Experimental Methods: This experiment is located in watersheds 4B and 20C at KPBS. Both watersheds were chosen for their burn frequencies and abundant Cornus drummondii. Starting in May 2015, 20 discrete shrub islands were chosen in the lowlands that spanned the length of each watershed. In each watershed, 10 shrub islands were randomly assigned the browsing treatment and the remaining shrub islands act as an unmanipulated control to the browsing treatment. Our browsing treatment is a simulated browsing treatment, where 50% of new aboveground meristematic growth is removed using clippers or hands. Browsing occurs once a month during the growing season (May - September), all biomass removed is moved away from the shrub islands. In watershed 4B a prescribed fire occurred in April 2017.

Data Collection Methods: Species Composition: This protocol is the same that is used in the PVC02 long-term data set but adapted for our study. In the center of each selected shrub island in watersheds 4B and 20C a surveyor's pin with a 1.78 m long chain is used to get canopy cover of all vascular plants within a 10 m² circular area. Canopy cover is estimated using a modified Daubenmire cover scale (Bailey and Poulton, 1968. Ecology 49:1 - 13), the cover categories are the same as described in PVC02.

Shrub Area: The area was calculated by taking two diameter measurements perpendicular to each other at the center of the shrub islands and then put into an ellipsoidal area equation.

Stem/Ramet Density: Every stem/ramet was counted within the shrub island area before the browsing treatment started and then at the end of the growing season in October. Additionally, stems/ramets were counted a month after the prescribed fire occurred in 4B.

Gas Exchange: During the growing season (May - September) gas exchange measurements were taken once a month around solar noon at each shrub island using a LICOR - 6400. Within each shrub island 2 measurements were recorded, the first at the edge of the shrub island and the second at the center of the shrub island. All measurements were taken on new fully expanded leaves.

Leaf Area Index (LAI): During the growing season (May - September) LAI measurements were taken once a month around solar noon at each shrub island using a LICOR-2000. Within each shrub island 2 measurements were recorded, the first at the edge of the shrub island and the second at the center of the shrub island. Additionally, 10 measurements in each watershed were taken in the space between shrub islands as a grassland control for LAI.

¹³C Leaf Stable Isotope: Once a month during the growing season several new fully expanded leaves were clipped from each shrub island in each watershed. These leaves were then dried for 48 hours at 60^oC and then powdered using a Wig-L-Bug^rgrinder. Once ground the leaf material was processed using an elemental analyzer in sequence with an isotope-ratio mass-spectrometer. See Nippert et al. 2013 (https://doi.org/10.1371/journal.pone.0081630) for an identical protocol. 

Diurnal Leaf Temperature: Leaf temperatures were taken every 6 hours during a 24 hour period (6AM, 12PM, 6PM, 12AM) in all of the shrub islands. These measurements were made at least once per month during the growing season (May - September) using an infrared radiometer (Apogee MI-220). Measurements were taken at 3 locations in each shrub island: southern side, center and northern side. At each location 2 measurements were made if possible, first in the upper canopy and the second in the understory. All measurements were made on fully expanded leaves.

Diurnal Soil Moisture: Soil moisture is taken every 6 hours during a 24-hour period (6AM, 12PM, 6PM, 12AM) in all of the shrub islands and at 10 locations between each shrub island in each watershed. Measurements were taken at 3 locations within each shrub island: southern side, center and northern side. All measurements were made at least once a month during the growing season (May - September) using a handheld Stevens Hydraprobe soil moisture sensor.

Root Non-Structural Carbohydrates: Several grams of root tissue were harvested at each shrub island each year prior to the shrubs leafing out and after leaf senescence. All root tissues were cleaned in DI water to remove all dirt impurities and then dried for 72 hours at 60^oC before being ground up using a Wig-L-Bug® grinder. Then all of the samples had the non-structural carbohydrates extracted and were analyzed colormetrically using a spectrophotometer (Hendrix, 1993.Crops Science 6:1306-1311).   

Fire Temperature: In watershed 4B prior to the prescribed fire 100 ceramic tiles with welder's temperature paint was placed in shrub islands and in the space between the shrubs. Each shrub island had 5 ceramic tiles placed in each cardinal direction (North, East, South, West) as well as 1 ceramic tile in the center of the shrub island. The remaining 20 tiles were placed randomly in the grassland spaces between the shrub islands. Each ceramic tile had 21 paint samples that ranged from 175 ^oF to 1150 ^oF in 50 degree increments.