Storage and display
All collected datasets (directly downloadable as tabular files), the bibtex file with all references, all reports and all kinetic bioaccumulation metric estimates are publicly available on Zenodo17. An rmarkdown file18,19 was created to build the overview table with information collected from the name of the dataset and from the dataset itself (e.g., column headers, number of data, number of replicates), as well as from the bibtex file. The R package DT was additionally used20 to combine all collected information in a user-friendly manner including a convenient search tool, and the rmarkdown file was finally compiled19 in HTML format for display to the user in packs of 10 lines by default. In such a way, each new dataset added into the repository will compile the rmarkdown file automatically for update.
In parallel, the database can also be accessed directly via http://lbbe-shiny.univ-lyon1.fr/mosaic-bioacc/data/database/TK_database.html, or from MOSAICbioacc clicking on the “More scientific TK data” button. An example of the output of the overview table is shown in Fig. 2, while the full table is provided in the supplementary information (Table S2). The collected raw TK data of the database consist in the time-course of several types of chemical substances bioaccumulated in various species via different exposure routes.
Screenshot of the first page of the overview table of the database available from MOSAICbioacc.
Datasets overview
Each dataset is summarized by:
the file name (raw data directly downloadable by clicking on the file name, in text or CSV format),
the genus of the tested organism,
the category of the organism (e.g., aquatic, terrestrial, etc.),
the tested chemical substance,
the duration of the accumulation phase,
the tested exposure routes (e.g., water, sediment, food, pore water),
the total number of observations in the dataset (plus the number of replicate(s) in brackets),
the kinetic bioaccumulation metric median value with its 95% uncertainty interval,
the report which contains all the outputs from MOSAICbioacc (in PDF format),
the link to the reference or the source of the data,
some additional comments (e.g., lipid fraction, growth, biotransformation, if exposure was done for chemical mixtures or not, if total radioactivity was used or not, etc.).
A summary of all datasets is presented in Table 1. Genus were separated in 12 categories: aquatic invertebrates (n = 105), fish (n = 42), insects (n = 17), aquatic worms (n = 10), terrestrial worms (n = 16), seawater sponges (n = 2), seawater plants (n = 1), aquatic algae (n = 1), terrestrial invertebrates (n = 1), vertebrates other than fish (n = 4), marine invertebrates (n = 8), and heterotrichea (n = 4). The most represented genus in the database are Gammarus (aquatic invertebrate, n = 43) and Daphnia (aquatic invertebrate, n = 27), followed by Oncorhynchus (fish, n = 15), genus that are classically used in ecotoxicological experiments. Recommended genus by OECD guidelines for bioaccumulation tests are Eisenia and Enchytraeus for terrestrial organisms (OECD 317)21, and Tubifex or Lumbriculus for aquatic invertebrates exposed to sediment (OECD 315)22; some datasets for these specific species are available in the database (n = 24).
Chemical substances were divided in 10 classes following at the best the nomenclature used in Standartox23: pesticides (n = 71), hydrocarbons (n = 37), metals (n = 20), nanoparticules (n = 23), polychlorobiphenyls (PCB, n = 22), flame retardants (brominated or chlorinated, n = 8), pharmaceutical products (n = 14), PFAS (n = 7), octyphenol (n = 2) and other (n = 7). Among all datasets, the majority of bioaccumulation tests were performed via spiked water (n = 137). Besides, 34 datasets account for biotransformation processes, considering from 1 to 8 metabolites.
According to ECHA (2017)2, BCF below 1,000 means that the chemical substance is not bioaccumulative, whereas one ranging between 1,000 and 5,000 corresponds to a bioaccumulative chemical substance: low bioaccumulative if BCF ∈]1,000; 2,000]; mid-bioaccumulative if BCF ∈]2,000; 5,000]. If BCF is >5000, the chemical substance is classified as very bioaccumulative. These ranges are reported in Table 1, where BCF median estimates are >5000 for 25 datasets, indicating a very bioaccumulative capacity of the corresponding chemical substances for the corresponding genus. Concerning BSAF and BMF estimates, their value must be compared to threshold 1. A median BSAF estimate >1 indicates that the corresponding chemical substance can bioaccumulate from soil or sediment into organisms at the base of the non-aquatic food chain24,25; a median BMF estimate >1 indicates that the corresponding chemical substance can biomagnify in the trophic relationship under consideration26. In the database, 16 datasets in 36 led to BSAF >1, for genus Eisenia (n = 2), Enchytraeus (n = 6), Gallus (n = 1), Lumbriculus (n = 2), Metaphire (n = 2), Physa (n = 1), Radix (n = 2)), while 8 datasets in 38 led to BMF >1, for genus Gallus (n = 1), Oncorhynchus (n = 5) and Perca (n = 2). On an ecotoxicological point of view, the highest BCF estimates were obtained for genus Culex and Sialis exposed to chlorpyrifos due to a very low estimate of the elimination rate, for genus Gammarus and Calanus exposed to hydrocarbons, and several aquatic invertebrates exposed to pesticides, especially chlorpyrifos (n = 4), attesting to the potential high bioaccumulation capacity and high risk of toxicity associated with this chemical substance for aquatic organisms. Overall, aquatic invertebrates seem to be the most sensitive category of organisms in terms of bioaccumulation of chemical substances representing 20 in the 25 datasets with a BCF estimates >5000.
Source: Ecology - nature.com
