Fast and invisible: Conquering Subsurface Stormflow
through an Interdisciplinary Multi-Site Approach
(DFG RU 5288)

Abstract:

Extreme rainfall events are likely to increase in intensity and frequency due to climatic changes and therefore the forecast of flooding events will become more important in the following decades. The flow properties of rainwater in the subsurface play a critical role in the flood formation process, but the underlying mechanism of this subsurface stormflow (SSF) formation has not been fully understood so far. Here, we explore the viability of environmental DNA (eDNA) as an indicator for small-scale flow pathway reconstruction. eDNA comprises genetic signatures from organisms across the Tree of Life (ToL), from whole microorganisms to molecular traces of higher taxa, such as plants or animals. The degree of similarity of biodiversity patterns indicates biological and therefore, in principle, also hydrological connectivity. As part of the SSF Research Unit we characterised 3 trenched hillslopes in 4 catchment areas in Germany and Austria through eDNA ToL-metabarcoding. With this broad-range approach, we aim to understand whether and how eDNA diversity patterns can inform subsurface flow pathways. We found three-dimensional connectivity patterns of biodiversity indicating systematic barriers as well as pathways of hydrological connectivity within each hillslope. Variation between catchments reflects their geographic differences as well as geological peculiarities. Although our results support the potential of eDNA to identify flow pathways and enhance our understanding of SSF, we are still at the beginning of understanding the viability of eDNA as a tracer in hydrological research. Nonetheless, making use of such natural occurring tracers can extend our understanding of hydrological phenomena and can contribute to a more accurate flood prediction.

Abstract:

Rainfall runoff contributes to a large proportion of the discharge in streams and therefore, heavily influences stream water quality but also flood generation. Rainfall runoff generation is usually a combination of overland and subsurface flow processes, the latter of which being especially difficult to trace. Here, we explored the viability of environmental DNA (eDNA) for subsurface water flow pathway reconstruction and simultaneous biodiversity assessment. The degree of similarity of community patterns indicates biological and therefore, in principle, also hydrological connectivity. We applied eDNA metabarcoding to characterise 10 drilling cores (0.7-3.2 m depth) on 3 hillslopes (10x50 m) in 4 catchment areas in Germany and Austria. In total, more than 2000 species across taxonomic groups could be identified down to species level. Analysis of alpha and beta diversity in the different catchments showed significant differences in spatial clustering patterns between taxonomic groups, but also between geomorphological and geochemical properties, such as the composition of dissolved organic carbon, of the respective catchment. We could assign indicator species sets in all taxonomic groups to various depth layers and identify habitat-specific communities that can be used as hydrological tracers. Although our results support the potential of eDNA to identify flow pathways and enhance our understanding of subsurface flow processes, we are still at the beginning of understanding the viability of eDNA as a tracer in hydrological research. However, our results show that making use of such naturally occurring tracers can expand our understanding of hydrological phenomena, especially those hidden in the subsurface.

Abstract:

The transport of water-soluble organic matter (WSOM) during stormflow events is an important link between hillslope hydrology and biogeochemical cycling, as it determines the movement of organic carbon from soils to streams. Hydrological dynamics in hillslopes, particularly subsurface stormflow (SSF), exhibit substantial spatial and temporal variability, making quantification and generalization challenging. SSF can account for up to 90% of rainfall input to stream discharge during storm events, highlighting its importance in catchment hydrology. Despite its significance, current research frequently overlooks WSOM dynamics during SSF, which are not only key components of carbon cycling but may also serve as tracers for identifying potential critical source areas. This emphasizes the importance of studying hillslope hydrological dynamics and determining the factors that contribute to SSF spatial and temporal variability. Furthermore, the specific flow paths within hillslopes remain poorly understood, which complicates the identification of spatial sources and transport mechanisms for organic carbon. To fill these knowledge gaps, we conducted a field study in the Black Forest, Germany, using a trench system to collect lateral subsurface flows at two depths (0-100 cm and 100-200 cm) over several rain events. We analysed WSOM concentration and quality using absorbance and fluorescence properties to assess the variability in critical source areas. We also conducted isotopic analyses of oxygen (δ¹⁸O) and hydrogen (δ²H) of the same water samples to infer flow pathways with a conservative tracer. This approach provides valuable insights into the temporal dynamics and spatial heterogeneity of SSF. Our findings will contribute to our understanding of flow paths, transit times, and the characteristics of WSOM export, offering a deeper understanding of subsurface flow processes in catchments. Finally, the findings of this study can help to improve biogeochemical models and improve scaling of hillslope processes models, particularly in understanding their contribution to organic carbon transport via SSF.

