Leaf morphological and physiological traits
To determine leaf morphological traits, we studied twenty randomly selected individuals at each elevational site. Leaf area and lengths of major and minor axes were determined from the scanned leaf images (GT‐S630, EPSON, Japan) with image analysis software (Image J, National Institutes of Health, USA). “Characteristic leaf dimension”, which is length of a leaf in the direction of airflow (d ), was calculated as the average length of the major and minor axes multiplied by 0.72 (Campbell & Norman, 2010). We determined lamina mass and trichome mass by a shaving technique (Tsujii et al., 2016). We defined leaf mass per area (LMA) associated with lamina and trichome as laminaLMA and trichomeLMA, respectively, according to Tsujii et al. (2016) (i.e., LMA = laminaLMA + trichomeLMA). On one‐half side of a mature leaf (separated by the midrib), we shaved leaf trichomes using a rubber thimble and punched out two 10‐mm‐diameter disks from both sides of the leaf. These leaf disks were dried at 70 °C for three days and used for weighing with a digital scale. TrichomeLMA was calculated from the differences in dry weight between the disks with and without trichomes. We determined lamina and trichome thicknesses by a light microscope (SZX-ILLB2-100, Olympus, Japan). We excised 10 × 3 mm segments from leaves and fixed them with 5% glutaraldehyde. Hand cut sections (<10 µm thickness) were used to measure trichome thickness (δt ) with a light microscope. Porosity and tortuosity were estimated from the trichome thickness by using the correlation between the porosity and the trichome thickness in M. polymorpha (Amada et al. 2017; Figure S1).
To determine the photosynthetic parameters [the maximum rate of RuBP (Ribulose-1,5-bisphosphate) carboxylation (Vcmax ) and the maximum rate of electron transport required to RuBP regeneration (Jmax )], we conducted gas-exchange measurements on ten randomly selected individuals at each elevational site with a portable photosynthesis system (LI‐6400XT, LI‐COR, USA) with the integrated fluorescence chamber head (chamber area=2 cm2, 6400‐40, LI‐COR, USA). The measurement was done on intact leaves between 6 and 12 am in July and August 2017. We determined ACi curve by stepwise decrease of [CO2] in the chamber from 400 to nearly 0 µmol mol-1 and followed by step increase from 0 to 1800 µmol mol–1 under the saturating photon flux density for photosynthesis (1000 µmol m–2s–1). Vcmax andJmax were determined by fitting Eqs. S27-30 (Appendix S3; Farquhar et al. 1980) using least square method.Rd was estimated as 0.015Vcmax according to Collatz, Ball, Grivet, & Berry (1991). The in vivo Rubisco kinetics parameters (Ko , Kc ,Γ* ) were taken from Bernacchi, Singsaas, Pimentel, Portis, & Long (2001) (Table 2). The values ofVcmax and Jmax at the 100, 1280, and 2400 m sites partially included the data from Amada et al. (2017).
To determine the temperature dependence of Vcmaxand Jmax and the stomatal parameters [the minimum stomatal conductance when A = 0 (g0 ) and the coefficient expressing sensitivity of stomatal conductance to photosynthetic rate and humidity deficit (a ); Appendix S3; Leuning 1995] in M. polymorpha , we conducted the gas-exchange measurements at four different leaf temperatures (15, 20, 25, 30 °C) on randomly selected ten individuals at the 2000 m site. We measured Vcmax andJmax at each leaf temperature with the same procedure described above. Then, the activation energies forVcmax , Jmax , andRd (HVc ,ΔSv , Hj ,ΔSj , HRd ) and the entropy factors for Vcmax and Jmax(ΔSv , ΔSj ) were fitted by using the modified Arrhenius models (see Eqs. S37–39 in Appendix S3; Medlyn et al. 2002). Using the sets of photosynthetic rates (A ) and humidity deficits (D ) for each individual in these measurements, the stomatal coefficients (g0 ,a ) were fitted by using the stomatal-conductance model from Leuning (1995) (Eq. S22 in Appendix S3).