EcoDL 2025 aims to explore the integration of AI, digital libraries, and FAIR data principles in ecological research to improve knowledge synthesis and predictive modeling. Ecology's complexity and data heterogeneity present challenges in generalization, requiring advanced computational tools for structured knowledge representation, search, and decision support. We invite researchers from ecology, AI, and digital information systems to discuss AI-driven data synthesis, semantic search, causal inference, and machine learning applications in biodiversity and conservation. Through interdisciplinary contributions, EcoDL 2025 seeks to foster innovation in ecological informatics, supporting open science and advancing digital methods for ecological research and environmental sustainability.

Ecology is data-intensive, yet its complexity and context-dependence make it difficult to generalize findings across environmental conditions, species interactions, and spatial scales [1]. Synthesizing knowledge from heterogeneous studies remains a significant challenge [2]. Addressing these issues requires robust data management strategies, AI-driven information systems, and digital libraries embracing the FAIR principles [3] to enhance ecological knowledge synthesis and refine predictive models [4].

Ecological data encompasses a wide range of information, including species occurrence records, remote sensing imagery, climate measurements, genetic sequencing data, habitat maps, and long-term ecosystem monitoring datasets. These diverse data sources contribute to understanding patterns of species distributions, ecosystem processes, and environmental changes over time.

Digital information systems and AI have become important tools for managing ecological complexity. Digital libraries store and retrieve vast, heterogeneous datasets, while semantic search, knowledge graphs, and AI-based retrieval extract critical insights. However, issues such as interoperability, fragmented data sources, and the difficulty of mapping evidence to theoretical frameworks persist. AI systems that integrate machine learning and natural language processing can assist in causal inference, hypothesis testing, and predictive modeling—approaches that are valuable for advancing ecological research and supporting conservation efforts.

This workshop aims to bring together experts from ecology and digital libraries, with contributions from AI research where applicable, to address the challenges of synthesizing diverse ecological data and improving predictive models. The integration of multi-scale data — often fragmented across different sources — poses a significant barrier to developing generalizable ecological insights. Recognizing the essential role of digital libraries in research infrastructure, we seek to explore AI-driven systems and FAIR data principles for enhancing ecological methodologies. While interdisciplinary contributions from fields such as climate science, biodiversity conservation, ecosystem restoration, environmental science, and geography are welcome, the primary focus remains on advancing ecological research through digital and computational tools. By bridging theory and practice, this workshop supports TPDL’s mission to drive digital innovation in ecological research and applications.

The EcoDL 2025 Workshop is co-located with the TPDL 2025 Conference (The 29th International Conference on Theory and Practice of Digital Libraries, https://tpdl2025.github.io/) and will take place on September 23, 2025, in Tampere, Finland. It is tentatively planned as a full-day event. Details about the workshop program will be announced in August.