e-CITIZEN SCIENCE COMMUNICATING: ARTIFICIAL INTELLIGENCE FEEDBACK BETWEEN OBSERVATIONS AND STAKEHOLDER INPUTS FOR CROSS-BORDER RESILIENT FISH MIGRATION IN EUROPE

Introduction

Migratory fishes connect rivers, estuaries and seas while crossing administrative and policy boundaries; however, European management remains fragmented and feedback between emerging science, local observations and operational decisions is often weak. HORIZON Project FISH2SUSTAIN addresses this challenge through life-cycle, multi-stressor research and bioclimatic modelling (Nekrasova et al. 2024) on three model migration strategies: long-distance catadromous European eel (Anguilla anguilla), river-sea-river anadromous Atlantic salmon (Salmo salar), and coastal-marine European seabass (Dicentrarchus labrax). Citizen science plays an important role in ensuring sustainable nature conservation (Bonney et al. 2014) and adaptive feedback between observations and stakeholder inputs for transboundary sustainable fish migration in Europe.

Materials and Methods

The Project FISH2SUSTAIN develops a stakeholder-centric machine-learning e-Citizen Science environment (Figure 1) based on mobile application with AI-Stakeholder co-creation and interaction, targeted e-Education, anonymous One-Click geolocated photo-reporting, real-time data collection and knowledge gamificated testing for different stakeholder MARG (Pupins et al. 2025). The present study aims to design and validate this e-Citizen Science AI-component as a bidirectional, evidence-aware interface linking citizen/stakeholder observations with the validated scientific outputs of the whole project and returning interpretable, uncertainty-aware adaptive feedback.

Results

As this research is implemented within the FISH2SUSTAIN work plan, results are formulated as expected and testable outputs rather than completed findings. The primary result will be an operational e-Citizen Science mobile application that shortens the pathway from a local observation of a migrating fish to a quality-controlled, spatially explicit and FAIR-compatible record. The AI assistant will make complex project evidence accessible without presenting predictions as observations or suppressing uncertainty. Validated citizen records will provide an additional stream for detecting migration events, spatial gaps, bottlenecks and unusual environmental conditions and will feed the adaptive feedback architecture. Performance will be quantified against expert-labelled observations and independent monitoring, while AI-stakeholder co-creation will test whether feedback improves understanding, reporting quality and sustained e-Engagement. The target architecture is a closed observation-to-action loop (Figure 1), in which user inputs improve monitoring and model evaluation and validated Project knowledge is returned to users in accessible form.

Discussion

The proposed system extends e-Ctizen Science from one-way data acquisition to continuous science-policy-society co-investigation. Its novelty lies in coupling low-burden geolocated reporting with expert-governed AI, project-wide mechanistic evidence, bioclimatic modelling, and a Digital Twin. This design can strengthen cross-border comparability and enable local observations for sustainable cross-border management (Pupins et al. 2026).

Figure 1. FISH2SUSTAIN stakeholder-centric machine-learning e-Citizen Science environment and AI Pipeline Loop.

Acknowledgment

This research is developed within the Horizon Europe FISH2SUSTAIN project "From Life-Cycle Science to Cross-Border Action for Resilient Fish Migration in Europe". We thank INTERREG project UrbUmbrella, Nature Conservation Agency for the cooperation.

References

Bonney R., Shirk J.L., Phillips T.B., Wiggins A., Ballard H.L., Miller-Rushing A.J., Parrish J.K. (2014) Next steps for citizen science. Science, 343, 1436-1437. https://doi.org/10.1126/science.1251554.

Nekrasova O., Pupins M., Tytar V., Fedorenko L., Potrokhov O., Skute A., Ceirāns A., Theissinger K., Georges J.-Y. (2024) Assessing prospects of integrating asian carp polyculture in Europe: a nature-based solution under climate change? – Fishes, 9, 148. https://doi.org/10.3390/fishes9040148.

Pupins M., Khvesko V., Čeirāns A., Garkajs A., Skute A. (2025) Latgales Zoo-Wetland AI-based mobile App. https://doi.org/10.13140/RG.2.2.31042.64961.

Pupins M., Nekrasova O., Čeirāns A., Arte I., Skute A. (2026) Joint Sustainable AI-based Urban Wetland Management System. Cross-Border Wetland socio-ecological e-Governance guidance. – Daugavpils University, 45 p. https://doi.org/10.13140/RG.2.2.16036.03205.