TY - JOUR
AU - Muhlfeld, Clint C
AB - Abstract Climate change is increasing the severity and extent of extreme droughts events, posing a critical threat to freshwater ecosystems, particularly with increasing human demands for diminishing water supplies. Despite the importance of drought as a significant driver of ecological and evolutionary dynamics, current understanding of drought consequences for freshwater biodiversity is very limited. We describe key barriers that hinder integrative drought research and monitoring across riverscapes. The primary constraint limiting understanding of ecological drought is an existing monitoring framework focused on human water consumption and flood risk in mainstem rivers. This approach is misaligned with escalating needs for research and data collection that illuminate exposure, sensitivity, and adaptive capacity (i.e., vulnerability) of biota to drought across entire riverscapes. We present a hierarchical framework for integrated ecological drought monitoring and research that addresses drought vulnerability across riverscapes and describe how this approach can directly inform natural-resource management. The increased frequency and magnitude of extreme weather events is one of the most visible consequences of climate change (Seneviratne et al. 2012). Drought (box 1), in particular, has become a critical threat to freshwater ecosystems and services (Lake 2003, Lake 2011). Globally, drought events appear to be increasing in severity and spatial extent (Dai 2013), reflecting ongoing changes in the prevailing climate. In North America, long-term paleoclimate records suggest that processes influencing hydrologic drought have changed substantially over recent decades, particularly toward increased aridity across the western United States (Pederson et al. 2013, Cook et al. 2015). Although decadal to multidecadal variation in hydroclimate is normal, recent multiyear droughts in the West have been severe and, in some cases, unprecedented in magnitude (e.g., Ault et al. 2013, Cook et al. 2015). These events have raised concerns about the prospects for megadroughts in freshwater ecosystems and about the resulting conflicts between ecological needs and escalating human demands for water resources that often equal or exceed supply (e.g., Barnett et al. 2008, Diffenbaugh et al. 2015). Drought (box 1) influences three critical axes of hydrologic variability in stream and river (i.e., lotic) environments: streamflow regime (timing, magnitude, and duration of flow), streamflow permanence (drying) and associated stream-network fragmentation, and stream temperature regime. Scientific and management communities increasingly recognize the vital role of hydrologic variability as a fundamental driver of freshwater ecosystems and the numerous services they provide (e.g., Poff et al. 1997, Poff and Zimmerman 2010). Hydrologic variability across space and time plays a critical role in determining the demographic and evolutionary trajectories of economically valuable species, as well as many plant and animal species of significant conservation concern or ecological importance (e.g., Lytle and Poff 2004). Of particular concern is the emergence of large-extent and high-magnitude drought events under warmer future conditions as major stressors to human and ecological systems (Crausbay et al. 2017). Box 1. What is drought? Drought can be described in meteorological, agricultural, hydrological, and ecological terms (Wilhite and Glantz 1985, Crausbay et al. 2017). For freshwater ecosystems, hydrologic drought is most directly pertinent to ecological dynamics, which, in turn, leads to the definition for ecological drought itself: “significant episodic shortages of water availability that can influence individual ecosystem services (e.g., species), overall ecosystem states, and trigger socio-ecological feedbacks” (Crausbay et al. 2017). Lake (2011) framed hydrologic drought—by definition an episodic process rather than long-term trend in conditions—in terms of seasonal and supraseasonal events (figure 1). Figure 1. Open in new tabDownload slide Examples of (a) seasonal (1973), (b) supraseasonal (1987–1990), and (c) decadal to multidecadal (750–2010) drought from the monthly gage record for the Big Hole River at Melrose, Montana. The red line in panels (a) and (b) serves as a reference highlighting the first quartile of the monthly flow data (9.7 cubic meters per second). The bottom plot shows a tree-ring-based reconstruction (the blue line) of naturalized flow (the black line), with a loess smoother (the red line) and 95% confidence intervals (the gray shading) highlighting the decadal to multidecadal drought and high-flow events. Figure 1. Open in new tabDownload slide Examples of (a) seasonal (1973), (b) supraseasonal (1987–1990), and (c) decadal to multidecadal (750–2010) drought from the monthly gage record for the Big Hole River at Melrose, Montana. The red line in panels (a) and (b) serves as a reference highlighting the first quartile of the monthly flow data (9.7 cubic meters per second). The bottom plot shows a tree-ring-based reconstruction (the blue line) of naturalized flow (the black line), with a loess smoother (the red line) and 95% confidence intervals (the gray shading) highlighting the decadal to multidecadal drought and high-flow events. Seasonal droughts are characterized by below-normal hydrological conditions and limited to an individual year, whereas supraseasonal drought events, also known as multiyear droughts, commonly persist anywhere from 2–10 years, but have been shown on rare occasion to extend well beyond 20 and even 60 years (Ault et al. 2013). These droughts have been described as megadroughts because of their severity and persistence over multiple decades. As depicted in figure 1, many streams experience predictable seasonal low flows, a critical environmental feature that influences the evolution of freshwater biodiversity (Lytle and Poff 2004). In contrast, the onset of supraseasonal drought can result in hydrologic changes (including discharge, streamflow permanence, and temperature) that can induce stress, lower abundance, or reduce distribution of species and disrupt normal ecosystem processes. The causes of supraseasonal droughts are diverse and largely driven by deficits in precipitation, changes in temperature (warming and cooling), and local catchment characteristics. Van Loon (2015) recognized at least eight types of hydrologic drought, on the basis of how they relate to meteorological anomalies (precipitation or temperature), their timing, and sequencing. Although rainfall deficits are often important, certain types of hydrologic drought can emerge from changes in temperature alone (e.g., a warm snow drought; Harpold et al. 2017). Meteorological changes initially translate into alterations in runoff and surface flows, and if anomalies persist, propagate into lagged groundwater responses. Consequently, hydrologic systems that respond quickly to meteorological changes (e.g., those with little storage in groundwater or lakes) often recover quickly, whereas systems with greater storage may respond more slowly and take longer to recover. In short, the components of the hydrologic processes influenced, as well as hydrologic characteristics of individual catchments, should drive expectations of response and recovery from meteorological drought. All forms of meteorological drought can influence hydrological regimes in multiple dimensions, including the timing, duration, magnitude, and frequency of physical events, all of which can have important implications for species and ecosystems. To this end, there are many descriptors of stream discharge (Poff et al. 1997), stream flow permanence (Jaeger et al. 2014), and temperature regimes (Steel et al. 2017), as well as spatial measures of hydrologic variability within (Fullerton et al. 2010) and among (Black et al. 2018) catchments. Lake (2011) emphasized the need to be as specific as possible in describing hydrologic drought in terms of processes and descriptors of responses, noting that most hydrological and ecological drought research actually addresses seasonal droughts, and not supraseasonal events, which are arguably what constitute true drought events. The effect of drought on riverine biota has been an area of considerable research and management focus. Existing research highlights that drought events have directly influenced, and often limited, the growth, survival, reproductive success, abundance, and distribution of many freshwater species throughout North America and elsewhere (e.g., Magoulick and Kobza 2003, Matthews and Matthews 2003, Lake 2011). Drought can also indirectly influence native aquatic biodiversity by altering, and often expanding, the distribution and abundance of invasive species (e.g., Beche et al. 2009, Muhlfeld et al. 2014), thereby affecting population-, species-, community-, and ecosystem-level dynamics (Lake 2011). Although much less appreciated for aquatic organisms (Matthews and Matthews 2003), drought events can strongly influence natural selection, genetic drift, and gene flow on contemporary time scales (Banks et al. 2013) and, therefore, evolutionary dynamics and their ecological consequences (Lowe et al. 2017). Indeed, drought is a key driver of natural selection in one of Earth's most famous systems for the study of evolution—Darwin's finches—in which drought events strongly influence natural selection acting on beak size and, in turn, hybridization among species (Grant and Grant 2002). Importantly, drought events can also strongly interact with a suite of existing human stressors (e.g., pollution, water consumption, land use), potentially leading to irreversible changes within freshwater ecosystems (Bogan and Lytle 2011, Power et al. 2015). Despite the importance of drought as a significant driver of ecological and evolutionary dynamics, understanding of drought consequences for aquatic biodiversity is limited, particularly for supraseasonal drought events in headwater streams—defined in the present article as tributary networks in first-, second-, and third-order streams. This reflects a fundamental limitation constraining research and management: The data necessary for quantifying drought impacts in freshwater—stream discharge, permanence, and temperature—are not available for the vast majority of headwater streams, because hydrological monitoring almost ubiquitously occurs in larger mainstem rivers (DeWeber et al. 2014). The focus on mainstem rivers reflects the historical legacy of a monitoring system devoted to informing human use (i.e., withdrawal) and associated regulation of water resources, or risk from flooding. Although it is effective for those purposes, this ecological blind spot is particularly concerning, given the critical value of headwater streams