The Queensland government regularly assesses the impact of water resource use on key ecological assets. One asset used to assess impacts is the Eastern snake-necked turtle (Chelodina longicollis). Modelling risks to C. longicollis requires a wetland commence-to-fill threshold (ML river discharge) and persistence (days) as input variables. Previously, these values have been coarsely estimated.
We developed an approach combining spatial analysis, time series analysis of Landsat imagery, river gauge flow observations, and post processing in Microsoft Excel to determine (a) commence-to-fill thresholds and (b) the persistence time of selected floodplain wetlands. A wetlands shape file (ArcMap) was sourced from Queensland Wetland Mapping (wetlandinfo.ehp.qld.gov.au/wetlands) and filtered to select >800 floodplain wetlands of interest within the Condamine-Balonne, Moonie and Dumaresq-Macintyre catchments. A newly optimised Landsat-based water mask was used to identify patterns of filling and drying within the selected wetlands for the period from 1988–2015. This resulted in a sequence of dates, listing each wetland as wet or dry, including the number of wet pixels in each wetland. Two Excel macros were then developed: Firstly, to correlate wetland filling dates with overbank flow events at the closest river gauge, and define a commence-to-fill range, and secondly to identify drying sequences to estimate wetland persistence. The approach successfully generated commence-to-fill thresholds and persistence times for wetlands throughout the three catchments, and was used to model risk profiles for C. longicollis under a number of water resource development scenarios.
The next step in this work is to conduct surveys/bathymetry and install depth loggers at a sub-set of wetlands to validate this method. Key wetland complexes providing breeding and foraging habitat for C. longicollis will be identified for this purpose.
This approach can potentially be applied to other catchments throughout Queensland and Australia to determine these parameters over broad spatial areas.