The wetlands of the Sierra Nevada were formed and are maintained by a feedback between soil, plant, and hydrologic processes. Primary production of plants builds soil organic matter and plant roots bind soil, preventing erosion during flooding. In turn, soil organic matter retains water and nutrients that support plant growth, while the hydrologic regime regulates soil organic matter decomposition, plant community makeup, and plant production. The relative stability of these interacting processes has built thick meadow soils over the past several thousand years. However, modern human impacts such as livestock grazing and water extraction have decoupled the interacting processes. Removal of plants by grazers exposes soil to water erosion and reduces production, the source of soil organic matter. Erosion gully formation and direct water extraction lower the wetland water table, speeding soil organic matter decomposition, altering plant community composition, and reducing production. Gully formation and loss of soil organic matter occur rapidly but are extremely slow to reverse by natural processes alone. Wetlands that have experienced these impacts enter alternative stable states that will not quickly return to their original configurations. In these cases, ecological restoration is necessary to repair human impacts and reestablish the stabilizing feedback of soil, plant, and hydrologic processes. This dissertation is composed of five chapters that explore wetland ecosystem function and restoration in the Sierra Nevada.

Chapter 1 is an overview of Sierra Nevada fens: groundwater-supported peat-accumulating wetlands. Underlying geology and topography exert strong control over the distribution and vegetation of these ecosystems. Groundwater chemistry is largely determined by watershed rock type and is a significant determinant of plant species composition. Wide-ranging values of pH (4.28–8.00) and dissolved cation concentrations (1.6–62.0 mg/L) span the categories of transitional poor–rich to extremely rich fens. Species richness is primarily (and negatively) correlated with altitude. Peat thickness (15–253 cm) is constrained in smaller catchments and on steeper slopes, and is positively correlated with soil organic matter content (16–92 %).

Chapter 2 describes a field exclosure experiment testing the effects of native deer and rodent herbivory on the meadow and streamside willow vegetation of Tuolumne Meadows. Streamside willows protected from deer gained a net average 16.7 ± 7.0 cm height, 13.8 ± 11.4 % shoot frequency, and 49.2 % (20.5 – 66.9%) flowering plants, compared to control plots exposed to herbivory. Meadow plots protected from herbivory by fencing grew an additional 106 ± 66 g m-2 of aboveground biomass compared to control plots. Bare ground in fenced plots dropped by 3.5 ± 3.1% areal cover, and survival of two species of sedges and lodgepole pine are significantly higher in fenced plots. Sedge cover is low and bare ground is high in Tuolumne Meadows compared with similar nearby meadows, likely due to the large size and accessibility of the meadow to shepherds in the late 1800s. Native herbivory is limiting sedge survival and maintaining a high proportion of bare ground.

Chapter 3 looks at the effect of the meadow herbivory exclosure on carbon dioxide fluxes into and out of the Tuolumne Meadows ecosystem. Models for gross primary production (GPP) and ecosystem respiration (ER) were built using hourly measurements of environmental variables to fill in gaps between direct field measurements of GPP and ER carbon fluxes. The modeled summer carbon flux shows ER flux approximately double that of GPP, resulting in net ecosystem exchange (NEE) ranging from 469 to 666 g CO2-C m-2 released from the meadow to the atmosphere. NEE was significantly higher in wet plots compared to dry plots. In the summer of 2014 the herbivore fencing treatment reduced NEE efflux to the air in the wet plots by 92.8 ± 58.9 g CO2-C m-2 while the fencing effect on NEE was not significant in the dry plots.

Chapter 4 examines the effects of groundwater pumping on the sustainability of a mountain wetland complex in Yosemite National Park. Daily head pressure and water table declines observed at sampling locations within 100 m of the pumping well were strongly correlated with the timing and duration of pumping. Predictive scenarios developed using a groundwater model showed that even in a dry year like 2004, distinct increases in fen water table elevation can be achieved with reductions in pumping. Site vegetation composition indicated that maintenance of a high water table during summers following low snowpack years had a more significant influence on vegetation composition than depth of water table in wet years or peat thickness.

Chapter 5 describes a field investigation of the effects of soil compaction on wetland plant growth, a field experiment to determine how the addition of wood chips affects soil compaction, and a greenhouse experiment to measure how phenolic compounds from wood chips affect wetland plant growth. Field soil compaction (MPa) was significantly negatively correlated with both wetland plant height and width, resulting in -9.8 and -11.9 cm MPa-1, respectively. Experimentally amending soil reduced compaction by 0.174 MPa per 10%-by-volume addition of wood chips. In the greenhouse, a high concentration of phenolic compounds derived from bark (211 mg/L) significantly reduced wetland plant growth and triggered the production of polyphenol oxidase (PPO). However, phenol concentrations similar to field conditions (0 – 12 mg/L) did not affect plant growth or PPO production.