Understanding when trees establish is critical for reconstructing recruitment histories, assessing population resilience, and evaluating water management decisions. In long-lived floodplain trees such as Eucalyptus camaldulensis (river red gum), reliable age estimation remains challenging because traditional approaches, including diameter-based estimates, dendrochronology and radiocarbon dating, are often imprecise, costly or invasive.
Here, we developed the first epigenetic clock for a hardwood tree species using DNA methylation data from known-age E. camaldulensis individuals spanning 1-154 years. Reduced representation bisulfite sequencing was applied to leaf, root and cambium tissues. We found that cambium provided the strongest age-associated methylation signal, while root and leaf tissues were less informative.
Age was predicted using elastic net regression models based on CpG, CHG and CHH methylation contexts. The most accurate model combined all three contexts and achieved strong predictive performance (R² = 0.83; RMSE = 18.5 years), substantially outperforming diameter at breast height (R² = 0.17; RMSE = 37.0 years). The final clock incorporated 79 age-informative cytosine sites and was primarily driven by progressive hypermethylation with age.
To realise its water-management potential beyond age determination, further refinement is required to improve the tool's precision and accessibility. As seed germination in river red gum is linked to flooding, the tool could identify specific flood events and subsequent climatic conditions associated with establishment if prediction error could be reduced to 1-2 years. Currently, the drivers of establishment remain poorly understood. Additional development is needed to provide a high-throughput, low-cost, accessible assay using DNA from cambium samples.
This tool uses non-invasive sampling and potentially provides a means of assessing whether river red gum populations are actively recruiting or persisting as ageing cohorts and evaluating environmental flow outcomes for floodplain forests. More broadly, it demonstrates the potential of epigenetic ageing as a powerful ecological monitoring tool for long-lived hardwood species.