A-Ci response curves: A-Ci response curves were measured between July 13 and July 21, 2022, using a LI-COR 6800 infrared gas analyzer. Four replicates of each species were measured in each watershed. We used the dynamic assimilation response curve method that uses dynamic equations to continuously change CO2 concentrations from 1600 ppm to 40 ppm. Leaves were allowed to stabilize to chamber conditions before starting each curve. Initial chamber conditions were set to: CO2 = 400 ppm, temperature = 30 °C, relative humidity = 50%, light = 2000 umol m-2 s-1. Maximum rates of electron transport (Jmax) and carboxylation capacity (Vcmax) were derived using the plantecophys package in R. Leaf area was calculated using imageJ for accurate parameter estimation.

Rapid light response curves: Rapid light response curves were measured between June 28 and July 5, 2022, using a LI-COR 6400 infrared gas analyzer. Four replicates of each species were measured in each watershed. Light intensities were set to 2000, 1600, 1200, 800, 400, 200, 100, 50, 30, 15, and 0 umol m-2 s-1. CO2 inside the chamber was set to 400 ppm and relative humidity was maintained between 40-50%. Each leaf was allowed 90-200 seconds to stabilize between light intensities. We derived quantum yield (f), light compensation point (LCP), light saturated photosynthetic rate (Asat) using the photosynthesis package in R. Leaf area was calculated using imageJ for accurate parameter estimation.

Pressure-volume curves: Pressure-volume curves were sampled between August 22 and August 31, 2022, using the bench dehydration method. Four replicates of each species were measured in each watershed. Terminal sections of branches were collected before sunrise between 400 and 530 hr, placed in a plastic bag with a moist paper towel, and stored in a refrigerator until measuring. All branches were collected on the same morning as they were measured. To start each curve, a portion of the stem with the youngest developed leaves were cut underwater with a razor blade and weighed. Leaf water potential was measured using a Scholander Pressure Chamber. The leaf was weighed again immediately after the water potential measurement, allowed to dry down on the bench top for 2 minutes, placed in a plastic bag to equilibrate for 20 minutes and then water potential was measured again. This process was repeated with dry down periods of 2 minutes to 2 hours until a pressure of 3.0 to 3.5 MPa was reached. Each leaf was dried at 60 °C for 48 hours and weighed. We derived turgor loss point for each sample using the pressure volume analysis spreadsheet tool (PVAST).

Leaf functional traits: Specific leaf area (SLA) and leaf dry matter content (LDMC) were measured in late July 2022. Five leaves from five replicates of each species were measured in each watershed. The petiole and rachis of each leaf were included. Leaves were collected, placed in a moist plastic bag, and stored in a cooler until processing. We measured fresh area using imageJ. Leaves were then rehydrated overnight and measured for water saturated mass. Leaves were dried for 48 h at 60 C and weighed them for leaf dried mass. SLA and LDMC were calculated for each individual leaf and then averaged for each individual shrub. Leaves for each replicate shrub were combined, ground, and measured for foliar C:N and d¹³C.

Wood density: Wood density was measured in August 2022 using the water displacement method. Five replicates of each species were measured in each watershed. A 10 cm long section of stem was cut 20 cm above the ground and measured for fresh volume. Stem sections were dried at 60 °C for 72 hours and weighed to calculate stem density (mg cm-3).