As droughts become more frequent and more severe under anthropogenic climate change, water stress due to diminished subsurface supplies may threaten the health and function of semi-arid riparian woodlands, which are assumed to be largely groundwater dependent. To better support the management of riparian woodlands under changing climatic conditions, it is essential to understand the sensitivity of riparian woodlands to depth to groundwater (DTG) across space and time. In this study, we examined six stands of riparian woodland along 28 km of the Santa Clara River in southern California. Combining remote sensing data of fractional land cover, based on spectral mixture analysis, with historical groundwater data, we assessed changes in riparian woodland health in response to DTG during the unprecedented 2012–2019 California drought. We observed a coherent ‘brown wave’ of tree mortality, characterized by decreases in healthy vegetation cover and increases in dead/woody vegetation cover, which progressed downstream through the Santa Clara River corridor between 2012 and 2016. We also found consistent, significant relationships between DTG and healthy vegetation cover, and separately between DTG and dead/woody vegetation cover, indicating that woodland health deteriorated in a predictable fashion as the water table declined at different sites and different times. Based on these findings, we conclude that the brown wave of vegetation dieback was likely caused by local changes in DTG associated with the propagation of precipitation deficits into a depleted shallow alluvial aquifer. These factors suggest that semi-arid riparian woodlands are strongly dependent on shallow groundwater availability, which is in turn sensitive to climate forcing.
Dryland riparian woodlands are considered to be locally buffered from droughts by shallow and stable groundwater levels. However, climate change is causing more frequent and severe drought events, accompanied by warmer temperatures, collectively threatening the persistence of these groundwater dependent ecosystems through a combination of increasing evaporative demand and decreasing groundwater supply. We conducted a dendro‐isotopic analysis of radial growth and seasonal (semi‐annual) carbon isotope discrimination (Δ13C) to investigate the response of riparian cottonwood stands to the unprecedented California‐wide drought from 2012 to 2019, along the largest remaining free‐flowing river in Southern California. Our goals were to identify principal drivers and indicators of drought stress for dryland riparian woodlands, determine their thresholds of tolerance to hydroclimatic stressors, and ultimately assess their vulnerability to climate change. Riparian trees were highly responsive to drought conditions along the river, exhibiting suppressed growth and strong stomatal closure (inferred from reduced Δ13C) during peak drought years. However, patterns of radial growth and Δ13C were quite variable among sites that differed in climatic conditions and rate of groundwater decline. We show that the rate of groundwater decline, as opposed to climate factors, was the primary driver of site differences in drought stress, and trees showed greater sensitivity to temperature at sites subjected to faster groundwater decline. Across sites, higher correlation between radial growth and Δ13C for individual trees, and higher inter‐correlation of Δ13C among trees were indicative of greater drought stress. Trees showed a threshold of tolerance to groundwater decline at 0.5 m year−1 beyond which drought stress became increasingly evident and severe. For sites that exceeded this threshold, peak physiological stress occurred when total groundwater recession exceeded ~3 m. These findings indicate that drought‐induced groundwater decline associated with more extreme droughts is a primary threat to dryland riparian woodlands and increases their susceptibility to projected warmer temperatures.