and, more broadly, the entire riverscape (sensu Fausch et al. 2002) to overall aquatic ecosystem functioning and the conservation of biodiversity (box 2). Several obstacles currently impede efforts to describe, predict, plan for, and ultimately mediate the effects of drought across riverscapes, from headwater streams to mainstem rivers. A key challenge of understanding ecological drought is the fact that it fundamentally assimilates processes ranging from physical to biological to social, and accordingly diverse expertise (Crausbay et al. 2017). In this article, we first identify and describe key barriers that limit our ability to integrate physical and biological data across scales (spatial and temporal) to advance scientific understanding of ecological drought in freshwater environments. Specifically, we focus on misalignments among natural-resource fields and the problems they cause for integrated drought science, challenges of modeling physical processes in small streams, and trade-offs associated with optimizing precision over the spatial extent of coverage in drought monitoring. We then outline a framework for integrated drought monitoring and research in freshwater that emphasizes ecological vulnerability as a means to integrate disciplines, advance research, and ultimately help address the escalating problem of drought in natural-resource management and conservation. Barriers to unified drought science across riverscapes Misalignment of physical and biological measurements Understanding the magnitude and consequences of ecological drought across North American riverscapes is hindered by the location, extent, and grain of current hydrologic monitoring. Specifically, there is a significant misalignment of physical and biological measurements across both space and time (e.g., Lake 2011). Hydrologic drought is well documented in North America at locations with continuous, long-term streamflow records (McCabe and Wolock 2014), but the streamflow gaging network, which is largely focused on monitoring discharge in mainstem rivers (Falcone et al. 2010, DeWeber et al. 2014), has limited spatial resolution (e.g., figure 3a) for describing fine-grained heterogeneity in streamflow responses to drought. Extrapolating the impacts of drought documented at mainstem gages to upstream locations is typically inaccurate because the hydrology of headwater streams is largely tied to local hillslope and geomorphic processes, which can vary markedly across headwater catchments (Montgomery 1999, Benda et al. 2005). Consequently, there is often little correspondence between the hydrologic regimes of small to medium catchments and large, mainstem rivers (Chezik et al. 2017). Furthermore, the distribution and number of permanent stream gaging stations is decreasing in many river basins, despite growing biological and social need for increasingly scant water resources (Ruhi et al. 2018). In contrast to hydrologic monitoring, biological data are often collected at spatial scales that target the ecological needs of species of interest. For example, systematic monitoring of trout, salmon, and char—a group of fishes worth billions of dollars annually—generally occurs in their spawning and rearing habitats in headwater streams that have little to no streamflow monitoring. This mismatch between biological and streamflow monitoring is consistent for other species of conservation concern (e.g., see figure 3). However, it should be noted that biological-monitoring strategies (including spatial and temporal frequency of sampling as well as methods) vary within and among federal, state, provincial, and tribal governmental agencies, a pattern that significantly hinders our understanding of how climatic variation influences ecological and evolutionary dynamics, including drought events (Kovach et al. 2016). That is, misalignments are present between disciplines (physical versus biological) and within disciplines. When hydrologic and biological data do align, few studies have involved the temporal scales needed to evaluate the short- and long-term effects of drought, and how these effects differ in severity over a single season, multiple seasons, or multiple years (i.e., supraseasonal drought events). Even when biological studies do overlap with drought periods, they rarely have pre- and postdrought data to effectively quantify immediate impact and long-term response relative to predrought conditions—that is, population-, species-, and ecosystem-level resistance and resilience (Bogan et al. 2015). That being said, long-term biological monitoring data are collected for numerous species in streams and rivers of North America (e.g., annual or semiannual estimates of abundance or density) especially for species of conservation concern or economic value. Those data will be critical for future ecological drought research across riverscapes and have proven essential for quantifying the effects of drought on species found in larger mainstem rivers where extensive flow data are available (e.g., Ruhi et al. 2015). Box 2. Ecological value of headwater streams. Although vastly underrepresented in terms of streamflow monitoring, headwater streams have a disproportionately large influence on the functioning of aquatic ecosystems across entire riverscapes. Headwater streams serve as the sources of nutrients, water, carbon, and sediment for mainstem rivers, all of which can vary considerably within and across catchments (Vannote et al. 1980). Headwater streams also fulfill the habitat requirements of different life-history stages for numerous freshwater species (Lowe and Likens 2005), many of which are declining or threatened with extinction (supplementary table S1). Indeed, over one hundred aquatic species listed as Threatened or Endangered under the US Endangered Species Act depend on headwater stream environments (figure 2, supplementary table S1), not to mention the numerous ecosystem services generated by headwater streams (Lowe and Likens 2005). Beyond the freshwater realm, riparian corridors within headwater streams are critical habitats for numerous terrestrial organisms and communities (Naiman et al. 1993) that are quite sensitive to drought-induced changes in hydrologic regimes (Tonkin et al. 2018). Importantly, the ecological and evolutionary consequences of drought can differ considerably between headwater streams and mainstem rivers, in part because the availability of refugia can vary markedly at numerous scales across riverscapes (Torgersen et al. 1999, Steel et al. 2017), emphasizing the need for riverscape perspective in ecological drought. Figure 2. Open in new tabDownload slide The importance of headwater habitats to species listed as Threatened or Endangered under the US Endangered Species Act (ESA). (a) The number of ESA-listed aquatic species that depend on headwater streams for all or part of their life cycle (see supplemental table 2 for a list). (b) The proportion of ESA-listed species within each taxonomic group that depend on headwater streams. Figure 2. Open in new tabDownload slide The importance of headwater habitats to species listed as Threatened or Endangered under the US Endangered Species Act (ESA). (a) The number of ESA-listed aquatic species that depend on headwater streams for all or part of their life cycle (see supplemental table 2 for a list). (b) The proportion of ESA-listed species within each taxonomic group that depend on headwater streams. Spatially intensive versus spatially extensive hydrology Misalignment in data collection and monitoring between physical and biological sciences reflects the historical development of programs serving different needs and users. The current observational network for streamflow in the United States was largely developed to serve water-management decisions at key points in river networks where there are dams, diversions, or wastewater discharges (Wahl et al. 1995). Water availability for in-stream and out-of-stream use and compliance with legal standards is typically determined at select locations intended to represent conditions across a river network. For example, the Colorado River Compact (1922)—a multistate water allocation agreement—is administered on the basis of streamflow in the Colorado River at Lee's Ferry, Arizona. Likewise, the curtailment of junior water users in a given river basin is triggered when streamflow drops below a threshold at a particular gage and on the basis of prior appropriations systems for administering water rights. Therefore, the current monitoring framework for streamflow hydrology in the United States is predicated on accurately defining human use, availability, and risk (of drought or flood). Specifically, a point-based water management system places high value on accurate streamflow measurements and forecasts at a small number of selected locations, which have been the historical focus for improving streamflow observations and models. A monitoring framework predicated on stream flow observations that are highly precise but spatially limited is often insufficient for assessing ecological-drought vulnerability. During drought events, aquatic organisms often respond to the spatial extent and connectivity of streamflow across river networks (Woodward et al. 2015). Therefore, hydrology focused on the spatial distribution of water across networks and its physical properties (e.g., temperature, dissolved oxygen, and depth) rather than on flux rates may be needed to understand the consequences of drought for aquatic ecosystems (Magoulick and Kobza 2003, Jaeger et al. 2014). However, the value of stream discharge data relative to coarser data describing drought condition (e.g., water availability) depends on the ecological context of drought (i.e., how different species use different portions of the riverscape at different life stages; Fausch et al. 2002). For ecological and evolutionary dynamics, other relevant summaries of drought include measures of frequency, duration, and timing (Steel et al. 2017) at spatial and temporal scales appropriate for organismal biology or management action. Challenges to modeling relevant physical processes in small streams Characterizing streamflow, intermittency, or temperature variation across riverscapes poses numerous practical challenges, particularly for low-flow conditions in headwater streams. Hydrologists primarily rely on hydrologic models calibrated from continuous flow gages to characterize current streamflow and to predict future drought scenarios (Bourdin et al. 2012, Jaeger et al. 2014). However, streamflow in larger streams and rivers is the product of upstream influences of many smaller streams and may not represent spatial heterogeneity in upstream processes (Benda et al. 2004, Larned et al. 2010). Similarly, during high-streamflow conditions, large-scale controls that govern runoff-routing