NSC Laboratory Preparation
We initially freeze-dried samples for 24-hours (FreeZone 2.5; Labconco, Kansas City, MO, and Hybrid Vacuum Pump, Vacuubrand, Wertheim, Germany), then ground them to a fine powder (mesh 10, Thomas Scientific Wiley Mill, Swedesboro, NJ, USA; SPEX SamplePrep 1600; MiniG, Metuchen, NJ) and stored them in sealed glass vials. Sugar and starch extraction protocols were adapted from Chow & Landhäusser (2004).
We extracted sugar from 20 mg of dried, ground tissue using 80% hot ethanol, followed by a colorimetric assay with phenol and sulfuric acid, and read using a spectrophotometer at 490nm (Thermo Fisher Scientific GENESYS 10S UV-Vis, Waltham, MA). We calculated sugar concentrations of mg sugar per g of dry wood using a 1:1:1 glucose-fructose-galactose standard curve (Sigma Chemicals, St. Louis, MO). We extracted starch using the tissue remaining after sugar extraction. We solubilized tissue in NAOH, then incubated it for 24-hours with alpha-amylase and amyloglucosidase digestive enzymes, which digested starch into glucose. We then assayed the solutions using a PGO-color reagent solution (Sigma chemicals) and read them on the spectrophotometer at 525nm. Starch concentrations of mg glucose-starch-equivalent per g dry wood were calculated based on a glucose standard curve (Sigma Chemicals).
For all lab analyses, we included at least two internal laboratory standards (Quercus rubra stemwood from Harvard Forest, MA; 42.01 ± 5.13 mg•g-1 Sugar, 30.17 ± 4.23 mg•g-1 starch). This acid methodology extracts all fructose, glucose, sucrose, and starch, as well as other oligosaccharides and other glucans (Landhausser et al., 2018). We then report these metrics as sugar and starch concentrations (mg•g-1) in the supplement, as well as the total nonstructural carbohydrates (TNC) concentration (sugar + starch) and theproportion of starch to total (starch/ (sugar + starch)) in the main text. All statistical analyses were performed in Rv.3.5.1(R. C. D. Team, 2018).