Species-specific leaf thermoregulation provenance effects
In our study, we observed provenance effects in at least one thermal trait (i.e., leaf width, leaf angle, or g s) in three of the four species, and provenance effects in measured Tleaf in two species. Assuming the expression of leaf thermal traits under common garden conditions would reflect underlying genetic adaptation associated with provenance, we expected higher Tleaf for cool-adapted, and lower Tleaffor warm-adapted provenances, which would be consistent with selective pressure to limit plant exposure to extreme leaf temperatures (Michaletz et al., 2016). However, patterns of provenance-differentiation in thermal traits and Tleaf followed expected trends (H1) only in one species, C. australe , with the opposite pattern observed for T. microcarpa , and no variation associated with provenance for the other two species. While this contrasts with other studies that have found warm-adapted plants to have a lower Tleaf compared to cool-adapted plants (Blasini et al., 2022; Kitudom et al., 2022; Kullberg et al., 2023), species-specific patterns in intra-specific variation of thermal traits have been reported previously (Kullberg et al., 2023; Manishimwe et al., 2022). This suggests that different species may have different selection pressures, genetic history, or trait trade-offs resulting in contrasting patterns of provenance-associated variation in leaf thermoregulation. It is also interesting to note that the species where we did observe the expected direction of provenance-differentiation in leaf thermoregulation, C. australe , was also the species that had the coolest realized thermal niche (S 1). As a result, the mean annual temperature of the lowland provenance collection site, and common garden planting location were at the warmer limit of the species distribution. This is similar to the provenance-trial set up and findings of Blasini et al (2022), and partly supports the theory that selection pressures leading to local adaptation are stronger at the climate extremes of a species distribution (Rehm et al., 2015; Zimmermann et al., 2009). However, further work assessing variation in thermoregulation in multiple provenances, across sites, and in a broader range of species would be needed to establish if this could explain the contrasting patterns observed across species tested here.