processes, such as drainage density and cumulative basin area overwhelm effects of local processes (Pallard et al. 2009). Relevant streamflow statistics can usually be effectively modeled in larger rivers using landscape and climate data collected at coarse spatial and temporal resolutions, but those predictions are often incongruous with actual dynamics in headwater environments. Although spatial heterogeneity in headwater streams is widely acknowledged, it can be very difficult to quantify. In headwaters, hydrologic responses to seasonal variation in precipitation or snowmelt are strongly tied to local processes affecting water storage and release (Smakhtin 2001). Evapotranspiration (ET) associated with riparian vegetation and interactions of groundwater with surface water are particularly important processes at local scales. The physical drivers of these processes (e.g., topography, soil, geology) often vary across steep gradients or exhibit spatial discontinuities, particularly in mountainous regions (Brown et al. 2009). The result of this spatial complexity is often a mosaic of losing and gaining reaches distributed longitudinally throughout the drainage network (Payn et al. 2009, Weekes et al. 2012). Ecologically, these reaches can be thought of as habitat patches defined by flow permanence and hydrologic and thermal suitability (Schultz et al. 2017). It is the spatial topology of these patches and how they vary over time that ultimately control habitat suitability, connectivity, and drought refugia—that is, areas of freshwater permanence where biota can persist during drought events (Labbe and Fausch 2000, Magoulick and Kobza 2003). Unfortunately, characterizing the local processes driving low-flow hydrology in headwater basins remains challenging. Even a fundamental understanding of location, extent, and permanence of the stream network is poorly quantified in most headwaters (Fritz et al. 2013). In addition, the volume of baseflow in headwater streams commonly depends on groundwater, a basic understanding of which is critical for predicting ecological dynamics (e.g., Snyder et al. 2015). Alas, direct measurements of groundwater–surface water interactions are labor intensive, expensive, and usually not practical at ecologically relevant extents. Modeling groundwater processes is also complicated by subsurface aquifer boundaries that cannot be observed from the land surface and often change in response to recharge conditions (Winter et al. 2003). Similarly, ET can have large effects on streamflow and water availability, particularly under low-flow conditions (Wondzell et al. 2007, Cooper et al. 2018). Therefore, modeling streamflow conditions in headwater environments remains elusive, necessitating a shift to cost-effective monitoring that accommodates conditions at the riverscape extent. Recently, there have been numerous technological and methodological advances for estimating streamflow, permanence, and thermal regimes at ungaged sites. Similarly, new sources of landscape data are available that provide higher-resolution depictions of terrain and vegetation that more effectively capture the fine-scale spatial discontinuities important in headwaters (table 1). When these tools are deployed or summarized at the appropriate spatial grain, their data can be used to parameterize streamflow models capable of assessing drought effects at the riverscape scale. Moreover, the coordinated integration of these approaches into research and monitoring networks across different landscape settings (e.g., snowmelt-driven versus rainfall-driven versus groundwater systems, alpine versus montane versus desert) can provide new insights into the previously intractable goal of understanding consequences of drought on organisms in headwater streams. Table 1. Tools and methods for measuring and modeling the three critical axes of drought impacts across riverscapes, including stream discharge, permanence, and thermal regimes. Drought attribute . Tools and method . Description . Spatial grain . Streamflow, water depth and velocity Pressure transducer Provide high-frequency measures of water depth. Can be used to estimate flow and flow-related statistics at ungauged sites. Watershed Salt dilution Technique for measuring streamflow in turbulent headwater streams at stream reach scale. Reach Remote sensing Increasingly accurate aerial imagery can be used to measure stream width and even bathymetry in lower order watersheds. Reach to watershed Particle image velocimetry Measurement of surface velocity using passive optical image data. Can be used to estimate streamflow. Reach to watershed Radar and acoustic sensors Measure surface-water velocities using suspended instruments to estimate discharge. Reach to watershed Stream temperature regime Temperature loggers Provide high-frequency water temperature at instrumented sites. Reach Thermal infrared cameras Thermal infrared cameras provide spatially continuous data on stream temperature. Watershed Groundwater–surface water interactions Salt dilution Salt dilution gauging in large reaches allows for estimation of groundwater and hyporheic exchange. Reach Temperature loggers High-frequency water temperature data can inform models that estimate groundwater influence using “heat as tracer” approaches. Reach Fiber-optic temperature sensing High spatial resolution temperature monitoring for longitudinal or vertical deployment and applied to numerical models following “heat as tracer” theory. Reach Conductivity sensors High-frequency specific conductance data, applied to stream discharge data, can be used to quantify magnitude of groundwater influence. Watershed Stream permanence Temperature loggers High-frequency water temperature data can be used to inform statistical models to estimate the probability of drying. Reach Electrical resistivity loggers Modified temperature sensors in which state changes in electrical resistance are proxy for presence or absence of flow. Reach Temperature and electric resistivity logger High-frequency water and electrical resistivity can be used to detect stream drying. Reach Evapotranspiration Water level loggers Diurnal variation in water level can be used to estimate evapotranspiration. Reach Lidar Measurement of vegetation provides inputs for transpiration, canopy interception, evaporation, and stream flow discharge. Watershed Stream channel mapping Lidar Improved mapping of stream channel informs ecologically relevant flow measures including channel extent, origination, and intermittency. Watershed Drought attribute . Tools and method . Description . Spatial grain . Streamflow, water depth and velocity Pressure transducer Provide high-frequency measures of water depth. Can be used to estimate flow and flow-related statistics at ungauged sites. Watershed Salt dilution Technique for measuring streamflow in turbulent headwater streams at stream reach scale. Reach Remote sensing Increasingly accurate aerial imagery can be used to measure stream width and even bathymetry in lower order watersheds. Reach to watershed Particle image velocimetry Measurement of surface velocity using passive optical image data. Can be used to estimate streamflow. Reach to watershed Radar and acoustic sensors Measure surface-water velocities using suspended instruments to estimate discharge. Reach to watershed Stream temperature regime Temperature loggers Provide high-frequency water temperature at instrumented sites. Reach Thermal infrared cameras Thermal infrared cameras provide spatially continuous data on stream temperature. Watershed Groundwater–surface water interactions Salt dilution Salt dilution gauging in large reaches allows for estimation of groundwater and hyporheic exchange. Reach Temperature loggers High-frequency water temperature data can inform models that estimate groundwater influence using “heat as tracer” approaches. Reach Fiber-optic temperature sensing High spatial resolution temperature monitoring for longitudinal or vertical deployment and applied to numerical models following “heat as tracer” theory. Reach Conductivity sensors High-frequency specific conductance data, applied to stream discharge data, can be used to quantify magnitude of groundwater influence. Watershed Stream permanence Temperature loggers High-frequency water temperature data can be used to inform statistical models to estimate the probability of drying. Reach Electrical resistivity loggers Modified temperature sensors in which state changes in electrical resistance are proxy for presence or absence of flow. Reach Temperature and electric resistivity logger High-frequency water and electrical resistivity can be used to detect stream drying. Reach Evapotranspiration Water level loggers Diurnal variation in water level can be used to estimate evapotranspiration. Reach Lidar Measurement of vegetation provides inputs for transpiration, canopy interception, evaporation, and stream flow discharge. Watershed Stream channel mapping Lidar Improved mapping of stream channel informs ecologically relevant flow measures including channel extent, origination, and intermittency. Watershed Note: See supplemental table S3 for relevant citations. Open in new tab Table 1. Tools and methods for measuring and modeling the three critical axes of drought impacts across riverscapes, including stream discharge, permanence, and thermal regimes. Drought attribute . Tools and method . Description . Spatial grain . Streamflow, water depth and velocity Pressure transducer Provide high-frequency measures of water depth. Can be used to estimate flow and flow-related statistics at ungauged sites. Watershed Salt dilution Technique for measuring streamflow in turbulent headwater streams at stream reach scale. Reach Remote sensing Increasingly accurate aerial imagery can be used to measure stream width and even bathymetry in lower order watersheds. Reach to watershed Particle image velocimetry Measurement of surface velocity using passive optical image data. Can be used to estimate streamflow. Reach to watershed Radar and acoustic sensors Measure surface-water velocities using suspended instruments to estimate discharge. Reach to watershed Stream temperature regime Temperature loggers Provide high-frequency water temperature at instrumented sites. Reach Thermal infrared cameras Thermal infrared cameras provide spatially continuous data on stream temperature. Watershed Groundwater–surface water interactions Salt dilution Salt dilution gauging in large reaches allows for estimation of groundwater and hyporheic exchange. Reach Temperature loggers High-frequency water temperature data can inform models that estimate groundwater influence using “heat as tracer” approaches. Reach Fiber-optic temperature sensing High spatial resolution temperature monitoring for longitudinal or vertical deployment and applied to numerical