Local adaptation in leaf water relations of a dominant aridland phreatophyte: Populus fremontii

Background/Question/Methods

Plant traits and water status can vary within a species, and seasonally within a plant. Leaf turgor loss point (ptlp) is recognized as one of the most reliable traits for evaluating drought tolerance. Fremont cottonwood (Populus fremontii) is a dominant riparian tree species in the southwestern United States that relies on shallow groundwater for growth and survival. Although P. fremontii is generally drought intolerant, this species occurs across a vast climate gradient including where temperatures approach 50 °C. Previous results indicate that exposure to intense heat has resulted in the selection of warm-adapted P. fremontii for high midday canopy transpiration rates to enhance leaf cooling relative to cool-adapted genotypes. Given these patterns, we tested the hypotheses: 1) warm-adapted P. fremontii displays lower mean ptlp than cool-adapted genotypes, and 2) ptlp is correlated with higher midday stomatal conductance (gs), lower midday leaf water potentials (Ymd) and higher leaf dry matter content (LDMC). The hypotheses were tested using genotypes planted in a common garden established in 2014 near Yuma, Arizona. Ten genotypes sourced from eight locations spanning an elevation gradient from 70 m to 1230 m were selected for the study. We measured leaf traits in spring and mid-summer in 2021.

Results/Conclusions

Mean ptlp varied from -2.45 MPa to -2.66 MPa in spring, and from -2.51 to -2.70 MPa in summer. However, contrary to hypothesis 1, there was no relation between ptlp and elevation of source population in either spring or summer. Across all populations, ptlp was marginally lower in summer than in spring (P=0.07). Likewise, from May to August, gs increased 85% (P< 0.0001) and LDMC increased slightly by 6% (P< 0.0001). Unlike ptlp, seasonal changes in gs was strongly negatively correlated with source population as a result of warm-adapted populations increasing gs by over 150% from spring to the much warmer summer period. Mean ptlp was moderately negatively correlated with LDMC in May (R2=0.18, P< 0.0001), supporting our second hypothesis, but no relationship was detected between ptlp and other measured traits, including gs and Ymd. LDMC was negatively correlated with gs in spring (R2=0.86, P< 0.0001), but not by mid-summer. Results indicate that warm-adapted plants prioritize leaf cooling via increased mid-summer stomatal conductance, but leaf water use patterns did not appear to correspond with leaf hydraulic function in P. fremontii genotypes sourced across a broad temperature gradient.