Determining potential future states of floodplain communities that are impacted by river regulation and modified flow regimes is a key requirement of making informed management decisions. Floodplain vegetation communities (e.g. river red gum forests, black box woodlands) and the maintenance of their condition via inundation are diverse in their requirements. Where modified hydrological conditions influence the duration, frequency, timing, and extent of inundation events, the response of vegetation communities and resulting condition will be spatiotemporally complex. Using the Floodplain Vegetation Condition Model (FVCM), a newly developed state and transition simulation model (STSM), we simulated futures of floodplain vegetation condition on the Murray River floodplain of south-eastern Australia in response to three scenarios of inundation (without development, observed flows and a management option).
The STSM approach provides a spatially explicit and dynamic analysis option for generating timeseries of vegetation condition in response to inundation sequences. From an assumed initial state of the floodplain (vegetation type and condition), the FVCM generated 124 annual projections of vegetation type and condition for gridded vegetation units across the Murray River floodplain. These simulations showed net increases to the annual area where vegetation improved in condition for both the ‘without development’ and ‘management option’ scenarios when compared to the observed inundation timeseries. However, strong spatially dependent effects were found for multiple community types concerning where on the floodplain benefits and impacts were modelled between inundation scenarios.
The FVCM provides a step-change in the ability to assess floodplain vegetation responses to changes in inundation. Current predictions align with expert opinion of the impacts of proposed inundation sequences though improve upon this by providing a spatially and temporally explicit framework for quantifying outcomes in ways that have not previously been possible. We close by presenting future directions of focus for developing models using this STSM framework.