Truncation of stream residence time: how the use of stable isotopes has skewed our concept of streamwater age and origin

Abstract Although early studies of streamwater residence time included the use of stable isotopes (deuterium, oxygen‐18) and tritium, work in the last decades has largely relied on stable isotopes (or chloride) alone for residence time determination, and derived scaling relations at the headwater and mesoscale watershed scale. Here, we review critically this trend and point out a significant issue in our field: truncation of stream residence time distributions because of only using stable isotopes. When tritium is used, the age distributions generally have long tails showing that groundwater contributes strongly to many streams, and consequently that the streams access considerably larger volumes of water in their catchments than would be expected from stable isotope data use alone. This shows contaminants can have long retention times in catchments, and has implications for process conceptualization and scale issues of streamflow generation. We review current and past studies of tritium use in watersheds and show how groundwater contributions reflect bedrock geology (using New Zealand as an example). We then discuss implications for watershed hydrology and offer a possible roadmap for future work that includes tritium in a dual isotope framework. Copyright © 2010 John Wiley & Sons, Ltd.