Divergence in forest structure and evapotranspiration over a chronosequence between Eucalyptus regnans and Acacia dealbata

Abstract

Abstract

Climate-induced changes in fire regimes in south east Australia have the potential to alter ecosystem and ecohydrological resilience in native Eucalyptus forests. A drier, hotter climate is triggering short-return, high-intensity, mega-fire events across Australia. In Victoria, over 189,000 ha of fire-sensitive, Eucalyptus regnans and Eucalyptus delegatensis forests have been burned two or more times within the past 18-years. Under these multiple burn conditions, fire-sensitive forests are highly vulnerable to ecological tipping points. Naturally occurring E. regnans is a dominant species in the higher rainfall regions of south east Australia. However, short interval fires (15-20 years) hinder the regeneration potential of this serotinous, obligate-seeder species, which then allows fast-growing, co-occurring understorey species such as Acacia dealbata to replace E. regnans as the dominant forest type. Replacing long-lived E. regnans forests (250-300 years) with short-lived A. dealbata (80-90 years) may have significant ecohydrological implications in water supply catchments.

A bottom-up, plot-based approach was used to measure Et in typical E. regnans and A. dealbata forests of various ages to develop empirical models of the hydrological response of forested water catchments to vegetation change. Field-based measurements of the various components of Et were taken from 10, 35 and 75/80-years old stands of each forest type. A forest inventory survey was carried out to quantify structural trajectories, including stand mean dbh, stand basal area and stand sapwood area in the two forest types. Transpiration was measured at the field using the heat ratio method together with micrometeorological conditions in the two forests. Throughfall, stem flow and evaporation from the forest floor were measured across the age sequence of each forest type. Throughout this research, eco-hydrologic processes along the chronosequence are modelled and then compared between the life cycles of E. regnans and A. dealbata forests.

Regeneration response to fire and ecohydrological properties of both forest types were similar during the initial stage of stand development up to age 10. However, stand structure, including mean stem diameter, basal area, stocking density and sapwood area, begins to diverge significantly between the two forest types after age 20. In both forest types, sapwood area reduces with stand age after age 20, but at a faster rate in A.dealbata. Once the forest structure starts to diverge, overstorey transpiration, overstorey Et and total Et of the two forest types also begin to diverge, driven primarily by divergence in sapwood area. In addition, mean sap velocity averaged for the 20, 35- and 75-80-years age classes was about 34% higher in E. regnans, although the difference was only statistically significant at age 20. This suggests that differences in sap velocity between the two forests also partly explained the divergence in overstorey Et between the two forest types. VPD is the strongest predictor of sap velocity in both forest types under non-water limited conditions. Daily sap velocity model for E. regnans could be further modified by accounting for forest age. The results provide a strong indication that after age 20, overstorey transpiration in Acacia-dominated forests is substantially lower than in the E. regnans forests they replace. Therefore, overstorey transpiration was the primary driver of differences in total Et between the two forest types, followed by differences in canopy interception. Soil evaporation contributed only 3% to the differences in Et between the two forest types.

Differences in Et partitioning between the two forest types imply a link between forest structure and the forest water budget. In senescing A. dealbata, understorey transpiration contribution of 29.8% to system Et was similar to that of overstorey transpiration (31.2%), indicating the understorey and overstorey contribute equally to total Et at the final stage of Acacia forests. This suggests that, after the Acacia life cycle finishes, the Et regime will transit into a new state that will be dominated by shrubby understorey species.

The findings of this research suggest that climate-driven high-frequency wildfires alter the composition and structure of E. regnans forest as a result of a change in the dominant overstorey species from E. regnans to A. dealbata. This species shift alters forest hydrological parameters, especially mean sap velocity and sapwood area, leading to changes in eco-hydrologic processes in the forest. These results highlight that species shift due to climate change can have important ecohydrological implications, resulting in evapotranspiration regime shift. Further, the present research suggests that climate-related species change from E. regnans to A. dealbata will alter the hydrologic response of water supply catchments. This type of eco-hydrologic response may be played out in many ecosystems in the future. By considering all studied changes in forest structure, evapotranspiration and water yield, this climate-induced species replacement is an ecologically significant vegetation change in the native E. regnans forests, reflecting extensive hydrological implications for the water supply catchments.