A. Sampling Site Samples were taken from the University of Michigan Biological Station (UMBS) (45˚30’N, 84˚42’W) in Pellston, MI. Samples were taken from two forest sites; the first forest, the Ameriflux forest, served as an undisturbed site while the FASET (Forest Accelerated Succession Experiment) served as the experimental site because of its disturbance gradient. Stuart-Haëntjens et al. 2015 characterized the level of disturbance in each plot of the FASET forest by recording the fraction of basal area senesced since the time of disturbance. Of the 21 total plots located in the FASET forest, 10 were selectively chosen for further study due to their place on the established disturbance gradient (Stuart-Haëntjens et al., 2015). Figure 2 below shows the spatial distribution of the plots in the FASET forest, and Table 2 below shows the level of disturbance for each plot as well as its dominant cover-type (Stuart-Haëntjens et al., 2015). Table 2: A summary of sampled plots as well as their relative disturbance level and dominant canopy cover-type. Plot (FASET) Fraction of Basal Area Senesced Cover Type A1 0.58 Bigtooth Aspen A2 0.54 Northern Hardwood-Hemlock B1 0.47 Bigtooth Aspen-Trembling Aspen C1 0.64 Bigtooth Aspen D1 0.38 Red Oak-Bigtooth Aspen D2 0.67 Red Oak-Bigtooth Aspen F1 0.13 Bigtooth Aspen-Red Oak F2 0.09 Bigtooth Aspen-Red Oak G2 0.19 Bigtooth Aspen-Red Oak G3 0.53 Bigtooth Aspen-Red Oak Plot (Ameriflux) Fraction of Basal Area Senesced Cover Type A1 0.00 Bigtooth Aspen F2 0.00 Trembling Aspen G4-5 0.00 Bigtooth Aspen Bryant Rd. 0.00 Bigtooth Aspen B. Basic Live Needle Harvesting Method Flat-ended steel forceps were used to collect approximately 50 second-year needles from the south side of each tree, as close to arm height (approximately 1.3m from the ground) as possible. The forceps used were rinsed in HPLC grade hexane before and after each sample was taken by dipping them into a 100ml plastic scre-cap vial. The needles were placed in a plastic Ziploc sampling bag (Thermo Scientific Sample Bags, LDPE, 10.2 x 15.2 cm, Product# 6255-0406). Pine needles are attached by what is called a fascicle, and Eastern white pine has 5 needles attached to each fascicle (Barnes and Wagner, 2011). For sampling purposes, 10 bundles of 5 needles each were removed from the tree and placed in the bag. Field blanks were taken between every 10 tree samples by opening the sample bag and placing the steel forceps into the bag 10 times to mimic real sampling. All samples were stored on dry ice immediately after collection in the field and were moved to a -80˚C freezer until processing to halt emissions because tissue damage has been shown to change VOC release from foliage (Lorento and Schnitzler, 2010). C. Validation of Sampling Method For Live Trees On 7/2/17, one singular tree was sampled at UMBS to evaluate whether or not the location on the tree of needles sampled from understory white pine had an effect on the terpene concentration for that specific tree. From this tree, one whorl 0.75m off of the ground and one whorl at DBH (approximately 1.4m). From each whorl, 2 samples were taken from each branch. The top whorl had four branches while the lower whorl only had three branches, so a total of 14 samples were taken in total. From each branch, one needle sample was taken from the interior of the tree near the trunk while the other sample was taken from the exterior, farthest from the trunk. Samples were removed in rapid succession and processed for terpene analysis on the same day. D. Sampling Description a. Live Needle Plot Sampling - FASET Samples were collected from the FASET site between the times of 9:30 a.m. and 4:15 p.m. on 7/7/17 and 7/8/17. Both sampling days had relatively similar weather: partially cloudy, 74˚F, and wind speeds of 10-15mph. From each “disturbance plot,” listed in Table 2, six understory white pine trees were sampled, totaling 60 trees from the FASET forest. From each plot, 3 of the 6 trees were characterized as being “small” (less than 2.2m in height) while the other 3 trees were characterized as being “medium” (between 2.3 and 4.5m in height). Samples were taken from areas that were relatively representative of the microclimate of each specific plot, however, in some plots few understory white pine were present, so sampling was limited to the few available trees. For each tree, the location, number of whorls (a measure of tree age), approximate height, DBH (diameter at breast-height), the relative light environment, types of vegetation, time of day, and ambient air temperature were all recorded. The relative light environment was an arbitrary analysis of the openness of the canopy above where the tree was sampled. A tree completely exposed to sunlight at all times during the day was given a canopy cover level of 1, a tree that was mostly exposed but received some shad was given a canopy cover level of 2, a tree that was relatively shaded received a canopy cover level of 3, and a tree that received little to no light received a canopy cover level of 4. In an effort to understand the microclimate of each plot, a singular soil moisture reading was recorded in each plot on the days of 7/21/17 and 7/22/17. b. Live Needle Sampling – Ameriflux On 7/11/17, samples were collected from the Ameriflux forest sites between the times of 9:30 a.m. and 1:00 p.m. Weather was similar to previous sampling days; 77˚F and a slight breeze. Trees were sampled in the same manner as in the FASET site with 6 trees sampled per plot, 3 of which were categorized as “small” and 3 categorized as “medium”. The same sampling procedure was used, and samples were also stored on dry ice in the field and then transferred to a -80˚C freezer until processing. Field blanks were collected between every 10 trees, as well as after the last 4 trees were sampled. 