Vernal pools are shallow, seasonal wetlands, underlined by an impermeable or slowly permeating layer, that form in depression pools. Vernal pools in California have been decimated by land developments. This has caused increased habitat fragmentation for vernal pool species. A thorough understanding of the vernal pools hydrology is imperative to their continued survival, as they are important refugia for endangered, endemic vernal pool species. The purpose of my research is to study the hydrological connectivity of two three–vernal pool series. My results show that vernal pools are hydrologically connected to each other and to the landscape. There are several variables that determine its degree of hydrological connectivity– shallow subsurface confining layer, perched groundwater and surface water flow, inundation, and location. Understanding the hydrology of this unique environment is imperative to the success of vernal pools and its species, as they undergo habitat loss and stresses associated with climate change.

The work presented here was motivated by the importance of the Sierra Nevada to California’s water resources and the need for understanding fluxes of water across expanded spatial and temporal scales through the critical zone, the area from bedrock to atmosphere, in which fluxes of energy and water support life on Earth. A new set of isotopic data is presented, characterizing each component of the hydrologic cycle (precipitation, evapotranspiration, runoff and storage), on two scales: headwater catchment scale and mountain range scale. These analyses are preceded by analyses of factors affecting precipitation signatures. From initial precipitation, links between headwater catchment evapotranspiration sources, subsurface storage and runoff, were tracked and, finally, the source-elevation of river water was determined across the entire Sierra Nevada using isotopic data from major rivers.

Precipitation input stable isotope signatures represent the starting point to track water through the critical zone’s atmosphere – vegetation – subsurface continuum and allow interpretation of streams, deeply stored saturated zone water, soil water and xylem water signatures. Elevation, season and canopy interception are factors affecting precipitation stable isotope signatures in the Sierra Nevada. Precipitation 𝛿18O signatures decreased with elevation at a rate of -3.3 (± 1.8) ‰/km in the northern central Sierra Nevada and -2.8 (± 1.8) ‰/km in the southern central Sierra Nevada. By measuring a total of 31 isotopic lapse rates along two elevation transects, temporal and spatial variability was observed. Individual isotopic lapse rates were compared to mean amount weighted lapse rates to determine that individual lapse rates do not represent the aggregate input into the system, concluding that aggregate lapse rates should be measured. Similarly, the Local Meteoric Water Line (LMWL) in the Sierra Nevada changed seasonally, but lacked interannual variability. The LMWL for the entire Sierra Nevada had a slope of 7.2 (± 0.1) and an offset of 3.1 (± 1.2). Seasonality in the Central Sierra Nevada was modeled with a sinusoidal function, which varied with elevation. Snowmelt signatures beneath canopy and in open areas were measured every ten days throughout two snowmelt seasons, showing that canopy interaction significantly affected snowmelt stable isotope signatures.

By measuring both water stable isotopes and tritium (3H) in each component of the hydrologic cycle on the catchment scale, it was determined that although evapotranspiration and deeper subsurface saturated zone water originated as snowmelt, vegetation in all seasons used a younger component of water compared to deeper subsurface saturated zone water. Even when stable isotope signatures of xylem water and saturated zone water matched, 3H data showed that the age of xylem water and saturated zone water were distinct, demonstrating the utility of using multiple tracers to track water through vegetation. Likewise, water stable isotopes 𝛿18O and 𝛿2H showed that vegetation responded to new water inputs over time, while saturated zone water did not, further confirming that vegetation accessed younger water compared to saturated zone water. These findings contribute to current debate in recent literature regarding the concept of ecohydrological separation, in which it is argued that vegetation uses a separate source water compared to runoff. Additionally, by understanding how vegetation source water is connected to saturated zone water on this expanded temporal scale, the implication is that older water provides drought resilience to runoff from the saturated zone and new precipitation provides drought resilience to vegetation.

Previous studies have shown that water from upper elevations in the Sierra Nevada contributed a disproportionate amount of water to runoff, partially due to additional subsurface stores of water and lower evapotranspiration at upper elevations. This dissertation uses isotopic data to confirm these previous findings through new Sierra Nevada river 𝛿18O and 𝛿2H data applied to isotopic lapse rates. Even across geologically heterogeneous catchments, river source waters originated above their mean catchment elevation, with higher elevation source waters correlating to higher elevation catchments. Isotopic results agree with results and findings derived from a spatially distributed mass balance approach.

