Abstract: Predicting species’ responses to drought is important in the context of climate change. Common methods to quantify responses are to perform greenhouse or lab experiments to determine critical thresholds like percent loss of conductivity (PLC) or turgor loss points (TLP). These measurements are often performed on small potted plants that undergo a rapid dry down of soils that may not represent mature plants in natural conditions that may experience a more gradual, and chronic, decrease in water availability. In addition, results can vary within species depending on plant provenance as well as among various methods to measure PLC. We used Tamarix, a non-native woody plant, that is widespread throughout the arid western United States, including the Cibola National Wildlife Refuge on the lower Colorado River where the source populations for this study were located. Several studies have examined the hydraulic function of this taxon to understand their invasive success, often as compared to other co-occurring native species. Previous work with the same source populations found differences in xylem anatomy and water use strategies in a greenhouse setting, and to further understand these population differences, we investigated differences in hydraulic conductivity (K), PLC, vessel anatomy, and response to an experimentally induced drought treatment which resulted in plant mortality.
We measured vessel length and diameter, and compared two methods to quantify PLC (microCT and the "hydraulic” method), on the same stem segments from plants that were dried down to increasing levels of dehydration over a few days. We created a mortality observation garden using plants from the same populations, where the plants received no supplemental irrigation and water potentials were monitored until dieback of the canopy was observed. Unexpectedly, we did not find differences in PLC or vessel anatomy between the two populations of Tamarix. We found that the microCT method consistently overestimated K and underestimated PLC when compared to the hydraulic method. In comparing the results from the mortality observation garden, we found good agreement for crown dieback and the hydraulic method. Our findings indicate that the hydraulic method was a good predictor of the water potentials where crown dieback was observed, i.e. how plants performed in the field, as well as other measurements of Tamarix in greenhouse and in situ studies. Linking different methods to plant field-based performance metrics is essential to validate different methods in plant hydraulics.
