Dynamics of Eucalyptus largiflorens growth and water use in response to modified watertable and flooding regimes on a saline floodplain

Abstract

Listen

Translate article

Reduced flooding and raised watertables have caused increased soil salinity and die-back of native forests on the floodplains of the lower River Murray of south Australia. Proposed management options include increasing flood frequency by regulating flows from upstream storages, and groundwater pumping to lower the watertable. This paper uses a soil-vegetation-atmosphere-transfer model (WAVES) to evaluate the impact of these proposals on soil salinisation processes and vegetation growth (black box, Eucalyptus largiflorens) for soils with different hydraulic properties. The changes in canopy leaf mass and plant available soil water were simulated for the period 1970–1994 using historical daily climate and river level records. The river level records were used to reconstruct the flooding and watertable history of sites where tree water use studies were conducted to calibrate the model. Then the watertable depth and/or flooding frequency was modified and the changes in the canopy leaf mass and soil water availability relative to the historical simulation were evaluated. The simulations suggest that, with the present watertable and flooding regime, very large floods (e.g. 205 days as in 1974–1976) are needed to sustain tree cover on the higher parts of the floodplain where die-back is most severe. They also indicate that soil hydraulic properties have a large influence on the magnitude and time scale of the growth response of salt stressed vegetation to floods and salt accumulation. Infrequently flooded vegetation exhibiting die-back was predicted to increase its canopy leaf area for up to 12 years following the large floods of 1974 and 1976, at sites where the soil was relatively permeable and groundwater highly saline (EC = 55 dS m −1). The changes in canopy leaf area in response to the floods was predicted to be relatively small on sites with heavier clay soils. The growth response of the vegetation to a long term lowering of watertable depth by 1 m was greater than that induced by the small potential increase in flooding frequency which is feasible given the current water storage limitations. The simulations predict that changes in the average annual soil water availability which arise from flood events and soil salinisation, drive a long term cycle in the annual average transpiration rate per unit leaf area suggesting the soil-plant-climate system is adjusting towards a hydrological equilibrium but is not in equilibrium. The proposed management options may control die-back in parts of the floodplain with more slowly salinising heavy clay soils and lower salinity groundwater but are unlikely to prevent die-back on relatively permeable soils with high salinity groundwater. However, they may assist vegetation survival between the long duration flood events which appear to be essential to sustain tree cover on the higher floodplain. The management options need to be evaluated further at the floodplain scale using the understanding from site specific conditions to test simple approaches which can be linked to a geographic information system (GIS) of the floodplain.