24 trees in total were sampled in the Ameriflux forest. c. Leaf Litter Sampling Leaf litter samples were collected from the FASET forest on 7/25/17 and from the Ameriflux forest on 7/31/2017. One sample was collected 5m north of the stake marking the center of each plot that was previously used to sample live trees. All needles were collected from a 13cm x 13cm square located in the lower left-hand corner of a larger 50cm x 50cm square. Within the 50cm x 50cm square, the type of ground cover was recorded, and a singular soil moisture and soil temperature reading was recorded at the center of the square. Soil moisture and the soil temperature were recorded at a depth of 15cm. The top layer of needles and deciduous leaves was carefully excised from the area by hand to expose needles that had fallen the previous fall. This year-old leaf litter was collected with flat-ended steel forceps and placed into plastic sampling bags (Thermo Scientific Sample Bags, LDPE, 10.2 x 15.2 cm, Product# 6255-0406). The next two layers of deciduous leaves were then removed by hand to expose the final layer of remaining needles that had fallen approximately 3 years prior. Finally, a soil sample was collected from the area beneath the excised needles with a flat steel garden trowel for future use in exploring microbial soil characteristics. d. Experimental Sampling Method To further explore the nature of Germacrene-D-4-ol formation, an experimental sampling method was developed to examine the role that sampling induced damage may play in the formation of this species. Four medium sized understory white pine trees were selected on UMBS Property from behind Lakeside Lab, as well as the trail leading from the lab to the UV Field, and sampling occurred on 7/30/2017. From each tree, 2 branches were selected for sampling on the south side of each tree. On “Branch 1” of each tree, one entire lateral branch on the right side of the main branch was dipped into a one liter dewar filled with liquid nitrogen until the needles were clearly frozen (indicated by extreme reduction of boiling by the liquid nitrogen). Once frozen, the sampling method used for live needles collected from FASET and Ameriflux was used. On that same branch, 50 needles were then collected from a different lateral branch on the right side of the tree in the same method without freezing. Needles on a third lateral branch on the right side of the main “Branch 1” were wiped with a Kimwipe that was dipped into HPLC grade hexane (transported into the field in a plastic screw cap vial), to remove the outer coating of the needle. These needles were then removed in the same fashion as all other samples. On “Branch 2” of the same tree, the same sampling methods were used, however they occurred in a different order. The table below illustrates this sampling method. Table 3: This table illustrates how one of the four experimental replicates would be sampled. Branch Number Lateral Branch Number Order of Sampling Removal Method 1 1 1 Freeze needles then remove them from the tree 1 2 2 Remove needles without freezing them (Control) 1 3 3 Remove needle coating and then remove needles 2 1 1 Remove needles without freezing them (control) 2 2 2 Freeze needles then remove them from the tree 2 3 3 Remove needle coating and then remove needles E. Sample Processing From each frozen sample, the fascicles were removed from each bundle of needles, and the needles were placed back into the sampling bag. Approximately 0.2g of randomly selected needles were removed from the sampling bag and cut into 5mm pieces and placed into an extraction vial (Fisherbrand Glass Tooled-Neck Vials with Polyethylene Closures; 21 x 50 mm OD x H; 3 dr. 11.1 mL) with 2.0 mL of HPLC grade hexane with a 50µM tridecane internal standard. Each vial was capped and wrapped in parafilm to prevent sample loss, and samples were incubated in a 23.0˚C water bath for 24 hours. After incubation, the solution was placed in an auto-sample vial (Thermo Scientific 12 x 32mm Amber target DP ID Vials) via glass transfer pipette. Samples were stored in a freezer until GCMS analysis. Analysis was accomplished within 12 hours of being placed into autosampler vials. Repeated injections showed that the hexane extracts did not degrade after 12 hours at room temperature on the autosampler tray. The extraction vials with the needles were placed in a drying oven and dried for 24 hours at 60˚C. Dry mass was then recorded. F. Sample Analysis Instrumentation used for analysis: GC: Thermo Scientific Trace 1310 Mass Spectrometer: Thermo Scientific ISQ LT Single Quadrupole Mass Spectrometer GC Column: Thermo Scientific DB-5 column (TG-5MS) Length: 30m I.D.: 0.25mm