Assessing ecosystem state with environmental DNA of aquatic insects

Description

Data associated to the publication "Environmental DNA sequences from aquatic insects indicate freshwater ecosystem state" Authors: Flurin Leugger, Martina Lüthi, Meret Jucker, Virginie Marques, Sarah Thurnheer, Zacharias Kontarakis and Loïc Pellissier Contact: flurinleugger@gmail.com loic.pellissier@usys.ethz.ch Project description: In this project, we aimed to compare kick-sampling-derived taxa lists and ecosystem state classifications (IBCH: Indece biologique Suisse) with eDNA metabarcoding and CRISPR-Dx. Additionally, we calculated the association with environmental variables and ecosystem state. We retrieved the kick-sampling-derived data from Haberthür et al. (2021) and Keck et al. (2023). We collected eDNA samples in 36 sites in the Rhine catchment in Switzerland (lowland) which are used for the national monitoring of freshwater in Summer 2024. We extracted the samples in the clean lab at ETH using the protocol described in Leugger et al. (2025). We used the 16S Ins primer (Elbrecht et al. 2016) for metabarcoding and for CRISPR analysis. To select indicator eDNA metabarcoding sequences, we applied a machine learning approach combining LASSO and rpart decision trees. For CRISPR-Dx, we used the rules of Leski et al. (2023) to in silicon predict detections based on the metabarcoding sequences and used the same machine learning approach as for the indicator metabarcoding eDNA sequences (see publication for more details). The data for the environmental variables is from the SWECO25 database (https://zenodo.org/communities/sweco25/records?q=&l=list&p=1&s=10&sort=newest; Külling et al. 2024).

Folders: catchments: Catchments in Switzerland, downloaded from https://data.geo.admin.ch/browser/index.html#/collections/ch.bafu.wasser-einzugsgebietsgliederung/items/wasser-einzugsgebietsgliederung?.language=en&.asset=asset-wasser-einzugsgebietsgliederung_2056-gpkg-zip The subfolder upstream_catchments includes the catchment for each buffer size tested.

comparison_classification_methods: File with the ecosystem state classification based on kick-sampling, indicator eDNA metabarcoding sequences (using the rpart decision tree) and CRISPR-Dx (using predicted, lab and retrained rpart trees).

eDNA: Folder metabarcoding: Raw and cleaned metabarcoding data (ASVs) per environmental extract (EVE), including taxa information from Ecotag reference data base. Additionally, the rpart trees for indicator sequences only from Ephemeroptera and all EPT are provided. Columns which might not be self explanatory taxon: identified taxon rank: rank of identified taxon nb_reads: read numbers EVS and metabarcoding: key columns to link with metadata files ASV_cleaned_by_site.csv: Read count by site (after cleaning) ASV_raw_by_site.csv: Read count by site (before cleaning), column 'NC' is for the negative controls sequence_count_per_site.csv: Comparison of reads per site before and after cleaning/filtering. 'NC' refers again to Negative Controls. Folder CRISPR: crispr_by_site.csv: Lab-based detection of the guide per site. Maximum number of detections is 6, as each site had two environmental extracts (EVE), and the PCR replicates were combined in 3 pools each. GRN refers to lab internal numbering of CRISPR-Dx assays. detection_matrix_Ephemeroptera_28_site.csv: Predicted detection based on eDNA metabarcoding sequences and the model of Leski et al 2023 per site. rpart_model_Ephemeroptera.rds: best performing rpart model for ecosystem state classification using the predicted CRISPR-Dx detections rpart_model_retrained.rds: best performing rpart model for ecosystem state classification using the actual lab-based CRISPR-Dx detections (thus "retrained" based on actual detections). IBCH: IBCH_classification_filtered_and_extended.csv: includes IBCH_cat (ecosystem category from kick-sampling with IBCH 2019 categorization). IBCH_taxa_kick-sampling.csv: contains taxa detect per site (site_code as column name to link with metadata or eDNA data) kicknet_Trend_IBCH_sites.csv: taxa detected by site, raw data

metadata: Folder containing files with overview of site names, extract number and coordinates. Columns named "metabarcoding" (and "EVS") are keys to link rows from the different tables. SWECO25: Folder containing raster stack from SWECO25 data (Külling et al. 2024) used in the article and a table with the extracted variable value per site/catchment with the different upstream buffer sizes tested. site_code: Code for the sampling site, see also metadata to link to EVS Temperature_avg: annual mean temperature in upstream catchment [°C] Precip_annual: annual precipitation [mm] Population_density: mean population density in 25m cells Traffic_noise: noise levels EVI, NDVI and LAI: each standard deviation (sd) and average (avg) value per upstream catchment Natural rivers: proportion of upstream river length which is classified as natural Little-disturbed_rivers: proportion of upstream river length which is classified as little disturbed Heavily-disturbed_rivers: proportion of upstream river length which is classified as heavily disturbed unnatural_rivers: proportion of upstream river length which is classified as unnatural culverted_rivers: proportion of upstream river length which is culverted Forest_edges: proportion of forest edges in upstream catchment Forest: proportion of forest in upstream catchment Crops: proportion of crops in upstream catchment Built_environment: proportion of built environment in upstream catchment buffer: buffer width for upstream catchment area [m] catchment_area_m2: upstream catchment area in square meters

References: Elbrecht, V., Taberlet, P., Dejean, T., Valentini, A., Usseglio-Polatera, P., Beisel, J. N., Coissac, E., Boyer, F., & Leese, F. (2016). Testing the potential of a ribosomal 16S marker for DNA metabarcoding of insects. PeerJ, 2016(4). https://doi.org/10.7717/peerj.1966 Haberthür, M. (2021). Ergebnisse der 3. Erhebung NAWA-Trend Los 2, Makrozoobenthos. https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.bafu.admin.ch/dam/bafu/de/dokumente/wasser/externe-studien-berichte/nawa_trend_biologie_2019-teil_makrozoobenthos.pdf.download.pdf/nawa_trend_biologie_2019-teil_makrozoobenthos.pdf&ved=2ahUKEwjI5crG-IaPAxUF_gIHHQisBZEQFnoECBgQAQ&usg=AOvVaw1QL3GRZdTllPsHLQ6hQtlP Külling, N., Adde, A., Fopp, F., Schweiger, A. K., Broennimann, O., Rey, P. L., Giuliani, G., Goicolea, T., Petitpierre, B., Zimmermann, N. E., Pellissier, L., Altermatt, F., Lehmann, A., & Guisan, A. (2024). SWECO25: a cross-thematic raster database for ecological research in Switzerland. Scientific Data, 11(1). https://doi.org/10.1038/s41597-023-02899-1 Leski, T. A., Spangler, J. R., Wang, Z., Schultzhaus, Z., Taitt, C. R., Dean, S. N., & Stenger, D. A. (2023). Machine learning for design of degenerate Cas13a crRNAs using lassa virus as a model of highly variable RNA target. Scientific Reports, 13(1), 6506. https://doi.org/10.1038/s41598-023-33494-4 Leugger, F., Lüthi, M., Schmidlin, M., Kontarakis, Z., & Pellissier, L. (2025). Rapid field-based detection of a threatened and elusive species with environmental DNA and CRISPR-Dx. Global Ecology and Conservation, 59, e03518. https://doi.org/10.1016/j.gecco.2025.e03518