models following “heat as tracer” theory. Reach Conductivity sensors High-frequency specific conductance data, applied to stream discharge data, can be used to quantify magnitude of groundwater influence. Watershed Stream permanence Temperature loggers High-frequency water temperature data can be used to inform statistical models to estimate the probability of drying. Reach Electrical resistivity loggers Modified temperature sensors in which state changes in electrical resistance are proxy for presence or absence of flow. Reach Temperature and electric resistivity logger High-frequency water and electrical resistivity can be used to detect stream drying. Reach Evapotranspiration Water level loggers Diurnal variation in water level can be used to estimate evapotranspiration. Reach Lidar Measurement of vegetation provides inputs for transpiration, canopy interception, evaporation, and stream flow discharge. Watershed Stream channel mapping Lidar Improved mapping of stream channel informs ecologically relevant flow measures including channel extent, origination, and intermittency. Watershed Drought attribute . Tools and method . Description . Spatial grain . Streamflow, water depth and velocity Pressure transducer Provide high-frequency measures of water depth. Can be used to estimate flow and flow-related statistics at ungauged sites. Watershed Salt dilution Technique for measuring streamflow in turbulent headwater streams at stream reach scale. Reach Remote sensing Increasingly accurate aerial imagery can be used to measure stream width and even bathymetry in lower order watersheds. Reach to watershed Particle image velocimetry Measurement of surface velocity using passive optical image data. Can be used to estimate streamflow. Reach to watershed Radar and acoustic sensors Measure surface-water velocities using suspended instruments to estimate discharge. Reach to watershed Stream temperature regime Temperature loggers Provide high-frequency water temperature at instrumented sites. Reach Thermal infrared cameras Thermal infrared cameras provide spatially continuous data on stream temperature. Watershed Groundwater–surface water interactions Salt dilution Salt dilution gauging in large reaches allows for estimation of groundwater and hyporheic exchange. Reach Temperature loggers High-frequency water temperature data can inform models that estimate groundwater influence using “heat as tracer” approaches. Reach Fiber-optic temperature sensing High spatial resolution temperature monitoring for longitudinal or vertical deployment and applied to numerical models following “heat as tracer” theory. Reach Conductivity sensors High-frequency specific conductance data, applied to stream discharge data, can be used to quantify magnitude of groundwater influence. Watershed Stream permanence Temperature loggers High-frequency water temperature data can be used to inform statistical models to estimate the probability of drying. Reach Electrical resistivity loggers Modified temperature sensors in which state changes in electrical resistance are proxy for presence or absence of flow. Reach Temperature and electric resistivity logger High-frequency water and electrical resistivity can be used to detect stream drying. Reach Evapotranspiration Water level loggers Diurnal variation in water level can be used to estimate evapotranspiration. Reach Lidar Measurement of vegetation provides inputs for transpiration, canopy interception, evaporation, and stream flow discharge. Watershed Stream channel mapping Lidar Improved mapping of stream channel informs ecologically relevant flow measures including channel extent, origination, and intermittency. Watershed Note: See supplemental table S3 for relevant citations. Open in new tab Toward a unified framework for ecological drought We describe a generic workflow and framework that seeks to improve understanding of drought processes and impacts on stream and river systems through coordinated and multidisciplinary data collection, synthesis, analysis, and predictions. Specifically, this workflow can be adapted to unique configurations of physical and biological conditions in riverscapes to inform drought vulnerability of freshwater species, which, in turn, can be used to identify robust management strategies for natural-resource managers. This overarching framework will better unify and strategically enhance future ecological drought monitoring and research across riverscapes of North America and beyond. An integrated drought monitoring and research workflow An integrated and effective ecological drought program must assess key stakeholder values associated with ecological drought (Kennen et al. 2018). Surprisingly, such efforts are rare, but when they do occur, fish and the values they provide often rise to the forefront for economic, ecological, and cultural reasons (e.g., McEvoy et al. 2018). Drought is of particular concern to communities and tribal sovereigns that rely on fish, such as salmon and trout (both of which depend on headwater environments), as major subsistence food sources or drivers of local economies (Cozetto et al. 2013). Indeed, recent drought-induced mortality events of salmon and other fish in the Pacific Northwest were notable phenomena that highlighted the need for a priori drought-adaptation plans, and the associated ecohydrologic monitoring networks needed to support those plans (Hand et al. 2018). Furthermore, drought observation networks must effectively address conservation and management needs across riverscapes of North America. This will necessitate identifying cost-effective means of monitoring and evaluating drought impacts across entire riverscapes, including headwater systems in various landscape settings, and in locations with immediate stakeholder needs (box 3). Doing so will require new tools and technologies that collate data from public (e.g., US Geological Survey) and private (e.g., watershed groups, nongovernmental organization) gages that already exist in headwater streams, and that expand and strengthen space-based, airborne, aquatic, and terrestrial data observation networks at multiple spatial and temporal scales (table 1). After development and refinement, this effort can be significantly enhanced by existing and future crowdsourcing efforts (e.g., Isaak et al. 2017). Next, coordinated and integrated observation networks will enable researchers to develop and synthesize data sets that improve understanding of drivers, responses, and interactions using appropriate methods. Improved understanding of the relationships between drought conditions and the resulting biotic responses—both spatially and temporally—will provide the observational foundation for improved forecasting and management of freshwater resources and the biota they support (Lake 2011). This approach will enable transfer of science and knowledge across rivers and streams of North America, with applications globally. Once key data sets and models are synthesized, a more comprehensive understanding of drought processes, ecological responses, and recovery times can be achieved, thereby facilitating delivery of decision support guides to address management needs for a wider range of decisions. Box 3. Developing an ecohydrological drought-monitoring network. Capturing abiotic heterogeneity across spatial scales is critical for understanding exposure to drought conditions, as well as biotic sensitivity, resistance, and resilience across landscapes (Picket and Thompson 1978, Fausch et al. 2002). The development of monitoring networks that appropriately describe drought conditions across riverscapes (e.g., figure 3c) will provide opportunities to address a series of questions critical for natural resource management and drought-scenario planning: Does water availability (quantity and quality) at a mainstem gage represent how drought is affecting upstream headwater streams, and how does this pattern vary across ecoregions? What are the physical and climatic sources of heterogeneity that disrupt relationships between the hydrology of upstream catchments and that of mainstem rivers? How should monitoring progress if there is ecologically significant variability between headwater streams and mainstem rivers within and between catchments? Variation in drought-induced conditions within and among catchments has significant implications for ecological drought, as well as climate-change resiliency; hydrologic heterogeneity increases the probability that some streams remain ecologically viable under long-term change or extreme events (e.g., drought), a phenomenon akin to the portfolio effect (sensu Chezik et al. 2017). Figure 3. Open in new tabDownload slide Misalignments in current ecological drought monitoring. (a) The current distribution of streamflow gages (the red circles) relative to federally designated critical spawning and rearing habitat (the pink lines) for (c) bull trout, a Threatened species, and (a) the known distribution of populations (the green circles) of (b) Lednia tumana (proposed Threatened). There are no hydrologic gages in the Middle Fork Flathead River (watershed area = 3004 square kilometers), Montana, that overlap with critical habitat designations, or provide information about drought-induced hydrologic conditions in headwater streams that support other imperiled biota including Lednia tumana. Future drought research and monitoring need to identify cost-effective means to address ecological drought across entire riverscapes (orange circles), idealized in panel (d). Figure 3. Open in new tabDownload slide Misalignments in current ecological drought monitoring. (a) The current distribution of streamflow gages (the red circles) relative to federally designated critical spawning and rearing habitat (the pink lines) for (c) bull trout, a Threatened species, and (a) the known distribution of populations (the green circles) of (b) Lednia tumana (proposed Threatened). There are no hydrologic gages in the Middle Fork Flathead River (watershed area = 3004 square kilometers), Montana, that overlap with critical habitat designations, or provide information about drought-induced hydrologic conditions in headwater streams that support other imperiled biota including Lednia tumana. Future drought research and monitoring need to identify cost-effective means to address ecological drought across entire riverscapes (orange circles), idealized in panel (d). Despite the clear need for a revised monitoring framework that more explicitly accounts for biota in riverine environments, monitoring of drought conditions across riverscapes poses many nontrivial logistical challenges. Overcoming these challenges will require integrated monitoring programs that work, initially, to adaptively test and implement novel methodologies. Fortunately, there have been numerous conceptual and technological