INTRODUCTION
Intermittent streams are closely tied to local climate and geology and reflect the hydrologic cycle within their watershed (Baker et al. 2004). Intermittent streams are naturally dynamic systems that often follow periodic cycles of drying and wetting (Arthington et al. 2014). Intermittent streams may fully dry, but often retain isolated pools during low flow periods (Busch et al. 2020). Intermittent streams are common and critical components of aquatic ecosystems that occur across most biomes and are estimated to account for up to 60% of the length of all defined stream channels (Larned et al. 2010; Datry et al. 2014; Costigan et al. 2016; Messager et al. 2021). Due to climate change, intermittent streams will likely become more widely distributed and many will likely dry for longer periods of time as precipitation patterns shift (Larned et al. 2010; Jaeger et al. 2014; Costigan et al. 2016; Zipper et al. 2021; Datry et al. 2022).
Despite drying for parts of the year, intermittent streams support diverse and complex community assemblages (Meyer et al. 2007, Datry et al. 2016), are valuable components of aquatic ecosystems (Colvin et al. 2019), play a role in the persistence of fishes by providing seasonal habitats (Vander Vorste et al. 2020), transport nutrients and material (von Schiller et al. 2017) and support terrestrial biodiversity (Sánchez-Montoya et al. 2022). The availability of habitats and resources provided by intermittent streams is variable within and between water years, influencing community dynamics (Fausch and Bramblett 1991) and driving patterns of community structure (Olden and Kennard 2010). Moreover, intermittent streams provide valuable habitat for native species and contain variable community compositions compared to perennial streams (Rogosh and Olden 2019). For example, wet season habitats can provide complementary habitats for different life stages of aquatic organisms (Labbe and Fausch 2000) and increase connectivity and therefore colonization rates (Franssen et al. 2006). In contrasts, during summer drying, fish communities shift as intermittent systems reduce to isolated pools, whether by reducing as conditions harshen, or by as available habitat shrinks (Capone and Kushlan 1991; Pires et al. 1999; Taylor and Warren 2001; Hopper et al. 2020). The longer a no-flow period persists, the more potential exists for long-lasting impacts on the aquatic community, as no-flow periods represent a significant disturbance (Poff et al. 1997).
While both high- and low-flow conditions are integral aspects of the flow regime (Poff et al. 1997), low-flow conditions such as no-flow duration, dry-down period and no-flow timing drive intermittent stream ecology (Poff et al. 1997; Olden and Poff 2003; Zipper et al. 2021). Ecologically relevant low-flow conditions can also be described by simple classifications such as whether or not a stream fully dries or dries to isolated pools, and what proportion of time the stream spends in each state (Gallart et al. 2017). The temporal and spatial components of low flow hydrology are closely linked and require understanding of both physical and anthropogenic factors (Smakhtin 2001). Climatic patterns, ground water, water table interactions, geology, and watershed shape, structure and size are all key components influencing low flow periods (Smakhtin 2001, Snelder and Biggs 2002, Hammond et al. 2020). At relatively fine spatial scales, geology and land use characteristics may be stronger drivers of flow patterns than climate (Snelder and Biggs 2002, Ssegane et al. 2012). Geology can impact stream flow through groundwater storage, recharge characteristics and flow capabilities (Mwakalila 2003). Anthropogenic alterations may also influence stream intermittency patterns, with changes in land cover influencing timing and magnitude of flow patterns, especially in regard to flow variability and low flow periods (Ficklin et al. 2018; Datry et al. 2022).
A wide range of metrics have been developed and used to characterize the flow regime and compare variations in hydrologic patterns across space and time. Many of these metrics are simple calculations that correspond to ecologically important aspects of flow, are often highly correlated (Smakhtin 2001), and serve to represent the complexity of a riverine system (D’Ambrosio et al. 2017). Olden and Poff (2003) examined 171 hydrologic indices to evaluate informativeness, redundancy and ecological importance and concluded intermittent flow regimes are best characterized by unique indices in comparison to more stable systems. The time scale at which flow patterns exist is highly variable, from hours to years, potentially requiring long-term observation to fully quantify variability (Poff et al. 1997; Richter et al. 1997; Leasure et al. 2016). Many streams have established, long-term monitoring in the form of gauging stations, such as those maintained by the United States Geological Survey (USGS), providing accurate discharge data (Falcone 2011). However, stream gauges are often not suitable for highly dynamic stream systems, and alternate data collection techniques have met varying levels of success (Blasch et al. 2004). Additionally, most intermittent streams are not gauged (Eng et al. 2015), limiting readily available long-term flow data.
Evaluating flow conditions in the absence of stream gauge data can be analytically and conceptually difficult, but is important given the increased interest in intermittent streams (Sivapalan 2003; Datry et al. 2016; Zipper et al. 2021) and in understanding how hydrology and hydrologic indices predict ecological patterns and processes (Clausen and Biggs 2000; Meyer et al. 2007; Booth and Konrad 2017). One approach to quantify dry periods of intermittent streams is to use water temperature data, which can be subject to interpretation (Blasch et al. 2004; Sowder and Steel 2012). Another technique is to model streamflow using nearby gauged streams, a data intensive methodology that may miss inherent variation between stream segments (Black 1972; Mwakalila 2003; Zimmer et al. 2020). Spatial scale is an integral aspect of flow modeling, as conditions between nearby streams can be highly variable, often as a result of local precipitation (Yuan 2013). The Soil and Water Assessment Tool (SWAT) is a widely used model for evaluating streamflow characteristics and potential changes in flow regime, predicting streamflow patterns and soil or nutrient transport from various processes including climate, hydrology, vegetation, management and precipitation (Cibin et al 2014; Gassman et al. 2014; Jajarmizadeh et al. 2015). In ungauged systems, SWAT models may not represent flow patterns accurately (Qi et al. 2020) and can be difficult to calibrate and validate due to a lack of empirical data (Sivalapalan et al. 2003; Beven and Smith 2015). Regardless, there is a need for low-cost techniques to monitor ungauged intermittent streams (Blasch et al. 2004; Chapin et al. 2014) to better understand the link between stream drying and ecological responses.
Our objective was to describe patterns of drying, rewetting, and flow variability across a gradient of intermittency at ungauged streams. We used two low-cost techniques to identify drying extent (e.g.,completely dry, isolated pools, or retain connectivity) and quantify metrics describing the flow regime of intermittent streams (timing, duration, magnitude, rate of change, and frequency of high and low flow events). In addition to describing ecologically relevant flow metrics that can be derived from low-cost field methods, we also identified landscape and geophysical drivers of stream intermittency and hydrologic variability. We hypothesized watershed size and slope would be primary drivers of intermittency, with smaller and steeper watersheds exhibiting earlier and more frequent periods of no-flow. We aim for the field methods, hydrologic metrics, and observed variability in stream drying described in this paper to help initiate larger efforts to monitor hydrologic variability of intermittent streams.