Abstract:

Hydrological dynamics of hillslopes, particularly subsurface stormflow (SSF), are highly complex and variable in space and time. Frequently, available studies are often limited to single slopes or few storm events. As a result, the transfer of these findings to other slopes or catchments is associated with great uncertainties. Thus, for upscaling and model validation, a quantification of the hydrological dynamics of hillslopes and the factors influencing the spatial and temporal patterns of subsurface stormflow is urgently needed. Closely related to the hydrological dynamics of hillslopes is the export of organic carbon from the soils to the adjacent stream. However, the spatial sources of carbon are still largely unclear because the exact flow paths of SSF within the slope are not well known. In order to address this knowledge gap, we took a hydro-biogeochemical approach, that measures the water-soluble organic matter (WSOM; concentration, absorbance and fluorescence) along 480 locations on 100 hillslopes, in four contrasting catchments – varying from low to high mountain ranges (Sauerland, Ore Mountains, Black Forest, Alps). This enables us to derive empirical relations among different landforms (i.e., convergent, divergent and planar slope shapes, flow path lengths and valley shapes), bedrock and soil properties, and to quantify the spatial variability and stability of subsurface hydrological process patterns (e.g., flow directions, transit times, hydrochemical and biochemical composition). Distributed sampling of WSOM along the soil profile (6 WSOM samples per profile; both during wet and dry conditions) will help to assess the vertical and lateral subsurface flowpaths of water in the unsaturated and saturated zone, and the spatial discretization of source areas for SSF. We will use an array of state-of-the-art laboratory equipment and methods (TOC-Analyzer, Fluorescence Spectrometry) to analyze WSOM. First results will show depth profiles of WSOM in the four contrasting catchments from the low to high mountain ranges (Sauerland, Ore Mountains, Black Forest, Alps). By these depth profiles source areas of SSF can be detected.

Abstract:

Subsurface stormflow (SSF) is a key runoff generation mechanism in small catchments, yet its dynamics and thresholds remain poorly understood due to the challenges of observing and measuring subsurface processes. To address this, we conducted a study in a headwater catchment in the Black Forest, Germany, where we investigated the depth of SSF flow paths, if SSF is dominated by event water and if all SSF events have similar tracer behavior. Three trenches were installed on slopes with different landuse and topography. They were excavated down to bedrock ( 15 m wide, 2.5 m deep), collecting SSF separately from upper and lower soil layers. Continuous measurements of SSF volume and temperature were combined with water sampling for natural tracers: stable water isotopes, dissolved organic carbon, electric conductivity, and major ions. During the study period, 6 large SSF events were recorded in each trench. Using heat as a tracer we found stable SSF flowpath depths across events. Multitracer analysis suggested that SSF consisted mostly of pre-event water, with antecedent wetness governing event water contributions. Our results also show that C–Q relationships for solutes varied considerably between events and trenches, with some (e.g., Si) remaining chemostatic while others (e.g. DOC, SO24− and NO−3 ) behaved chemodynamically. This indicates that the chemical composition of SSF depends on both the antecedent hydrometeorological conditions and local factors such as landuse, reinforcing the complexity of these processes. Our results highlight the dynamic nature of subsurface stormflow and the challenges involved in its characterization.