advances (Table 1) that facilitate the direct measurement or estimation of flow and flow processes in small watersheds (Keenan et al. 2018). These advances provide powerful opportunities to shift current drought monitoring in mainstem rivers to the entire river network (figure 3a; Fausch et al. 2002), thereby facilitating the development of ecological drought relationships across scales. The experimental implementation of ecohydrological monitoring networks that collect drought-relevant data across riverscapes will require testing various methods at similar spatial scales (i.e., replicated study designs) across diverse ecoregions with varying propensity for and sensitivity to both meteorological and hydrological drought. This will provide robust comparisons of drought conditions in different landscapes while helping to identify what types of transferable and unique abiotic and biotic data are needed in different environmental, ecological, and evolutionary settings. Issues of scalability will pose immediate challenges, but an adaptive approach that targets pilot watersheds on the basis of past or future drought exposure (figure 4) will address decision-making needs while informing the development of more inclusive watershed monitoring approaches into the future. Figure 4. Open in new tabDownload slide Recent (a) and predicted future supraseasonal drought risk (b) across the United States. Panel (a) depicts the average warm season (May–August) drought conditions (2001–2016) based on the Palmer Drought Severity Index (scPDSI 3.25; Osborn et al. 2017). (b) The risk of future supraseasonal to decadal drought events over the United States using information on natural drought variability from paleodata combined with future scenarios from climate models (adapted from Ault et al. 2014). Figure 4. Open in new tabDownload slide Recent (a) and predicted future supraseasonal drought risk (b) across the United States. Panel (a) depicts the average warm season (May–August) drought conditions (2001–2016) based on the Palmer Drought Severity Index (scPDSI 3.25; Osborn et al. 2017). (b) The risk of future supraseasonal to decadal drought events over the United States using information on natural drought variability from paleodata combined with future scenarios from climate models (adapted from Ault et al. 2014). Defining biological drought vulnerability Developing an integrated framework for ecological drought monitoring, modeling, and research provides a critical first step toward delivering scientific information that can be used to inform both proactive and reactive management. In particular, an integrated drought strategy can be used to develop drought-adaptation strategies (that is, societal adaptation to drought) that accommodate the needs of aquatic species as well as human uses. We advocate for the use of a vulnerability framework as a robust means to conceptually and quantitatively capture the environmental, ecological, and evolutionary processes that interact to influence ecological drought at population, species, community, and ecosystem scales. This multidisciplinary approach incorporates a variety of knowledge to explicitly address the three primary elements of vulnerability—exposure, sensitivity, and adaptive capacity—as well as the potential for human responses or interventions (Dawson et al. 2011, Foden et al. 2013; figure 5a). Figure 5. Open in new tabDownload slide Exposure (the blue ring), sensitivity (the green ring), and adaptive capacity (the orange ring) collectively represent climate-change vulnerability (a). In the context of drought, species and ecosystems that experience greater exposure and have higher sensitivity and lower adaptive capacity are most vulnerable. The key elements of ecological-drought vulnerability (exposure, sensitivity, and adaptive capacity) have clear conceptual overlap with integrated drought research (b). Figure 5. Open in new tabDownload slide Exposure (the blue ring), sensitivity (the green ring), and adaptive capacity (the orange ring) collectively represent climate-change vulnerability (a). In the context of drought, species and ecosystems that experience greater exposure and have higher sensitivity and lower adaptive capacity are most vulnerable. The key elements of ecological-drought vulnerability (exposure, sensitivity, and adaptive capacity) have clear conceptual overlap with integrated drought research (b). Exposure represents the type, magnitude, and rate of environmental change. In the case of hydrologic drought (box 1), this includes changes to discharge, thermal regimes, and streamflow permanence. These sources of exposure can be related to and exacerbated by other abiotic factors that also respond to drought (e.g., wildfire). Sensitivity represents the degree to which the fitness, persistence, or performance of species or constituent populations depends on future drought exposure, especially the aspects of hydrological drought (e.g., frequency, magnitude, and duration) that are likely to change (Dawson et al. 2011). In aquatic habitats, this may include how ecosystem processes (e.g., primary productivity, respiration) respond to changes in the size and spatial distribution of suitable habitat, stream flows, and water temperatures (Davis et al. 2013). Species may be influenced through physiological, behavioral, and demographic responses, as well as by interactions among species, including competition, predation, disease, and parasites. Adaptive capacity refers to the ability of a species or population to cope with and persist under new environmental conditions (Nicotra et al. 2015, Beever et al. 2016). Aquatic species may exhibit adaptive capacity by accommodating drought-associated environmental changes through dispersal, phenotypic plasticity, and evolution by natural selection (Nicotra et al. 2015). However, drought itself can limit adaptive capacity by altering both phenotypic and genetic variation (Hendry 2016). Importantly, a vulnerability framework represents a clear and logical extension of integrated ecological drought research (figure 5b). Hierarchical vulnerability framework for ecological drought The dendritic nature of riverscapes imposes hierarchical structure on biodiversity (Frissell et al. 1986, Fausch et al. 2002). In the present article, hierarchy refers to different levels of constraint on biodiversity observed at any given location within a riverscape. For example, local conditions may be suitable for a given species, but the species may be absent because of higher-level constraints imposed by dispersal barriers or unsuitable landscape conditions. Accounting for hierarchical levels of control across the entire riverscape is critical for natural-resource management and conservation (Fausch et al. 2002). Ecological-drought vulnerability analyses need to be able to quantitatively compare the vulnerability of biological diversity at various hierarchical scales, including local populations metapopulations species and, ultimately, ecosystems (Wiens and Bachelet 2010). Coherent, regional expressions of climate, such as meteorological drought, translate through local hydrological and ecological processes to drive ecological drought exposure and vulnerability. Within riverscapes affected by drought, there may be locations that are resistant or resilient and capable of providing refuge during such conditions (Schultz et al. 2017). To capture these critical responses, future ecological drought monitoring and subsequent vulnerability analyses must effectively capture the mean and variance of exposure, sensitivity, and adaptive capacity within and among streams and watersheds. The variance in vulnerability elements across potential levels of hierarchical controls may be especially critical (Black et al. 2018), but to date has been almost universally ignored in existing climate-change vulnerability analyses. Spatial or temporal variability in ecological-drought exposure can highlight differential hydrologic sensitivity to meteorological drought or climate change (e.g., Chezik et al. 2017), whereas variance in sensitivity or adaptive capacity may determine the scope for management action (Mills and Lindberg 2002). Hierarchical drought vulnerability analyses will require a shift toward integrated ecological drought monitoring and research across riverscapes (box 3). For example, monitoring of headwater stream environments that many aquatic species depend on for persistence (figure 3c) will provide the data necessary to describe drought vulnerability within and among local populations for species of conservation concern or ecological and economic value (figure 6a), as well as being relevant to a host of other issues, species, and user groups. This hierarchical approach can be appropriately scaled to address various questions and issues surrounding ecological drought, both from scientific and applied perspectives. In so doing, this method of ecological-drought vulnerability analysis provides wider range of ecological and human-related information to inform a range of management decisions, including conservation prioritization at various hierarchical scales. Moreover, this approach can effectively partition risk among the major vulnerability elements (exposure, sensitivity, and adaptive capacity; Wade et al. 2017a), thereby helping identify the primary environmental, ecological, or evolutionary challenges confronting species and management. For example, bull trout in the Middle Fork of the Flathead River likely have moderate drought vulnerability, a pattern that predominately reflects their sensitivity to drought conditions rather than a propensity for significant drought exposure in that watershed (figure 6). Together, a prioritization framework coupled with quantitative sensitivity analysis provides robust integrated results that can directly inform decision-making for biological conservation and natural resource management, including human water use. However, the collection and synthesis of disparate data, of varying quality, collected within and among watersheds, requires appropriate quantitative frameworks that can account for and propagate uncertainty and error. Figure 6. Open in new tabDownload slide A heuristic example of hierarchical drought vulnerability for bull trout in the Middle Fork of the Flathead River (see figure 3) informed by riverscape-scale ecodrought monitoring. (a) Drought vulnerability (see figure 5) for each local population, where larger rings depict greater exposure (blue) and sensitivity (green), and lower adaptive capacity (orange). Vulnerability itself is represented by the size of the grey circle surrounding the sum of the vulnerability elements. Total watershed (i.e., metapopulation) vulnerability is shown in the top left corner, where ring size refers to the mean effect and the width of the line represents spatial