Determining depth of water uptake in piñon-juniper woodlands using stable isotopes in the Sevietta LTER drought experiment

Background/Question/Methods

Forest mortality is a growing concern globally.  Understanding the physiological mechanism behind tree mortality can help us predict which species will be most vulnerable and where die-offs will occur.  In 2000-2003, the piñon-juniper woodlands (Pinus edulis-Juniperus monosperma) in Southwestern USA experienced a severe drought that precipitated pine mortality rates >90% in some areas, while junipers suffered little mortality.  One hypothesis of the differential success rates between the two species is that pine roots are unable to access the deeper water sources that junipers maintain access to during drought.  In order to test this hypothesis we used water isotope analysis of xylem water and soil water as a natural tracer that can help us constrain where in the soil profile each tree acquired its water.  Over five years, trees were sampled from ambient and experimental drought plots in the Sevilleta study site in southern New Mexico.  This method limits destructive sampling (like root excavation) and provides an integrated signal of the entire root structure.

Results/Conclusions

Juniper trees acquired water from deep sources (>60cm) at a much higher frequency than pine.  For 37% of the dates measured, the junipers range of water uptake went down to 80cm depth, in contrast to 4% of the measured dates for pine.  This is consistent with empirical observations of lower transpiration responses to precipitation pulse events.  Furthermore, pine and juniper water xylem was more enriched than the highest soil water measured 25% and 7% of the time, respectively, consistent with model simulations of greater hydraulic failure in pine.  Combined, these results indicate hydraulic isolation of pine trees from the soil water.  We also saw treatment effects between the ambient and drought plot with Juniper acquiring deeper water sources in the ambient plot 66% of the time and 54% for pine.  Additionally, this study helped us to gain further insight on how to sample in continuing drought mortality studies.