Abstract:

Headwater streams account for 70% or more of total stream length in most catchments, making it crucial to better understand the processes and controlling factors governing streamflow generation as well as water quality. In this context, stream water chemistry longitudinal profiles can provide valuable insights. This study examines longitudinal stream chemistry profiles across six headwater catchments in three mid-mountain ranges in Germany: The Ore Mountains (catchments OM 1 and OM 2), Black Forest (BF 1 and BF 2), and Sauerland (SL 1 and SL 2). Three to four snapshot sampling campaigns were conducted per catchment across different seasons and catchment wetness conditions. During the campaigns, water samples were collected from 22 stream monitoring points in the Ore Mountains catchments, 14 in the Black Forest, and 14 in Sauerland, and the samples were analyzed for major cations, anions, and dissolved organic carbon. Subsequently, the longitudinal profiles observed were grouped into spatial and temporal patterns. In the Ore Mountains, solute concentrations were generally stable over time. However, the spatial patterns varied between the two neighbouring catchments (OM 1 and OM 2). OM 2 exhibited chemostatic longitudinal profiles for most solutes, while OM 1 showed pronounced spatial variability in solutes such as nitrate, dissolved organic carbon (DOC), chloride, and sodium. This variability is usually linked to monitoring points located near springs, tributaries, and drainage systems. However, some spikes in ion concentrations along the stream were not linked to these obvious inflows, thus potentially indicating hotspots for groundwater inflow. The Sauerland catchments showed elevated concentrations of DOC, magnesium, calcium, and sodium in July 2023, a period associated with lower streamflow. An increase in concentration from upstream to downstream was here seen in both streams for solutes like calcium and sodium, during all snapshot campaigns. However, other solutes, like nitrate and sulfate, showed different longitudinal patterns and notable shifts in solute concentration during the snapshot campaigns in SL 2. The shifts in patterns indicate a dependency on time-variant factors like seasonal changes in water input, and land use practices. BF 1 catchment in the Black Forest showed a decreasing pattern in DOC, from upstream to downstream, while the neighbouring catchment BF 2 showed a chemostatic trend. These trends could be influenced by the land use changes within the catchments. Notable increased nitrate concentrations were seen along reaches adjacent to grassland areas and at sampling points near tile drains in OM 1, BF 1, SL 1, and SL 2. Overall, solute spatial and temporal patterns were stream-specific, with no universal behaviour observed across all catchments. This variability likely results from the interplay of factors such as geology, soils, land use, stream morphology, and climate. High-resolution spatial sampling enabled the identification of point sources and hotspots of groundwater inflow which could be missed by sparse sampling. These findings enhance our understanding of the processes regulating water quality and flow in headwater systems, providing a basis for better management of these systems.

Abstract:

Hydrologic connectivity is the linkage of separate regions of a catchment via water flow. Knowledge of hillslope–stream connectivity (both at the surface and in the subsurface) is essential for understanding and predicting runoff responses and streamwater quality. Connectivity can be very dynamic: hillslopes may connect to the stream only during certain events or seasons. While surface connectivity is often discussed, particularly in the context of sediment transport, subsurface connectivity is more difficult to describe and assess. This difficulty has led to a wide variety in methodologies that are used in various contexts. Field approaches have focused on intensive monitoring of processes on the hillslope or the fingerprint of connectivity in the stream. Combining experimental studies with modeling allows for testing of hypotheses with respect to thresholds and controls on connectivity, and extrapolation from the hillslope scale to the catchment scale. However, as most modeling approaches are based on datasets from a few intensively studied hillslopes, this carries the inherent risk of oversimplification because it assumes that the observed hillslope responses are representative for the catchment or even the region. Focussed efforts on catchment scale assessment of hillslope–stream connectivity, as well as site intercomparisons and the search for similarity measures may allow us to capture the wider picture of the mechanisms and factors that control hillslope–stream connectivity, and its effects on flow and transport at the catchment scale. This overview focuses on how hillslope–stream connectivity has been studied and describes the advantages, disadvantages, and challenges of the different methods. WIREs Water 2015, 2:177–198. doi: 10.1002/wat2.1071 This article is categorized under: Science of Water > Hydrological Processes Science of Water > Water Quality