The work presented herein examines how and to what extent turbidity, channel bed movement and stream source water ratios change over different temporal scales in four forested mountain headwater catchments in the Sierra Nevada, California. This work focuses on the water quality topics of sediment and source waters as part of a larger study on the effects of forest fuels treatments. No effects on water quality were expected because treatments were set far back from the stream channel and relatively light (8.0% and 7.5% decrease in mean Leaf Area Index (LAI) for the northern and southern site respectively). Therefore, this work was designed to characterize sediment and source waters rather than explore treatment impacts. In doing so important insights were gained around sediment and stream water sources.

Sediment movement was found to be highly episodic and tied to low-frequency, short-duration discharge events with pronounced seasonality. Turbidity hysteresis patterns indicated localized (bed and banks) sediment sources that were mobilized quickly but became progressively depleted. Continuous measurement of channel bed elevations in the thalweg pointed to seasonal mobilization (connectedness) of sediment during winter, a gradual depletion through spring and early summer, and disconnectedness in low flow season when material builds up again at the base of banks mirroring patterns seen in the turbidity hysteresis loops. Stream water chemistry data showed higher concentrations during base flow, which was exacerbated by consecutive years of drought. The high base flow ion concentrations were tied to high ratios of groundwater. End-member mixing analysis showed that wet and dry years had similar source water ratios during high flow (winter and spring) but during the low flow season (late summer), a shift toward higher ratios of groundwater was seen for drought years. Recent tree mortality observed across the Sierra Nevada after four consecutive years of drought underscores the critical role of groundwater in maintaining baseflow and sustaining forest ecosystems. Better understanding of sediment processes, source water contributions, and drought effects in small, forested, mountain, headwater catchments provides an important foundation for sustainable land use management, more effective channel restoration design, and improved mitigation of downstream water quality.

The Merced River, originating in the Sierra Nevada in California, drains a watershed with an area of approximately 3,305 km2. The stream has been highly altered due to diversions, gold-dredged mining, damming, and subsequent modification to the hydrograph. Over the course of a year, groundwater-surface water interactions were studied to elucidate the hydrological connection between the Main Canal, an unlined engineered channel containing Merced River water and flowing parallel to the river (average elevation 89 m) and also a highly conductive previously-dredged floodplain, and the Merced River average (elevation 84 m). Upstream of the study reach, located in an undredged portion of the floodplain, are a state run salmon hatchery and a privately run trout farm that have been operating for approximately 40 years. Exchanges between the hyporheic and surrounding surface, groundwater, riparian, and alluvial floodplain habitats occur over a wide range of spatial and temporal scales. For this study, pressure transducers were installed in seven wells and four ponds located in the dredged floodplain. All wells were drilled to the Mehrten Formation, a confining layer, and screened for the last 3 m. These groundwater well water levels as well as the surface water elevations of the Main Canal and the Merced River were used to determine the direction of sublateral surface flows using Groundwater Vistas as a user interface for MODFLOW. The wells, the canal, the river, and seepage from the river banks were sampled for major anion and cation, dissolved organic carbon, total nitrogen, total iron, and total dissolved iron concentrations to determine water sources and the possibility of suboxic water. Field analysis indicated that water in all wells and ponds exhibits low dissolved oxygen, high conductivity, and oxidation/reduction potentials that switched from oxidizing to reductive during the course of the monitoring period. Chemical analysis indicates that there are three sources of water for this floodplain: the Merced River and Main Canal (which are chemically very similar), the waters from the trout farm, and precipitation. The well closest to the trout farm had a C:N of 1, typical of a highly carbon-limited system. MODFLOW particle tracer experiments designed with homogeneous soils were performed and did not capture the near-surface preferential flow paths. These results indicate that travel time between the Main Canal and Merced River is approximately 10-15 years, while chemistry results indicate seasonal fluctuations. Based on the well levels and chemistry, this water system responds on a much faster scale than indicated by the particle tracer experiments. Reconciling these results, there must be significant preferential flow paths. Candidate flow paths are abandoned channels from the dredging era. The hydraulic gradient set up by the groundwater connection between Main Canal and the Merced River ensures that any effluent released by the trout farm will be transported to the Merced River. Conclusions of the year-long study are that the waters that seep from the Main Canal to the Merced River in this area can be suboxic, which is not conducive to spawning and incubation for native Chinook salmon (Oncorhynchus tshawytscha) an indicator species for the overall ecosystem's health. This study reach has been historically important for salmon spawning and rearing, as the area examined is where more than 50% of the Chinook salmon of the Merced River spawn. Currently, salmon restoration efforts are focusing on gravel augmentation and adding a side channel, but ignoring groundwater influences. Due to the causal connections between the hydrological system of the Merced River floodplain and the riverine system, habitat rehabilitation should target not only the surface water but also important subsurface hydrological components.