variability. (b) The site-specific (the light lines) and watershed-level (the bold lines) mean and variance of each vulnerability element. The bold lines represent hyperparameters in a Bayesian hierarchical modeling framework, and their values are directly informed by the probability distributions for each local population. Figure 6. Open in new tabDownload slide A heuristic example of hierarchical drought vulnerability for bull trout in the Middle Fork of the Flathead River (see figure 3) informed by riverscape-scale ecodrought monitoring. (a) Drought vulnerability (see figure 5) for each local population, where larger rings depict greater exposure (blue) and sensitivity (green), and lower adaptive capacity (orange). Vulnerability itself is represented by the size of the grey circle surrounding the sum of the vulnerability elements. Total watershed (i.e., metapopulation) vulnerability is shown in the top left corner, where ring size refers to the mean effect and the width of the line represents spatial variability. (b) The site-specific (the light lines) and watershed-level (the bold lines) mean and variance of each vulnerability element. The bold lines represent hyperparameters in a Bayesian hierarchical modeling framework, and their values are directly informed by the probability distributions for each local population. Hierarchical Bayesian models (HBMs) will be particularly useful for ecological-drought vulnerability because they are explicitly designed to deal with sparse data across space and time, account for uncertainty in the data (observations), are well suited for combining multiple data sources and model structures, and, most importantly, provide robust estimates of the mean and variance across spatial or temporal scales (Kery and Shaub 2011, Steinschneider et al. 2012). HBMs can deal with sparse data by modeling local processes as a component of an overarching distribution. The model hyperparameters define parameter distributions across all locations (e.g., figure 6b). For locations with substantial data, local parameters will be mainly driven by the local data. In contrast, locations with sparse data will borrow information from the hyperparameters and shrink toward the hyperparameter or average values. In this framework, locations with variable amounts of data are useful because all locations contribute to the hyperparameters and local estimates. HBMs will likely prove particularly important for linking drought estimates to ecological processes and developing effective models, given that incomplete sampling is widespread in both hydrology and ecology. Uncertainty in the modeling process can arise from error in observations, model structure, or parameter estimates (Renard et al. 2010). State space models, a class of HBMs, explicitly separate the observation model from the process models (Zipkin et al. 2014), thereby allowing uncertainty in the observation model to inform the process model. Although observation error in gage data is usually minimal, structural and parametric error can be substantial in hydrologic models (Renard et al. 2010), and this issue will likely expand as novel, cost effective, but relatively imprecise data collections are used to monitor and characterize hydrologic conditions across riverscapes. HBM provide techniques to fully characterize sources of model uncertainty (Steinschneider et al. 2012), which is particularly important as models are combined for explicit quantification of ecological vulnerability. Together, a hierarchical conceptual framework coupled with explicitly hierarchical quantitative methodologies provides a powerful framework that addresses many of the challenges inherent in this field (Wade et al. 2017b), for both drought and assessments of climate change vulnerability (e.g., Leasure et al. 2018). Extending the framework into terrestrial and social systems To fully understand and prepare for ecological drought along entire riverscapes it will be necessary to extend assessments of ecological-drought vulnerability into nearby terrestrial environments, and to fully incorporate how human water demands will interact with and often exacerbate the ecological consequences of drought (Crausbay et al. 2017). Rivers and their riparian corridors are known to play an important role in regional biodiversity (Hauer et al. 2016). Riparian areas are used obligately by many terrestrial species and strongly influence groundwater–surface water interactions. Up to 70% of a region's vertebrate species require riparian corridors during their lifetime, emphasizing that riparian areas play critical roles in maintaining diversity at landscape and regional extents (Gregory et al. 1991). Even more importantly, human water use already exceeds supply over much of North America (Vorosmarty et al. 2000), a pattern that will only intensify under future climate change and human population expansion. Fortunately, the hierarchical vulnerability framework we outline can readily expand to include drought in terrestrial environment and incorporate social–ecological interactions (Dunham et al. 2018). Like many climate-change vulnerability frameworks (e.g., Foden et al. 2013), our ecological-drought vulnerability framework can be readily applied to ecological processes in terrestrial environments; the only limiting factor is data availability. With strategic development of ecohydrological monitoring networks (box 3) it will be possible to quantify drought exposure, species sensitivity, and adaptive capacity along entire riverscapes, including surrounding riparian areas. Furthermore, it will be feasible to identify and measure links between terrestrial and aquatic environments that influence ecological-drought vulnerability itself (e.g., ET). Again, HBMs hold considerable power and promise as they can explicitly include links among submodels (e.g., species interactions, aquatic–terrestrial links) and accommodate multiple species-specific ecological-drought vulnerability analyses simultaneously. That is, a single HBM could provide ecosystem-level estimates of ecological-drought vulnerability summarizing the vulnerability of multiple constituent species in terrestrial and aquatic habitats (i.e., species specific vulnerability estimates would be hyperparameters). This flexibility can allow for conservation prioritization and decision-making at various spatial and biological scales. Finally, it is critical that ecological-drought vulnerability analyses address human water demands, because this stressor will almost necessarily act synergistically with both meteorological and hydrological drought. Ideally, vulnerability analyses will address the extent of existing water use, future water use scaled to future human population expansion, future water use under various drought scenarios, and the scope for societal adaptation (e.g., existing and future water law, conservation measures, storage). At a minimum, human water consumption can be directly incorporated in ecological-drought vulnerability analyses through metrics describing exposure (Crausbay et al. 2017). Ecological-drought vulnerability analyses should also acknowledge how human water consumption (and other human stressors) also influences species sensitivity and adaptive capacity (see figure 1 of Beever et al. 2016). For example, stream and river fragmentation due to human dewatering can shrink and isolate populations thereby magnifying population stochasticity and increasing risk of inbreeding depression (sensitivity) while also decreasing both genetic and life-history variation (adaptive capacity). Therefore, ignoring human water use and other compounding stressors (e.g., invasive species, habitat degradation, pollution) may grossly underestimate ecological-drought vulnerability now and into the future. Conclusions Drought and climate change together offer unprecedented threats to freshwater biodiversity, especially when coupled with invasive species expansions, habitat modification, and water withdrawal. To effectively address the escalating threat of ecological drought, there is a clear need to reenvision how abiotic and biotic conditions are monitored and how these data are integrated across riverscapes. The current system of monitoring in the United States and throughout North America is insufficient for addressing ecological vulnerability to drought throughout the vast majority of aquatic ecosystems. The fields of climatology, hydrology, biology, and evolutionary ecology must better integrate efforts so that robust prioritization efforts and management plans for social–ecological systems concerning freshwater become feasible and holistic. This will require adopting a new framework that prioritizes data collection, monitoring, synthesis, and drought-vulnerability analysis across entire riverscapes, especially in headwater streams that are critical for many terrestrial and aquatic species and ecosystem services. The integrated and hierarchical ecological-drought vulnerability framework we describe can appropriately account for existing ecological theory concerning riverine environments (e.g., Frissel et al. 1986, Fausch et al. 2002), leverage recent statistical advances that better account for and describe heterogeneity across scales, account for human dimensions, and more realistically account for the needs of aquatic organisms in freshwater systems. Ultimately, shifting toward a fully integrated drought-vulnerability framework will improve the capacity for societal drought adaptation, the conservation of numerous species dependent on stream and river environments, and our understanding of ecohydrological processes in a changing world. Acknowledgments This work was supported by the USGS Fisheries Program and Ecosystems Mission Area. Any use of trade, produce, or firm names is for descriptive purposes only and does not imply endorsement by the US government. Author Biographical Ryan P. Kovach (rpkovach@gmail.com) is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, in Missoula, Montana. Jason B. Dunham is affiliated with the US Geological Survey, Forest and Rangeland Ecosystem Science Center, in Corvallis, Oregon. Robert Al-Chokhachy is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, in Bozeman, Montana. Craig D. Snyder is affiliated with the US Geological Survey, Leetown Science Center, in Kearneysville, West Virginia. Benjamin H. Letcher is affiliated with the US Geological Survey, Leetown Science Center, S. O. Conte Anadromous Fish Research Laboratory, in Turners Falls, Massachusetts. John A. Young is affiliated with the US Geological Survey, Leetown Science Center, in Kearneysville, West Virginia. Erik A. Beever is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, in Bozeman, Montana. Greg T. Pederson is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, in Bozeman, Montana. Abigail J. Lynch is affiliated