Precipitation pulse responses and mortality in a five-year drought experiment

Background/Question/Methods

Mortality mechanisms may vary across the continuum between isohydric and anisohydric stomatal regulation, but few studies have examined the relationship between plant response to rain events during drought and processes leading to mortality. The ability to respond quickly to rain events may differentiate between trees that die and survive. To address this question, we analyzed precipitation pulses and sapflow-estimated transpiration responses over five years in relatively anisohydric Juniperus monosperma (juniper) and isohydric Pinus edulis (piñon) in ecosystem-scale drought, water-addition, ambient, and cover-control treatments. We evaluated growing season pulses of at least 3 mm between 2007 and 2011, and calculated the response of transpiration after each event. We anticipated three possible scenarios: 1) a single model can predict the transpiration response across treatments based on parameters that vary with treatment (soil moisture before and after each pulse, pulse size, duration of drought prior to pulse, etc.); 2) different models are required for different treatments; or 3) treatment effects are additive and intensify with treatment duration such that one model may fit all treatments at the beginning, but eventually responses diverge to the point that treatment-specific models are required. 

Results/Conclusions

Our results revealed differences in each species’ ability to respond to precipitation pulses during an on-going drought. For isohydric piñon, the pre- and post-pulse soil water content and pulse size each significantly influenced the magnitude of post-pulse transpiration. Supporting scenario 1, the relationships were similar for all treatments, including the drought plot piñon which were attacked by bark beetle, infected with Ophiostoma fungi, and died after approximately 1 year. Trees that died experienced lower soil moisture and smaller pulses, and therefore reduced transpiration. In anisohydric juniper, almost all of the same drivers significantly influenced post-pulse transpiration, but response varied by treatment. Maximum transpiration was constrained in the drought plot and a year-by-year comparison showed that this was only after the drought treatments were installed. Thus, scenario 2 appears supported by our juniper data. Droughted juniper exhibited progressive branch die-back during the experiment, sometimes leading to complete mortality. This suggests that the shift in juniper pulse response observed in the drought treatment is linked to physiological changes that may result in mortality depending on drought duration. These results provide insights into the dynamics of long-term drought and the field conditions mature trees can endure compared to those causing mortality.