Meadows comprise less than 1 percent of the land area of the Sierra Nevada, however, they play an important role, often linking the forested hillslope to the stream, in hydrologic processes of Sierra Nevada catchments. This study investigates evapotranspiration and groundwater patterns in montane meadows in the Sierra Nevada, CA. The dissertation is divided into three sections. 1) The first section uses direct eddy covariance measurements and Applied Neural Network models in order to determine annual evapotranspiration (ET) from 2008 to 2014; annual ET rates in Long Meadow ranged from 321 to 392 mm and was higher in warm, dry years. Based on literature estimates of ET increases of 25-50%, due to meadow restoration, the overall increase in catchment ET, due to restoration efforts, would only amount to 0.03-0.05%. 2) The second section uses in situ measurements of groundwater table fluctuations and evapotranspiration in order to constrain soil specific yield. This study found that the variability of soil specific yield generally decreased with increasing depth to water; variability was also minimized when the surface slope was between 3-7 degrees. A method is proposed using these parameters for selecting monitoring well locations to estimated soil specific yield and, subsequently, ET for relatively low cost and impact. This method was demonstrated at 5 locations in two additional meadow systems. Average daily ET from July 1-31, 2013 ranged from 2.7 to 6.3 mm d-1 in these two meadows, compared to an average daily ET rate measured in Long Meadow of 4.6 mm d-1. 3) The third section investigates groundwater sources and patterns in a montane meadow and how the meadow links up-gradient hydrologic sources to the down-gradient stream. This was accomplished using a combination of physical measurements—including hydrostatic head, stream flow, snow depth, and soil moisture content, and analysis of water chemistry—including major ions, alkalinity, and stable and radioactive isotopes of water. Results show that meadow groundwater is comprised of two sources, a shallow subsurface flow with groundwater gradient signals dependent on snowpack, and a deeper subsurface flow that persists through the growing season—even after several years of drought. Individually and in combination, the studies in this dissertation significantly add to the knowledge of meadow hydrology in the Sierra Nevada. Both the first and third chapters found meadow systems still being supplied with groundwater, even after several years of the current drought in California. This suggests that meadow resilience in the face of drought is tied to significant groundwater storage capacity within a meadow catchment. Systems where deeper groundwater sources can be identified through analysis of stream chemistry or where significant groundwater storage capacity can be identified through geophysics or physiographic analysis should be prioritized for restoration efforts.

We observed polymictic behavior in stream pools in a low gradient montane meadow in the southern Sierra Nevada Mountains, California. Thermal stratification in stream pools has been observed in various environments; stratification generally persists where the buoyancy forces created by a variation in water density, as a function of water temperature, are able to overcome turbulent forces resulting from stream flow. Because the density gradient creates a relatively weak buoyancy force, low flow conditions are generally required in order to overcome the turbulent forces. In some studies, a cold water source in to the pool bottoms can help to increase the density gradient and perpetuate thermal stratification. Our study took place in Long Meadow, Sequoia National Park, California. Long Meadow lies in the Wolverton Creek watershed and is part of the Southern Sierra Critical Zone Observatory. The 1-4 m diameter and 1-2 m deep pools in our study stratified thermally during the day and mixed completely at night. The low gradient of the meadow provided low stream flows. Piezometers in the meadow indicated groundwater discharge into the meadow in the months during which stratification was observed. Radon-222 activity measured in the pools also indicated groundwater influx to the pool bottoms. We used Fluent, a computational fluid dynamics equation solver, to construct a model of one of the observed pools. Initially we attempted to model the physical mechanisms controlling thermal stratification in the pool including stream flow, groundwater discharge, solar radiation, wind speed, and air, stream and ground water temperatures. Ultimately we found the model best agreed with our observed pool temperatures when we considered the light attenuation coefficients as a function of the dissolve organic carbon (DOC) concentration. Elevated DOC concentrations are expected in low stream flow regimes associated with highly organic soils such as a montane meadow. DOC concentrations measured in samples collected from the meadow stream, pools, and ground water wells ranged from 3.09 to 5.25 mg/L. We used a power equation taken from the literature to vary the visible light attenuation with DOC values measured in the meadow system. Light attenuation coefficients determined from measured DOC concentrations ranged from 0.507 m-1 to 0.899 m-1. Conditions consistent with the meadow environment--low gradient and stream flow, high solar loading, groundwater discharge, and high productivity--are such that thermal stratification of stream pools may be prolific. Understanding how these parameters influence thermal stratification of stream pools may allow investigators to appropriately assess thermal stratification when using large scale measurements, such as DTS or FLIR, and may help assess stream pool utilization by temperature sensitive species.