Periodic Reporting for period 1 - PROTECT (A machine learning conservation apPROach to evaluaTE extinCTion risk in freshwater biodiversity)
Reporting period: 2024-09-01 to 2026-08-31
Biodiversity loss is accelerating worldwide, yet the true magnitude of this crisis remains uncertain, as many species have not yet been assessed and therefore remain unprotected. Accurate assessments of extinction risk are essential to guide conservation priorities, but the current IUCN Red List of Threatened Species is taxonomically biased towards well-studied groups. Freshwater invertebrates, despite their crucial ecological roles, are among the most neglected. The PROTECT project aimed to address this gap by developing a predictive framework to approximate extinction risk assessments for freshwater biodiversity using machine learning tools. Focusing on the family Hydrobiidae—one of the most diverse and threatened groups of freshwater snails—the project sought to identify ecological, evolutionary, and genetic predictors of species vulnerability. By integrating molecular, morphological, and ecological data, PROTECT provided complementary insights into extinction risk patterns in poorly known taxa and contributed to advancing large-scale conservation research.
The results demonstrated the feasibility of using machine learning approaches (IUCNN) to predict extinction risk in freshwater gastropods. Among the non-evaluated Hydrobiidae species, a large proportion were classified into one of the threatened categories. Habitat type, taxonomic placement and extent of occupancy were identified as the most influential factors shaping these classifications. Range reconstructions based on historical records and field surveys conducted during the project revealed that the distribution of several Iberian species is shrinking, likely due to harsh climatic conditions, water pollution and the presence of the invasive species Potamopyrgus antipodarum. Furthermore, the project highlighted the value of museum collections as genomic resources for conservation research.
The results demonstrated the feasibility of using machine learning approaches (IUCNN) to predict extinction risk in freshwater gastropods. Among the non-evaluated Hydrobiidae species, a large proportion were classified into one of the threatened categories. Habitat type, taxonomic placement and extent of occupancy were identified as the most influential factors shaping these classifications. Range reconstructions based on historical records and field surveys conducted during the project revealed that the distribution of several Iberian species is shrinking, likely due to harsh climatic conditions, water pollution and the presence of the invasive species Potamopyrgus antipodarum. Furthermore, the project highlighted the value of museum collections as genomic resources for conservation research.
Throughout the project, the researcher carried out a comprehensive programme combining training, data collection, modelling and dissemination activities. The work was organised in several research modules, each addressing different aspects of extinction risk prediction in freshwater gastropods.
The project compiled an extensive dataset for over 800 hydrobiid species, including geographical, ecological and evolutionary information, which served to train and validate automated extinction risk assessments using the IUCNN machine learning framework. Fieldwork conducted in central and eastern Spain provided fresh specimens for genomic analyses and new occurrence data for range-size reconstruction. This led to the sequencing and assembly of a high-quality reference genome, while 22 lower-coverage genomes from seven species were successfully obtained from museum-preserved material, demonstrating the feasibility of using historical collections as valuable sources of genetic material for conservation studies.
Training activities strengthened the researcher’s expertise in neural network modelling, bioinformatics, and IUCN Red List assessment protocols, as well as in fieldwork safety and conservation management. The project also contributed to national conservation efforts by updating the Libro Rojo de los Invertebrados Amenazados de España for 11 hydrobiid species and by participating in collaborative reports and publications with the IUCN Specialist Groups.
Dissemination and outreach activities increased the project’s visibility among both scientific and general audiences. The problem, methodology and results were presented at international and national conferences, published through institutional channels, and used to educate student teachers and the public.
Although the project concluded earlier than initially planned, all major scientific and training objectives were achieved, laying a strong foundation for future research on freshwater biodiversity conservation in Spain and beyond.
The project compiled an extensive dataset for over 800 hydrobiid species, including geographical, ecological and evolutionary information, which served to train and validate automated extinction risk assessments using the IUCNN machine learning framework. Fieldwork conducted in central and eastern Spain provided fresh specimens for genomic analyses and new occurrence data for range-size reconstruction. This led to the sequencing and assembly of a high-quality reference genome, while 22 lower-coverage genomes from seven species were successfully obtained from museum-preserved material, demonstrating the feasibility of using historical collections as valuable sources of genetic material for conservation studies.
Training activities strengthened the researcher’s expertise in neural network modelling, bioinformatics, and IUCN Red List assessment protocols, as well as in fieldwork safety and conservation management. The project also contributed to national conservation efforts by updating the Libro Rojo de los Invertebrados Amenazados de España for 11 hydrobiid species and by participating in collaborative reports and publications with the IUCN Specialist Groups.
Dissemination and outreach activities increased the project’s visibility among both scientific and general audiences. The problem, methodology and results were presented at international and national conferences, published through institutional channels, and used to educate student teachers and the public.
Although the project concluded earlier than initially planned, all major scientific and training objectives were achieved, laying a strong foundation for future research on freshwater biodiversity conservation in Spain and beyond.
The project advanced the state of the art in biodiversity conservation by integrating machine learning and genomic approaches to predict extinction risk in one of the most diverse and understudied groups of freshwater organisms, the Hydrobiidae. By applying and extending the IUCNN framework, the project not only demonstrated the feasibility of automated extinction risk assessments but also developed new analytical modules incorporating explainable artificial intelligence tools to explore how ecological and evolutionary variables influence species classification within IUCN threat categories. These advances make IUCNN a more transparent and versatile tool for large-scale conservation assessments. Furthermore, the development of genomic protocols for small, non-model invertebrates and the successful use of museum-preserved material as a source of DNA represent methodological innovations with applications extending to other taxa.
At the societal level, the project directly contributed to conservation policy and management by updating the Libro Rojo de los Invertebrados Amenazados de España (Red Book of Threatened Invertebrates of Spain) and informing IUCN Red List evaluations of multiple species. By focusing on spring ecosystems, the research also highlighted the link between biodiversity protection and the sustainability of freshwater resources essential to rural communities. Overall, the project delivered new data, tools, and conservation outputs that strengthen the scientific and institutional capacity for safeguarding freshwater biodiversity.
At the societal level, the project directly contributed to conservation policy and management by updating the Libro Rojo de los Invertebrados Amenazados de España (Red Book of Threatened Invertebrates of Spain) and informing IUCN Red List evaluations of multiple species. By focusing on spring ecosystems, the research also highlighted the link between biodiversity protection and the sustainability of freshwater resources essential to rural communities. Overall, the project delivered new data, tools, and conservation outputs that strengthen the scientific and institutional capacity for safeguarding freshwater biodiversity.