with the US Geological Survey, National Climate Adaptation Science Center, in Reston, Virginia. Nathaniel P. Hitt is affiliated with the US Geological Survey, Leetown Science Center, in Kearneysville, West Virginia. Chris P. Konrad is affiliated with the US Geological Survey, Washington Water Science Center, in Tacoma, Washington. Kristin L. Jaeger is affiliated with the US Geological Survey, Washington Water Science Center, in Tacoma, Washington. Alan H. Rea is affiliated with the US Geological Survey, National Geospatial Program, in Boise, Idaho. Adam J. Sepulveda is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, in Bozeman, Montana. Patrick M. Lambert is affiliated with the US Geological Survey, Southwest Region, in Denver, Colorado. Jason Stoker is affiliated with the US Geological Survey, National Geospatial Program, in Reston, Virginia. Joseph J. Giersch is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, Glacier National Park, in West Glacier, Montana. Clint C. Muhlfeld is affiliated with the US Geological Survey, Northern Rocky Mountain Science Center, Glacier National Park, in West Glacier, Montana. References cited Ault TR , Cole JE, Overpeck JT, Pederson GT, St George S, Otto-Bliesner B, Woodhouse CA, Deser C. 2013 . The continuum of hydroclimate variability in western North America during the last millennium . Journal of Climate 26 : 5863 – 5878 . Google Scholar Crossref Search ADS WorldCat Ault TR , Cole JE, Overpeck JT, Pederson GT, Meko DM. 2014 . Assessing the risk of persistent drought using climate model simulations and paleoclimate data . Journal of Climate 27 : 7529 – 7549 . Google Scholar Crossref Search ADS WorldCat Banks SC , Cary GJ, Smith AL, Davies ID, Driscoll DA, Gill AM, Lindenmayer DB, Peakall R. 2013 . How does ecological disturbance influence genetic diversity? Trends in Ecology and Evolution 28 : 670 – 679 . Google Scholar Crossref Search ADS PubMed WorldCat Barnett TP et al. . 2008 . Human-induced changes in the hydrology of the western United States . Science 319 : 1080 – 1083 . Google Scholar Crossref Search ADS PubMed WorldCat Beever EA et al. . 2016 . Improving conservation outcomes with a new paradigm for understanding species’ fundamental and realized adaptive capacity . Conservation Letters 9 : 131 – 137 . Google Scholar Crossref Search ADS WorldCat Beche LA , Conners PG, Resh VH, Merenlender AM. 2009 . Resilience of fishes and invertebrates to prolonged drought in two California streams . Ecography 32 : 778 – 788 . Google Scholar Crossref Search ADS WorldCat Benda LEE , Poff NL, Miller D, Dunne T, Reeves G, Pess G, Pollock M. 2004 . The network dynamics hypothesis: How channel networks structure riverine habitats . BioScience 54 : 413 – 427 . Google Scholar Crossref Search ADS WorldCat Benda LEE , Hassan MA, Church M, May CL. 2005 . Geomorphology of steepland headwaters: The transition from hillslopes to channels . Journal of the American Water Resources Association 41 : 8232 ; 835 – 851 . Google Scholar Crossref Search ADS WorldCat Black BA et al. . 2018 . Rising synchrony controls western North American ecosystems . Global change Biology 24 : 2305 – 2314 . Google Scholar Crossref Search ADS PubMed WorldCat Bourdin DR , Fleming SW, Stull RB. 2012 . Streamflow modeling: A primer on applications, approaches, and challenges . Atmosphere-Ocean 50 : 507 – 536 . Google Scholar Crossref Search ADS WorldCat Bogan MT , Boersma KS, Lytle DA. 2015 . Resistance and resilience to seasonal and supraseasonal drought in invertebrate communities of arid-land headwater streams . Freshwater Biology 60 : 8232 ; 2547 – 2558 . Google Scholar Crossref Search ADS WorldCat Bogan MT , Lytle DA. 2011 . Severe drought and novel community trajectories in desert stream pools . Freshwater Biology 56 : 270 – 281 . Google Scholar Crossref Search ADS WorldCat Brown LE , Hannah DM, Milner AM. 2009 . ARISE: A classification tool for alpine river and stream ecosystems . Freshwater Biology 54 : 8232 ; 1357 – 1369 . Google Scholar Crossref Search ADS WorldCat Chezik KA , Anderson SC, Moore JW. 2017 . River networks dampen long-term hydrological signals of climate change . Geophysical Research Letters 44 : 7256 – 7264 . Google Scholar Crossref Search ADS WorldCat Cook BI , Ault TR, Smerdon JE. 2015 . Unprecedented 21st century drought risk in the American Southwest and Central Plains . Science Advances 1 : e1400082 – e1400082 . Google Scholar Crossref Search ADS PubMed WorldCat Cooper MG , Shaperow JR, Cooley SW, Alam S, Smith LC, Lettenmaier DP. 2018 . Climate elasticity of low flows in the maritime western U.S. mountains . Water Resources Research 8 : 5602 – 5619 . Google Scholar Crossref Search ADS WorldCat Cozetto K et al. . 2013 . Climate change impacts on the water resources of American Indians and Alaska Natives in the US. Climatic Change 120 : 569 – 584 . Google Scholar Crossref Search ADS WorldCat Crausbay SD et al. . 2017 . Defining ecological drought for the twenty-first century . Bulletin of the American Meteorological Society 98 : 2543 – 2550 . Google Scholar Crossref Search ADS WorldCat Dai A. 2013 . Increasing drought under global warming in observations and models . Nature Climate Change 3 : 52 – 58 . Google Scholar Crossref Search ADS WorldCat Davis JM , Baxter CV, Rosi-Marshall EJ, Pierce JL, Crosby BT. 2013 . Anticipating stream ecosystem responses to climate change: Toward predictions that incorporate effects via land–water links . Ecosystems 16 : 909 – 922 . Google Scholar Crossref Search ADS WorldCat Dawson TP , Jackson ST, House JI, Prentice IC, Mace GM. 2011 . Beyond predictions: Biodiversity conservation in a changing climate . Science 332 : 53 – 58 . Google Scholar Crossref Search ADS PubMed WorldCat DeWeber JT , Tsang Y, Krueger DM, Whittier JB, Wagner T, Infante DM, Whelan G. 2014 . Importance of understanding landscape biases in USGS gage locations: Implications and solutions for managers . Fisheries 39 : 155 – 163 . Google Scholar Crossref Search ADS WorldCat Diffenbaugh NS , Swain DL, Touma D. 2015 . Anthropogenic warming has increased drought risk in California . Proceedings of the National Academy of Sciences 112 : 3931 – 3936 . Google Scholar Crossref Search ADS WorldCat Dunham JB et al. . 2018 . Rivers are social–ecological systems: Time to integrate human dimensions into riverscape ecology and management . Wiley Interdisciplinary Reviews: Water 5 : e1291 . Google Scholar Crossref Search ADS WorldCat Falcone JA , Carlisle DM, Wolock DM, Meador MR. 2010 . GAGES: A stream gage database for evaluating natural and altered flow conditions in the conterminous United States . Ecology 91 : 621 . Google Scholar Crossref Search ADS WorldCat Fausch KD , Torgersen CE, Baxter CV, Li HW. 2002 . Landscapes to riverscapes: Bridging the gap between research and conservation of stream fishes . BioScience 52 : 483 – 498 . Google Scholar Crossref Search ADS WorldCat Foden WB et al. . 2013 . Identifying the world's most climate change vulnerable species: A systematic trait-based assessment of all birds, amphibians and corals . Plos One 8 ( art. e65427 ). Google Scholar OpenURL Placeholder Text WorldCat Frissell CA , Liss WJ, Warren CE, Hurley MD. 1986 . A hierarchical framework for stream habitat classification: Viewing streams in a watershed context . Environmental Management 10 : 199 – 214 . Google Scholar Crossref Search ADS WorldCat Fritz KM , Hagenbuch E, D’Amico E, Reif M, Wigington PJ Jr, Leibowitz SG, Comeleo RL, Ebersole JL, Nadeu J . 2013 . Comparing the extent and permanence of headwater streams from two field surveys to values from hydrographic databases and maps . Journal of the American Water Resources Association 49 : 867 – 882 . Google Scholar OpenURL Placeholder Text WorldCat Fullerton AH , Burnett KM, Steel EA, Flitcroft RL, Pess GR, Feist BE, Torgersen CE, Miller DJ, Sanderson BL 2010 . Hydrological connectivity for riverine fish: Measurement challenges and research opportunities . Freshwater Biology 55 : 2215 – 2237 . Google Scholar Crossref Search ADS WorldCat Grant PR , Grant BR. 2002 . Unpredictable evolution in a 30-year study of Darwin's finches . Science 296 : 707 – 711 . Google Scholar Crossref Search ADS PubMed WorldCat Gregory SV , Swanson FJ, McKee WA, Cummins KW. 1991 . An ecosystem perspective of riparian zones: Focus on links between land and water . BioScience 41 : 540 – 551 . Google Scholar Crossref Search ADS WorldCat Hand BK et al. . 2018 . A social–ecological perspective for riverscape management in the Columbia River basin . Frontiers in Ecology and the Environment 16 : S23 – S33 . Google Scholar Crossref Search ADS WorldCat Harpold AA , Dettinger M, Rajagopal S. 2017 . Defining snow drought and why it matters . Eos 98 . Google Scholar OpenURL Placeholder Text WorldCat Hauer FR , Locke H, Dreitz VJ, Hebblewhite M, Lowe WH, Muhlfeld CC, Nelson CR, Proctor MF, Rood SB. 2016 Gravel-bed river floodplains are the ecological nexus of glaciated mountain landscapes . Science Advances 2 : e1600026 . Google Scholar Crossref Search ADS PubMed WorldCat Hendry AP. 2016 . Eco-evolutionary dynamics . Princeton University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Isaak DJ et al. . 2017 . The NorWeST summer stream temperature model and scenarios for the western US: A crowd‐;sourced database and new geospatial tools foster a user community and predict broad climate warming of rivers and streams . Water Resources Research 53 : 8232 ; 9181 – 9205 . Google Scholar Crossref Search ADS WorldCat Jaeger KL , Olden JD, Pelland NA. 2014 . Climate change threatens desert fishes . Proceedings of the National Academy of Sciences 111 : 13894 – 13899 . Google Scholar Crossref Search ADS WorldCat Kennen JG , Stein ED, Webb JA. 2018 . Evaluating and managing environmental water regimes in a water-scarce and uncertain future . Freshwater Biology 63 : 733 – 737 . Google Scholar Crossref Search ADS WorldCat Kéry M , Schaub M. 2011 . Bayesian Population Analysis Using WinBUGS: A Hierarchical Perspective . Academic Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kovach RP , Muhlfeld CC, Al-Chokhachy R, Dunham JB, Letcher BH, Kershner JL. 2016 . Impacts of climatic variation on trout: A global synthesis and path forward . Reviews in Fish Biology and Fisheries 26 : 135 – 151 . Google Scholar Crossref Search ADS WorldCat Lake PS. 2003 . Ecological effects of perturbation by drought in flowing waters . Freshwater Biology 48 : 1161 – 1172 . Google Scholar Crossref Search ADS WorldCat Lake PS. 2011 . Drought and Aquatic Ecosystems: Effects and Responses . Wiley–Blackwell . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Labbe TR , Fausch KD. 2000 . Dynamics of intermittent stream habitat regulate persistence of a threatened fish at multiple scales . Ecological Applications 10 : 1774 – 1453 . Google Scholar Crossref Search ADS WorldCat Larned ST , Arscott DB, Schmidt J, Diettrich JC. 2010 . A framework for analyzing longitudinal and temporal variation in river flow and developing flow-ecology relationships . Journal of the American Water Resources Association 46 : 541 – 553 . Google Scholar Crossref Search ADS WorldCat Leasure DR et al. . 