Non-native plants may rely on different patterns of water use to succeed and where this occurs, large-scale shifts from native to non-native plant dominance may be changing hydrological cycles.  Yet, the difficulties of belowground research have limited measurements of the timing, location, and extent of water use. Here we use a depth-controlled tracer technique to quantify water use by dominant native and non-native forbs and grasses in a shrub-steppe ecosystem in Washington State, USA. To do this, deuterium oxide (D₂O), was injected to five different depths at two times (10, 20, 30, 50 and 90 cm and 5, 40, 60, 80, and 120 cm in May and June 2009, respectively) in adjacent native and non-native dominated communities.

Results/Conclusions

Both native and non-native plants relied mostly on shallow soil water, accessing over 70% of their water from depths above 50 cm.  In May, native species absorbed 43%, 15%, 15%, 21 and 3% of the tracer at 10, 20, 30, 50, and 90 cm respectively while non-native plants absorbed 60%, 13%, 14%, 11% and 0.7% of tracer from the same depths. Leaf area index was higher in the native communities (0.6) than in non-native communities (0.2) but mean stomatal conductance was higher for non-native species (128 mmol/m²s) than for native species (116 mmol/m²s) suggesting that transpiration was greater in the native community. Our measurements confirm previous natural abundance data that showed non-native plants using a combination of shallow soil water early in May and deeper soil water in June. Soil sampling a week after tracer injection confirms that the vertical movement of the tracer did not exceed 10 cm in either direction. The tracer addition method shows promise for overcoming the lack of resolution associated with natural abundance isotopes and other enrichment approaches and for providing detailed measurements of plant water use space and impacts of land-use changes on regional hydrology and climate.