A spatio-temporal analysis of canopy dynamics and intra-stand competition in a riparian forest, south-eastern Australia

Abstract Widespread high stem density stands of riparian River Red Gum ( Eucalyptus camaldulensis Dehnh.) (RRG) result from land management legacies in inland riverine ecosystems of Australia. These stands generally exhibit reduced structural complexity with many slender stems and few large trees. Higher competition for water among populations of RRG in dense stands may affect stand-level canopy condition, particularly when water is scarce. We investigated whether stem density and water availability affected canopy condition dynamics in RRG forests of Murray Valley National Park, New South Wales. We collected eight years of satellite derived Foliage Projective Cover (FPC) data (Landsat 5 TM (2008–2011) and Landsat 8 OLI (2013–2016)) encompassing some of the driest and wettest years on record in south-eastern Australia. We used Generalised Linear Mixed Effects Modelling to investigate the drivers of RRG canopy condition at the plot-level. Canopy condition dynamics were driven by different sets of variables under different climatic phases, with water availability at the plot and regional scale, as well as climate, being the primary drivers of trends in RRG canopy condition. Stem density was not found to be a significant predictor of overall RRG stand canopy condition, although results suggested that individuals in higher density stands have lower canopy condition on average. Live basal area (LBA) was positively correlated with FPC in all climatic phases, and stands with high LBA tended to have a greater abundance of large trees than low LBA stands. Research efforts should be made towards integrating site-specific tree size class distributions with remotely sensed metrics to guide management for biodiversity. Findings from this study provide a baseline understanding against which ecological restoration actions can be evaluated and a foundation upon which an ongoing monitoring program can be developed.