2018 . Hierarchical multi-population viability analysis . Ecology 100 : e02538 . Google Scholar Crossref Search ADS PubMed WorldCat Lowe WH , Likens GE. 2005 . Moving headwater streams to the head of the class . BioScience 55 : 196 – 197 . Google Scholar Crossref Search ADS WorldCat Lowe WH , Kovach RP, Allendorf FW. 2017 . Population genetics and demography unite ecology and evolution . Trends in Ecology and Evolution 2 : 141 – 152 . Google Scholar Crossref Search ADS WorldCat Lytle DA , Poff NL. 2004 . Adaptation to natural flow regimes . Trends in Ecology and Evolution 19 : 94 – 100 . Google Scholar Crossref Search ADS PubMed WorldCat Magoulick DD , Kobza RM. 2003 . The role of refugia for fishes during drought: A review and synthesis . Freshwater Biology 48 : 8232 ; 1186 – 1198 . Google Scholar Crossref Search ADS WorldCat Matthews WJ , Matthews EM. 2003 . Effects of drought on fish across axes of space, time, and ecological complexity . Freshwater Biology 48 : 1232 – 1253 . Google Scholar Crossref Search ADS WorldCat McCabe GJ , Wolock DM. 2014 . Spatial and temporal patterns in conterminous United States streamflow characteristics . Geophysical Research Letters 41 : 6889 – 6897 . Google Scholar Crossref Search ADS WorldCat McEvoy J et al. . 2018 . Ecological drought: Accounting for the non-human impacts of water shortage in the upper Missouri headwaters basin, Montana, USA . Resources 7 : 14 . Google Scholar Crossref Search ADS WorldCat Mills LS , Lindberg MS. 2002 . Sensitivity analysis to evaluate the consequences of conservation actions . Pages 338–366 in Beissinger SR, McCullough DR, eds. Population Viability Analysis . University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Montgomery DR. 1999 . Process domains and the river continuum . Journal of the American Water Resources Association 35 : 397 – 410 . Google Scholar Crossref Search ADS WorldCat Muhlfeld CC et al. . 2014 . Invasive hybridization in a threatened species is accelerated by climate change . Nature Climate Change 4 : 620 – 624 . Google Scholar Crossref Search ADS WorldCat Naiman RJ , Decamps H, Pollock M. 1993 . The role of riparian corridors in maintaining biodiversity . Ecological Applications 3 : 209 – 212 . Google Scholar Crossref Search ADS PubMed WorldCat Nicotra AB , Beever EA, Hofman GE, Robertson AL, O’Leary J. 2015 . Assessing the components of adaptive capacity to improve conservation and management efforts under global change . Conservation Biology 29 : 1268 – 1278 . Google Scholar Crossref Search ADS PubMed WorldCat Osborn TJ , Barichivich J, Harris I, van der Schrier G, Jones PD. 2017 . Monitoring global drought using the self-calibrating Palmer Drought Severity Index . Bulletin of the American Meteorological Society 98 : S32 – S33 . Google Scholar OpenURL Placeholder Text WorldCat Pallard B , Castellarin A, Montanari A. 2009 . A look at the links between drainage density and flood statistics . Hydrology and Earth System Sciences 13 : 1019 – 1029 . Google Scholar Crossref Search ADS WorldCat Payn RA , Gooseff MN, McGlynn BL, Bencala KE, Wondzell SM. 2009 . Channel water balance and exchange with subsurface flow along a mountain headwater stream in Montana, United States . Water Resources Research 45 : W11427 . Google Scholar Crossref Search ADS WorldCat Pederson GT , Betancourt JL, Mccabe GJ. 2013 . Regional patterns and proximal causes of the recent snowpack decline in the Rocky Mountains, US . Geophysical Research Letters 40 : 1811 – 1816 . Google Scholar Crossref Search ADS WorldCat Pickett STA , Thompson JN. 1978 . Patch dynamics and the design of nature reserves . Biological Conservation 13 : 27 – 35 . Google Scholar Crossref Search ADS WorldCat Poff NL , Zimmerman JKH. 2010 . Ecological responses to altered flow regimes: A literature review to inform the science and management of environmental flows . Freshwater Biology 55 : 194 – 205 . Google Scholar Crossref Search ADS WorldCat Poff NL , Allan JD, Bain MB, Karr JR, Prestegaard KL, Richter BD, Sparks RE, Stromberg JC. 1997 . The natural flow regime: A paradigm for river conservation and restoration . BioScience 47 : 769 – 784 . Google Scholar Crossref Search ADS WorldCat Power ME , Bouma-Gregson K, Higgins P, Carlson SM. 2015 . The thirsty eel: Summer and winter flow thresholds that tilt the Eel River of northwestern California from salmon-supporting to cyanobacterially degraded states . Copeia 103 : 200 – 211 . Google Scholar Crossref Search ADS WorldCat Renard B , Kavetski D, Kuczera G, Thyer M, Franks SW. 2010 . Understanding predictive uncertainty in hydrologic modeling: The challenge of identifying input and structural errors . Water Resources Research 46 : W05521 . Google Scholar Crossref Search ADS WorldCat Rieman BE , Dunham JB. 2000 . Metapopulations and salmonids: A synthesis of life history patterns and empirical observations . Ecology of Freshwater Fish 9 : 51 – 64 . Google Scholar Crossref Search ADS WorldCat Ruhi A , Holmes EE, Rinne JN, Sabo JL. 2015 . Anomalous droughts, not invasion, decrease persistence of native fishes in a desert river . Global Change Biology 21 : 1482 – 1496 . Google Scholar Crossref Search ADS PubMed WorldCat Ruhi A , Messager ML, Olden JD. 2018 . Tracking the pulse of earth's freshwaters . Nature Sustainability 1 : 198 – 203 . Google Scholar Crossref Search ADS WorldCat Schultz LD , Heck MP, Hockman-Wert D, Allai T, Wenger S, Cook NA, Dunham JB. 2017 . Spatial and temporal variability in the effects of wildfire and drought on thermal habitat for a desert trout . Journal of Arid Environments 145 : 60 – 68 . Google Scholar Crossref Search ADS WorldCat Seneviratne SI et al. . 2012 . Changes in climate extremes and their impacts on the natural physical environment . Pages 109–230 in Field CB, et al. , eds. Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation . Cambridge University Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Smakhtin V. 2001 . Low flow hydrology: A review . Journal of Hydrology 240 : 147 – 186 . Google Scholar Crossref Search ADS WorldCat Snyder CD , Hitt NP, Young JA. 2015 . Accounting for groundwater in stream fish thermal habitat responses to climate change . Ecological Applications 25 : 1397 – 1419 . Google Scholar Crossref Search ADS PubMed WorldCat Steel EA , Beechie TJ, Torgersen CE, Fullerton AE. 2017 . Envisioning, quantifying, and managing thermal regimes on river networks . BioScience 67 : 506 – 522 . Google Scholar Crossref Search ADS WorldCat Steinschneider S , Polebitski A, Brown C, Letcher B. 2012 . A statistical framework to test the significance of hydrologic alteration under future climate scenarios . Water Resources Research 48 : WR011318 . Google Scholar OpenURL Placeholder Text WorldCat Tonkin JD , Merritt DM, Olden JD, Reynolds LV, Lytle DA. 2018 . Flow regime alteration degrades ecological networks in riparian ecosystems . Nature Ecology and Evolution 2 : 86 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat Torgersen CE , Price DM, Li HW, McIntosh BA. 1999 . Multiscale thermal refugia and stream habitat associations of chinook salmon in northwestern Oregon . Ecological Applications 9 : 301 – 319 . Google Scholar Crossref Search ADS WorldCat Vannote RL , Minshall GW, Cummins KW, Sedell JR, Cushing CE. 1980 . The river continuum concept . Canadian Journal of Fisheries and Aquatic Sciences 37 : 130 – 137 . Google Scholar Crossref Search ADS WorldCat Van Loon AF. 2015 . Hydrological drought explained . Wiley Interdisciplinary Reviews: Water 2 : 359 – 392 . Google Scholar Crossref Search ADS WorldCat Vorosmarty CJ , Green P, Salisbury J, Lammers RB. 2000 . Global water resources: Vulnerability from climate change and population growth . Science 289 : 284 – 289 . Google Scholar Crossref Search ADS PubMed WorldCat Wade AA , Hand BK, Kovach RP, Luikart G, Whited DC, Muhlfeld CC. 2017a . Accounting for adaptive capacity and uncertainty in assessments of species’ climate-change vulnerability . Conservation Biology 31 : 136 – 149 . Google Scholar Crossref Search ADS WorldCat Wade AA , Hand BK, Kovach RP, Muhlfeld CC, Waples RS, Luikart G. 2017b . Assessments of species’ vulnerability to climate change: From pseudo to science . Biodiversity and Conservation 36 : 8232 ; 223 – 229 . Google Scholar Crossref Search ADS WorldCat Wahl KL , Thomas WO, Hirsch RM. 1995 . The Stream-Gaging Program of the US Geological Survey . US Geological Survey . https://pubs.usgs.gov/circ/circ1123 . Google Scholar OpenURL Placeholder Text WorldCat Weekes AA , Torgersen CE, Montgomery DR, Woodward A, Bolton SM. 2012 . A process-based hierarchical framework for monitoring glaciated alpine headwaters . Environmental Management 50 : 982 – 997 . Google Scholar Crossref Search ADS PubMed WorldCat Wilhite DA , Glantz MH. 1985 . Understanding the drought phenomenon: The role of definitions . Water International 10 : 111 − 120 . Google Scholar Crossref Search ADS WorldCat Winter TC , Rosenberry DO, LaBaugh JW. 2003 . Where does the ground water in small watersheds come from? Groundwater 41 : 989 – 1000 . Google Scholar Crossref Search ADS WorldCat Wiens JA , Bachelet D. 2010 . Matching the multiple scales of conservation with the multiple scales of climate change . Conservation Biology 24 : 51 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat Woodward G et al. . 2015 . The effects of climatic fluctuations and extreme events on running water ecosystems . Philosophical Transactions of the Royal Society B 371 : 20150274 . Google Scholar Crossref Search ADS WorldCat Wondzell SM , Gooseff MN, McGlynn BL. 2007 . Flow velocity and the hydrologic behavior of streams during baseflow . Geophysical Research Letters 34 : L24404 . Google Scholar Crossref Search ADS WorldCat Zipkin EF , Thorson JT, See K, Lynch HJ, Grant EHC, Kanno Y, Chandler RB, Letcher BH, Royle JA. 2014 . Modeling structured population dynamics using data from unmarked individuals . Ecology 95 : 22 – 29 . Google Scholar Crossref Search ADS PubMed WorldCat Published by Oxford University Press on behalf of American Institute of Biological Sciences 2019 This work is written by (a) US Government employee(s) and is in the public domain in the US. Published by Oxford University Press on behalf of American Institute of Biological Sciences 2019
TI - An Integrated Framework for Ecological Drought across Riverscapes of North America
JF - BioScience
DO - 10.1093/biosci/biz040
DA - 2019-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/an-integrated-framework-for-ecological-drought-across-riverscapes-of-ocBefss9R0
SP - 418
EP - 431
VL - 69
IS - 6
DP - DeepDyve
ER -