Using a novel chain of models to mimic source aquifer depressurisation impacts across the Doongmabulla Springs Complex , Queensland, Australia

The Doongmabulla Springs Complex (DSC) is a cluster of groundwater-dependent wetlands located in central Queensland, Australia. The DSC has over 160 individual springs ranging from over 9 ha, with pools of water, down to vents less than 10 cm across, supporting individual grasses. The springs are home to a variety of plant species (many endemic) that are adapted to the unique physico-chemical parameters of the groundwater discharge on which they rely. Wetlands and scalds of the DSC support a listed Threatened Ecological Community of species dependent on this natural discharge of groundwater from underlying artesian aquifers of the Galilee Basin. The springs are protected under Australia’s State and Commonwealth Environmental legislation. 

Aquifer depressurisation due to mine dewatering occurring to the east of the springs has the potential to threaten spring biodiversity in future by reducing groundwater discharge and consequent wetland persistence. Assessing the likelihood and possible magnitude of these threats requires a multi-disciplinary modelling approach to address complex groundwater – surface water – ecosystem interactions: 1) define groundwater pressure change probabilities; 2) simulate likely wetland response; and 3) evaluate potential ecological impacts. Once such a modelling chain is in place, impact mitigation may be assessed, considering the effect of interventions at each stage of the chain.

To support the numerical modelling, extensive hydrological, ecological and physio-chemical data collection and analysis was undertaken to understand spring wetland area and species microhabitats and distributions over multiple scales and time frames, including seasonal and inter-annual variability. Generation of realistic and defensible conceptualisations for the springs is critical and is described in a companion paper (Cresswell, et al., these proceedings).

Regional numerical groundwater modelling (MODFLOW) generated potential groundwater pressure change responses relevant to the DSC source aquifers. Potential groundwater depressurisation over time at the individual spring locations was utilised in a wetland water balance model (GoldSim) to describe wetland persistence, size and seasonality. Water balance outputs were translated into spatial representations of predicted spring hydrology (TUFLOW), generating area and shape configurations that could be compared to historical wetland persistence data and then utilised to predict potential effects on suitable habitat for key flora species under a range of scenarios including mining-related groundwater drawdown and climate change using maximum entropy species distribution modelling (MaxEnt). This latter process relates known species’ occurrences to measurable variables that describe the environment (such as soil moisture, soil salinity and pH) to predict the presence or absence of a species at unsampled locations and under future groundwater drawdown scenarios. 

The results of the application of this novel chain of models have informed a revised impact assessment for the potential impacts of mine dewatering on the unique vegetation communities at the Doongmabulla Springs Complex and enables development of targeted mitigation approaches for individual wetlands and species.