Data and code from: Data integration advances reproductive phenology research across temporal, spatial, and taxonomic scales

Climate change is altering plant reproductive phenology; however, a scarcity of long-term, systematic monitoring hinders our ability to quantify and predict these responses in many parts of the world. We addressed this gap by demonstrating how data integration can be used to produce a synthesised record of reproductive phenology observations (flowering and fruiting) that spans longer time periods, larger spatial scales, and includes more species than any single source alone. Using Australian subtropical rainforest trees as a case study, we integrated reproductive phenology observations from both common data sources—published datasets, herbarium specimens, and citizen science records—and previously untapped expert botanical knowledge, including private photographic collections, field notes, and seed collections. Data integration yielded 110,657 records of flowers or fruits from 915 species (representing half of all subtropical rainforest tree species) spanning 255 years (1770-2025). We found that different data sources provided unique information across temporal, spatial, and taxonomic dimensions. Herbarium specimens provided the longest taxonomic coverage, while citizen science contributed the most recent observations. Critically, 197 species (21.5 %) were represented from only a single source, including 154 species represented solely by herbarium specimens and 46 species in expert botanist collections. While 46.6 % of species had fewer than 50 observations, for many species, these represent the only available historical phenology data. This integrated dataset may be the only available resource for establishing pre-industrial baselines for the reproductive phenology of Australian subtropical rainforest trees. This would not have been possible without the engagement and contributions of the local botanical community, which greatly expanded the research capacity beyond conventional data sources.