Discussion

The ability to represent the entire flow regime in a simple yet scientifically robust series of six environmentally relevant flow metrics represents a significant simplification in from the superfluous manner in which the flow regime of seasonal rivers are represented and assessed. This simplification is evident both in regards to the number of flow metrics used (e.g. compared to 56 in VanLaarhoven and van der Wielen (2009) or 120 in Kennard et al. (2010)) and in regards to the complexity of the metrics themselves. The results of the correlation assessment show that there is enough differentiation between the metrics that they all individually add significant information to the overall narrative. While it is clear that the additional functions of the flow regime (e.g. Yarnell et al., 2020) could have been identified (e.g. overbank flows, variability within flow season etc.), it was determined that these would have added excess complication to the process without adding sufficient additional information to the assessment. This was considered especially important as a significant role of these metrics will be to communicate flow regime changes to community groups and other non-scientific stakeholders.
The metrics that been identified are all simple metrics reported as annual values. This provides benefits over previous complex metrics or metrics that are calculated over the entire time series. Time series metrics allow for more detailed investigations into changes through time, especially relevant when considering the implications of a changing climate. The impacts of climate change on flow regime are likely to be a significant focus both for these metrics and in the broader seasonal rivers field akin to Dhungel et al. (2016).
The use of set threshold values identified from outside the flow data itself was possible for the Barossa as there was only two locations where this was required to inform the process, with the remaining values portioned by catchment area. This provided a tangible link to the physical environment and a clear line between observed information, the metrics and associated EWRs. This may prove difficult for larger areas where information may not be available (e.g. permanent pool depth for pool volumes) or where the number of locations to be assessed is prohibitively large to do individual threshold assessments.
For the development of low flow thresholds, it was noted that the threshold values were close to the lower inflection point of the flow duration curve for the site. This could provide a method for identifying the low flow threshold from within the flow data. This will remove the direct link to a tangible on ground measure but should retain the overall functions defined for the low flow metric described in table XX. With the high correlation observed in other flow metric sets within the same flow season, it is highly likely that any flow threshold used close to the lower inflection point of the flow duration curve will be highly correlated with a value identified from the physical assessment used here. Therefore, they would be likely to convey the same information for flow regime assessment. The same argument could be made with the commence-to-flow threshold used for the assessment of the break-of-season. The identification of the threshold could be achieved by identifying the inflection point of the cumulative flow post April 1st and averaging the flow value over the time series.
Ultimately, the six flow metrics were able to provide a basis for the establishment of EWRs for the region based on the premise that the ecology of the region was stable between 1997 and 2016. This ability to not only characterise the flow regime, but the relationship of the environment to it, and quantify thresholds that can be used as management triggers or levers provides a powerful tool for managers and researchers alike but is simple enough to be interpreted by the broader public. By quantifying these thresholds based on a moving average rather than an individual year, it provides a more meaningful representation of the impacts on the riverine ecosystems in an inherently highly variable system.