Opinion

The two water worlds hypothesis: ecohydrological separation of water between streams and trees? Jeffrey J. McDonnell1,2,3∗ Recent work in ecohydrology has shown that in some forested watershed systems, streams and trees appear to return different pools of water to the hydrosphere. Thus far, evidence for this has come exclusively from wet Mediterranean climates. This short opinion article outlines the hypothesis and a research agenda for future work. The most pressing issue is the need to gather more data points whereby dual isotope-based studies in forested catchments compare samples of plant water and tightly bound soil water as well as mobile waters (soil, groundwater, and streamflow) in the catchment. New work is needed to test the hypothesis across different climates and vegetation regimes, especially places that contrast with the Mediterranean climates and forest types where twowater worlds have been found. These include, but are not restricted to humid areas where plant water use and precipitation input are in phase, wetter zones where seasonality of precipitation is low, and drier zones where water stress is higher. Of equal importance to these basic research issues are the practical issues surrounding the sampling methods of plant and soil waters. Studies are needed to compare extraction techniques for low and high mobility soil waters and to understand the effect of sampling protocol on water isotope composition. Once these issues are resolved, high frequency sampling of soil and xylem waters will be especially instructive in development of mechanistic models of ecohydrological interaction—and an explanation for the hypothesis that is still wanting. © 2014 Wiley Periodicals, Inc. How to cite this article