Knowledge Graphs in Ecology: Linking Data, Concepts, and Systems

About: This repo contains part of the materials for a presentation given at the TU Munich on 8 May 2025 to students of ecology Abstract: Ecology often deals with complex, interconnected systems and increasingly with large, diverse datasets. Knowledge graphs offer a practical approach to organizing and linking such information - connecting ecological principles with species, environments, interactions and processes in a way that supports integration and reasoning. In this lecture, we introduce the concept of knowledge graphs - without assuming a technical background - and explore their growing relevance in ecology, particularly in areas like invasion biology, biotic interactions and ecosystem modeling. We will then look at some examples of how knowledge graphs are being used to structure ecological data and make it more discoverable and interoperable. Alongside this, we will also sketch out how graph-based approaches relate to other elements of the broader AI landscape. Throughout the lecture, attendees are invited to share their own ideas on the matter or to pose questions. The slide-based presentation was complemented by some practical examples from https://scholia.toolforge.org/ and https://globalbioticinteractions.org . Some additional starting points for exploring the topic (with some formatting lost upon transfor from the source etherpad): 10-min no-slide talk in German https://youtu.be/l2cYtCxipK4 featuring Strelitzia reginae Poster: Biodiversity in NFDI Knowledge Graphs. https://doi.org/10.5281/zenodo.10997504 Slides: 12 years of Wikidata - zooming in on its scholarly content. Zenodo. https://doi.org/10.5281/zenodo.13997230 https://w.wiki/D$n5 - list of ecological concepts https://w.wiki/E32X - a graph of their co-occurrence in the literature consider changing the LIMIT to 4000 consider adding in images if available https://w.wiki/E45C corresponding map list of co-occurring species formulas of co-occurring chemicals Phragmites australis Invasion biology and the success of social collaboration networks, with application to Wikipedia https://scholia.toolforge.org/work/Q56524142 A conceptual classification scheme of invasion science https://scholia.toolforge.org/work/Q130735544 Knowledge graph examples studying pollination in a given geographic area with a known population of flowers and pollinators start with one individual from one pollinator species let it visit one flower of one plant then another flower of the same plant then another plant eventually, it may come back to a flower it had already visited then it may go back to its nest bring in more individuals from this one pollinator species bring in more pollinator species consider microbes involved in pollination consider multiple pollination mechanisms consider using different methods (e.g. optical vs. acoustic vs. genetic vs. literature vs. mathematical modelling etc.) add disturbances like predators fires floods invasive species plant herbivore pollinator predator deforestation agricultural pesticides climate change do it at another time (of the day/ week/ year) and/ or location (lat/ long/ alt/ proximity to key biogeographic features/ ecological gradients) start building a network of relevant ecological concepts combine this with ways to generate research questions and hypotheses and track progress on them