofilm biomass. The emergence of C. riparius declined at a CBZ concentration of 400 µg/L (without the co-exposure to TBY) and at 80 µg/L in combination with TBY. The difference in sensitivity between laboratory and indoor stream experiments is indicative of a potential underestimation of risk when toxicity data are extrapolated to field conditions. The present results suggest the inclusion of non-chemical and chemical stressors in environmental hazard and risk assessments.

When Fluorescence Is not a Particle: The Tissue Translocation of Microplastics in Daphnia magna Seems an Artifact.

Previous research reported the translocation of nano- and microplastics from the gastrointestinal tract to tissues in Daphnia magna, most prominently of fluorescent polystyrene beads to lipid droplets. For particles >300 nm, such transfer is biologically implausible as the peritrophic membrane retains these in the daphnid gut. We used confocal laser scanning microscopy to study tissue transfer applying the setup from a previous study (neonates exposed to 20 and 1000 nm polystyrene beads at 2 µg L-1 for 4 and 24 h), the same setup with a fructose-based clearing, and a setup with a 1000-fold higher concentration (2 mg L-1 ). We used passive sampling to investigate whether the beads leach the fluorescent dye. Although the 1000 nm beads were visible in the gut at both exposure concentrations, the 20 nm beads were detectable at 2 mg L-1 only. At this concentration, we observed fluorescence in lipid droplets in daphnids exposed to both particle types. However, this did not colocalize with the 1000 nm beads, which remained visible in the gut. We further confirmed the leaching of the fluorescent dye using a passive sampler, a method that can also be applied in future studies. In summary, we cannot replicate the original study but demonstrate that the fluorescence in the lipid droplets of D. magna results from leaching of the dye. Thus, the use of fluorescence as a surrogate for particles can lead to artifacts in uptake and translocation studies. This highlights the need to confirm the stability of the fluorescence label or to localize particles using alternative methods. Environ Toxicol Chem 2019;38:1495-1503. © 2019 SETAC OPEN PRACTICES: The present study has earned Open Data/Materials badges for making publicly available the digitally shareable data necessary to reproduce the reported results. Learn more about the Open Practices badges from the Center for Open Science: https://osf.io/tvyxz/wiki.

The Effects of Natural and Anthropogenic Microparticles on Individual Fitness in Daphnia magna.

Concerns are being raised that microplastic pollution can have detrimental effects on the feeding of aquatic invertebrates, including zooplankton. Both small plastic fragments (microplastics, MPs) produced by degradation of larger plastic waste (secondary MPs; SMPs) and microscopic plastic spheres used in cosmetic products and industry (primary MPs; PMPs) are ubiquitously present in the environment. However, despite the fact that most environmental MPs consist of weathered plastic debris with irregular shape and broad size distribution, experimental studies of organism responses to MP exposure have largely used uniformly sized spherical PMPs. Therefore, effects observed for PMPs in such experiments may not be representative for MP-effects in situ. Moreover, invertebrate filter-feeders are generally well adapted to the presence of refractory material in seston, which questions the potential of MPs at environmentally relevant concentrations to measurably affect digestion in these organisms. Here, we compared responses to MPs (PMPs and SMPs) and naturally occurring particles (kaolin clay) using the cladoceran Daphnia magna as a model organism. We manipulated food levels (0.4 and 9 µg C mL-1) and MP or kaolin contribution to the feeding suspension (<1 to 74%) and evaluated effects of MPs and kaolin on food uptake, growth, reproductive capacity of the daphnids, and maternal effects on offspring survival and feeding. Exposure to SMPs caused elevated mortality, increased inter-brood period and decreased reproduction albeit only at high MP levels in the feeding suspension (74% by particle count). No such effects were observed in either PMP or kaolin treatments. In daphnids exposed to any particle type at the low algal concentration, individual growth decreased by ~15%. By contrast, positive growth response to all particle types was observed at the high algal concentration with 17%, 54% and 40% increase for kaolin, PMP and SMP, respectively. When test particles comprised 22% in the feeding suspension, both MP types decreased food intake by 30%, while kaolin had no effect. Moreover, SMPs were found to homoaggregate in a concentration-dependent manner, which resulted in a 77% decrease of the ingested SMPs compared to PMPs. To better understand MP-processing in the gut, gut passage time (GPT) and evacuation rate of MPs were also assayed. SMPs and PMPs differed in their effects on daphnids; moreover, the particle effects were dependent on the MP: algae ratio in the suspension. When the MP contribution to the particle abundance in the medium changed from 1 to 4%, GPT for daphnids exposed to SMPs increased 2-fold. Our results suggest that MPs and, in particular, SMPs, have a greater capacity to negatively affect feeding in D. magna compared to naturally occurring mineral particles of similar size. Moreover, grazer responses observed in experiments with PMPs cannot be extrapolated to the field where SMPs dominate, because of the greater effects caused by the latter.