Low-flow driven temperature variation in headwater streams: insights from spatial and long‑term temporal data in the Sierra Nevada

Abstract:

Midpoint snowmelt runoff in the Sierra Nevada is predicted to advance up to two months by the end of the century as a result of climate change. However, it is uncertain how this advance could affect other key physical variables of stream ecosystems. Headwater streams in the Sierra are particularly vulnerable to low flows because they experience large proportional decreases in flow volume and heat capacity, and are at high risk of becoming fragmented. Water temperature could increase rapidly if low flows coincide with peak air temperature; however, stream characteristics such as channel morphology, groundwater influence, and canopy cover could moderate this relationship. Here we studied variation in water temperature across stream networks and over decadal spans to understand the drivers of temperature variation and the scale-dependencies of such variation. The study took place in four headwater watersheds located in the Kings River Experimental Watershed (KREW), in California. We used spectral analyses on long-term water temperature data to identify periodicity and partition seasonal from interannual variability (i.e., anomalies). Annual maximum daily temperature occurred in July or August, and varied greatly by site when comparing the driest to the wettest year (2015 vs. 2017). Nine sites experienced higher annual maximum daily temperature during the lowest flow year, with a maximum difference of 6.67 °C. However, the one meadow site responded in the opposite direction (2.88 °C colder). These results support that interannual variability is high, and the relationship between flow and temperature anomalies (e.g., maximum water temperature) is complex and varies spatially within and across watersheds. We propose that canopy cover and discharge explain the most interannual variation between sites. We also contend that small headwaters dominated by meadows may experience a higher temperature range over small spatial scales during low flow, unless groundwater dominates during this time. We used sixty water temperature sensors distributed throughout the Bull Creek stream network to test this hypothesis. Identifying headwater habitats that will be vulnerable or resilient to climate change is critical as snowmelt advances, air temperature rises, and human water abstraction increases intermittency globally.