Film: 0.25µm Maximum Temp.: 330/350˚C Part Number: 26098-1420 Thermo Scientific DB-5 column (TG-SQC) Length: 30m I.D.: 0.25mm Film: 0.25µm Maximum Temp.: 330/350˚C Part Number: 26070-1300 *Both columns were spliced together with a Restek press-tight union. The 15m column was attached to the mass spectrometer while the 30m column was attached to the GC inlet. a. Standards used for Calibration: Monoterpene Standards: • (1R)-(+)-α-Pinene (98 %, ALDRICH) • Camphene (98.5 %, SUPELCO) • (1S)-(-)-β-Pinene (99 %, ALDRICH) • β-Myrcene (≥95.0 %, Fluka) • α-Phellandrene (≥85.0 %, ALDRICH) • R-(+)-Limonene (≥99.0 %, Fluka) • Terpinolene (≥95.0 %, Fluka) • 3-Carene (≥90.0 %, ALDRICH) Sesquiterpene Standards: • β-Caryophyllene (≥80.0 %, SAFC) • α-Humulene (≥96.0 %, ALDRICH) • Trans-Farnesol (≥96.0 %, ALDRICH) b. Standards used for Retention Time Determination • Cadinene (Unlisted purity (code 504607), Vigon International) • Gamma-Terpinene (Unlisted purity (code 500386), Vigon International) • Germacrene D (natural 40 %, GLCC, CO.) • Angelica seed oil (60 % of β-Phellandrene, SKK, ltd.) c. Internal Standard • Tridecane (99 %, ALDRICH) d. GCMS Method Program GC Inlet Settings: Temperature: 240˚C Operating Mode: Split Split Flow: 11.0 (mL/min.) Purge Flow: 0.500(mL/min.) Gas Saver Control: Off Vacuum Correction: On Flow at Equilibration Time: 1.750 (mL/min.) Autosampler Settings: Draw Speed: Slow Fill Strokes: 3 Air Volume” 1.00µl Sample Depth: Bottom Pre-washes: 2 per solvent (A and B) Sample-washes: 2 Post-washes: 2 per solvent (A and B) Solvent A: Acetone Solvent B: Hexane GC Oven Settings: Retention Time (min) Rate (˚C/min.) Target Value (˚C) Hold Time (min.) 0.000 Run 4.000 0.000 70.0 4.00 19.000 4.000 130.0 0.00 29.000 7.000 200.0 0.00 30.000 Stop Run Run Time: 30 minutes Mass Spectrometer Settings: Scan Start Time: 5.00 min. Mass Range (amu): 20-250 Scan Time: 0.2 sec. Transfer Line Temperature: 280˚C Ion Source Temperature: 320˚C e. Software and Processing Method • GCMS Software: Chromeleon 7.2 • Processing method: Retention Time (min.) Parameter Name Parameter Value Injection Type Channel Initial Consider Void Peak Off All TIC Initial Smoothing Width 0.117 min. All TIC Initial Baseline Noise Range 5.001….5.126 All TIC 5.001 Minimum Area 20000 [Signal]*min All TIC 10.600 Minimum Rider Ratio 10% All TIC 10.600 Detect Shoulder Peaks On All All Channels 10.657 Rider Detection On All All Channels 10.663 Baseline Type Valley to Valley All All Channels 10.830 Detect Shoulder Peaks Off All All Channels f. GCMS Calibration Method The GCMS was calibrated using a series of calibration solutions containing alpha-pinene, camphene, beta-pinene, beta-myrcene, alpha-phellandrene, D-limonene, terpinolene, beta-caryophyllene, alpha-humulene, and trans-farnesol (See section (a) above: Standards used for Calibration). Five terpene solutions of differing concentrations were made, and the concentrations were as follows: 500uM, 250uM, 100uM, 50uM, and 5uM. A 50uM tridecane internal standard was added to each concentration, and solutions were made from serial dilution of the 500uM and 250uM stocks. Each calibration solution was run three times. To create calibration curves, the area in Counts*minutes was recorded for the internal standard peak as well as each terpene component peak in each sample. A calibration curve was formed using the ratio of the peak area/internal standard area, and an orthogonal distance regression was formed for each compound. An orthogonal distance regression is useful for calculating a best fit line based on the assumption that errors exist in both x- and y- values. There are definitely uncertainties associated with calibrating the GCMS using calibration solutions, for example the uncertainty in micro pipetting volumes of standards, the error associated with volumetric flasks and serial dilutions, and the precision of the GCMS itself. The orthogonal distance regression calculates errors perpendicular to the best-fit line. After calibration curves were formed, the equation of the best fit line was then used to determine the concentration present of each component in actual samples. In total, 27 terpene components were recorded, but because a standard was not commercially available for many of them, the relative concentration of un-calibrated components was determined based on calibration curves created for other compounds. Tricyclene was calibrated for using camphene because both are monoterpenes that are solid at room temperature. Beta-phellandrene and gamma-terpinene were calibrated for using the best fit line for alpha-phellandrene because of the similar structure shared by all three compounds. Finally, all sesquiterpenes (not counting beta-caryophyllene, alpha-humulene, and trans-farnesol), were calibrated for by using the average slope and intercept between the best fit lines for alpha-humulene and beta-caryophyllene. Calibration curves for each terpene component are located in Appendix A. Appendix B has two sample chromatograms: one from a 250uM calibration solution, and one from sample N1B2-A, showing the components and their retention times. g. Statistical Analysis The significance level for all tests was 0.05 and all statistical analyses were conducted using SPSS software (IBM).