Introduction

The flow regime of a river, defined as the magnitude, timing, duration and frequency of flow events, has long been established as one of the key driving influences on riverine ecosystems (Naiman et al., 2008, Poff et al., 1997). Previous efforts to define the flow regime has provided a multitude of flow metrics that can be used to quantify nearly any aspect of the flow regime (Kennard et al., 2010, VanLaarhoven and van der Wielen, 2009). The altering of the flow regime through water resource development (e.g. dams) has been linked to declines in riverine ecosystem condition, most commonly associated with reductions in flows (Poff et al., 2007, Poff et al., 2010, Poff and Zimmerman, 2010). It has been further identified that some aspects of the flow regime have a disproportionate impact on riverine ecosystems, in particular for intermittent rivers (Datry et al., 2014a, Datry et al., 2014b). It follows that the assessment of impacts of changes to the flow regime of intermittent rivers can be summarised using a suite of key, ecologically relevant flow metrics, and that these changes in flow metrics can be meaningfully translated into a level of potential ecological impact.
Seasonal rivers, as a class of intermittent rivers, are defined as rivers that have a predictable cease to flow period. The vast majority of these follow the typical Mediterranean flow regime of cease to flow periods over the hot and dry summer months and a predictable flowing period over the cooler, wetter winter months (Kennard et al., 2010). These seasonal rivers are the dominant type of river found in southern Australia and are often heavily developed for their water resources to provide water for agricultural activities (Malerba et al., 2021). The level of development in some of these regions has led to the need to regulate and manage the capture, extraction and use of water through legislation (e.g. Landscape South Australia Act, Government of South Australia, 2019).
The sustainable management of these water resources depends on archiving an equitable balance between users, while keeping the total requirements for water below a sustainable level. Generally these requirements are broken down into social, economic, cultural and environmental water requirements. For water planning in South Australia, there is a significant emphasis placed on environmental water requirements as a priority user for water resources. Generally, environmental water requirements (EWRs) are defined as the water regime needed to sustain the ecological values of aquatic ecosystems, including their process and biological diversity, at a low level of risk (VanLaarhoven and van der Wielen, 2009). Previous assessments of environmental water needs have used various different methods for describing the EWRs for seasonal rivers, ranging from qualitative descriptions (e.g. Barossa, Natural Resources AMLR, 2009) through to complex series of 50+ flow metrics (e.g. Western and Eastern Mt Lofty Ranges, Natural Resources AMLR, 2013, Natural Resources SAMDB, 2013).
The notion of establishing flow assessment methods is not novel, with hundreds of methods developed and reported (Tharme, 2003). The most recent attempt at quantifying EWRs for seasonal rivers in South Australia provides a pass/fail style threshold for 56 flow metrics. Subsequent assessment of these flow metrics identified significant correlation between some of the flow metrics, especially between flow metrics within the same flow season (Maxwell et al., 2015). This correlation was indicative of potential bias within the metrics, especially when using percentage passing/failing as the overall indicator, as metrics would fail in batches rather than independently. There are also a considerable number of metrics using non-zero daily flows which has been identified as an issue when comparing datasets. The changing proportion of data between different scenarios included in a non-zero flow assessment results in metrics that are not directly comparable.
The goal was to define a suite flow metrics that were representative of all aspects of the flow regime that were ecologically relevant and not correlated with each other. This is in a similar vain to the functional flows approach used by Yarnell et al. (2020) with a focus both on function (e.g. intermittency, high flows) and season (e.g. low flow season, transition seasons). The number of metrics was determined by a combination of consideration of the aspects of the flow regime that were deemed to be important along with assessment of previous suites of metrics. A PCA of the metrics (normalised) from VanLaarhoven and van der Wielen (2009) suggested that the first six principal components contained ~91% of the variance of the overall dataset. While not directly relatable to individual metrics, six was chosen as the number of metrics to develop.
Previous assessments of flow data from seasonal rivers in South Australia has shown that there is significant correlation between flow metrics per season (Maxwell et al., 2015) so single metrics per season were identified to avoid this correlation. Based on assessment of previous suites of flow metrics used (e.g. Kennard et al., 2010, Poff and Zimmerman, 2010, VanLaarhoven and van der Wielen, 2009, Yarnell et al., 2020), the six key areas of interest were identified as (1) intermittency, (2) low flows, (3) break of season, (4) spring flows, (5) medium flows and, (6) high flows. The links between these flow metrics and riverine ecosystem function is summarised in Table 1.
Table 1: summary of ecological functions identified for each of the flow regime areas summarised from conceptual understanding of seasonal rivers and literature review.