Appendix 1. Detailed Methods

Appendix 1. Detailed Methods

Aboveground stand characteristics

To quantify carbon (C) and nitrogen (N) pools for individual tree components, we used component-specific equations developed by Alexander and others(2012) and multiplied the component biomass by the C and N concentration (foliar C and N concentrations reported in this study and wood, bark, and twig values obtained from H. Alexander, unpublished). No component-specific allometry data was available for shrub foliage, so we estimated foliage to be: total biomass – (branch + stem). For a portion of the smallest shrubs, this calculation yielded a negative number, possibly due to our shrubs being smaller than those used to develop the equations. For these shrubs, we calculated component-specific biomass by multiplying the total biomass by an estimated proportional contribution of foliage and stems (Mack and others 2008). Shrub C and N concentration data was obtained from H. Alexander (unpublished). Standing dead tree and shrub biomass was calculated the same way as live biomass, except that the wood biomass values were multiplied by 0.5 to account for the observation that many dead trees had snapped, losing the crown and upper stemwood, which we did not quantify in this study. We also assumed that wood C and N concentration did not differ between live and dead wood and applied the same values to both.

Tree cross-sections used for aging the stands and estimating aboveground tree net primary productivity (ANPPtree) were collected at diameter at breast height (DBH, 1.4 m) from a total of 15 trees per species. In the lab, cross-sections were dried and sanded to obtain a smooth, clear surface, then scanned. Ring number and width were determined using WinDendro software (Regent Instruments Inc., Quebec, Canada). To calculate ANPPtree, the mean width of the preceding 5 years’ growth was used with the wood allometric equations to estimate secondary growth.

To estimate specific leaf area, a subsample of approximately 100 fresh spruce needles of the current year’s growth and 10 fresh birch leaves werescanned using WinRhizo software (Regent Instruments Inc., Quebec, Canada). For the spruce needles, projected leaf area estimates were multiplied by 1.55 to account for their rhombus shape (Bond-Lamberty and others 2003). For nutrient analysis, foliage was dried at 60oC, then ground using a wiley mill (Thomas Scientific model 3383-L10, Swedesboro, NJ). A portion of this material was analyzed for percent C and N using a Costech Analytical ECS 4010 Elemental Analyzer (Valencia, CA) at the University of Florida and an additional subsample was shipped to the Louisiana State University AgCenter for analysis of total phosphorus (P), calcium, magnesium and potassium. Briefly, 5 mL of concentrated nitric acid was added to 0.5 g of ground sample. After 50 minutes, 3 mL of hydrogen peroxide was added and the sample was left to digest for 2.75 hours on a heat block. Samples were then cooled and diluted, then run on a Spectro ARCOS iCAP inductively coupled plasma spectrometer (Germany).

Downed woody debris was quantified using the line intercept method (Brown 1974) along three, 10 m transects placed at random locations across each plot. Fine woody debris was categorized into 5 size classes and the number of intercepts of each of class were recorded and converted to wood mass using multiplier values reported in Nalderand others(1997). The diameter of coarse woody debris (≥ 7 cm) was also recorded and converted to an area basis as outlined in Ter-Mikaelianand others(2008). Biomass estimates were converted to C and N pools using the same concentration information used to estimate standing tree wood biomass.

Soil Characterization

To estimate moisture content, a weighed subsample of each composite SOL sample was dried at 60oC and each mineral sample at 110oC, then re-weighed. Subsamples of both horizon types were dried at 60oC and ground prior to analysis of C and N concentration using the same approach detailed previously for the foliage and litter analysis. Exchangeable base cations were measured by mixing 50 mL of 1 M ammonium chloride (Robertson and others 1999)with 5 g of field moist organic horizon sample and 10 g of mineral soil. Samples were shaken for 1 h on a shaker table, then filtered through a GF/A filter via vacuum filtration and frozen until analysis at the Louisiana State University AgCenter (Baton Rouge, LA) on a Spectro CIROS inductively coupled plasma spectrometer (Germany).Mehlich P concentration was determined using a double acid (0.05 M hydrochloric acid (HCl) + 0.0125 M sulfuric acid) extraction procedure, with 20 mL of acid solution added to 5 g of air dried soil and shaken for 5 minutes (Kuo 1996). Samples were then filtered through Whatman No. 5 filter paper and P concentration was measured colorimetrically using a BioTek Instruments PowerWave XS microplatereder (Winooski, VT). Soil pH was measured using a Thermo Scientific Orion 2 Star pH meter. For organic soils, 5 g of air-dried material were mixed with 50 mL of deonized water in a cup and for mineral soils 20 mL of water were added. Each sample was well mixed, allowed to equilibrate for 30 minutes, then pH was recorded once the value stabilized.

To calculate potential N mineralization and nitrification, ammonium (NH4+) and nitrate (NO3-) were extracted from a subsample of the initial composite soils, as well as for the 30-day and 90-day incubations detailed in the Soil C and N fluxes section of the main text and below. Briefly, 5 g of moist organic soiland 10 g of mineral soil were mixed with 50 mL of 2 M potassium chloride for 1 hr(Robertson and others 1999). Samples were then filtered through a GF/A filter via vacuum filtration and immediately frozen until analysis on an Astoria-Pacific International Autoanalyzer (Clackamas, OR) at the University of Florida.

Near total element digestion was performed on all mineral soils (0-10 cm plot composite samples and all deeper,≤ 1 m mineral soil increments) by ALS Minerals (Reno, NV) using a four acid near total digestion method. First, 0.25 g of ground sample was digested with HCl, perchloric, nitric, and hydrofluoric acids. The residue was brought to volume with dilute HCl, then analyzed on an inductively coupled plasma atomic emission spectrometer.

To radiocarbon date the soils, green moss was first removed from the SOL surface of black spruce SOLs using scissors. The core was then cut in half vertically. One half of the core was carefully cut into 5 cm depth increments and each segment was analyzed for moisture content, bulk density, and C and N using the same methods as described for SOL processing previously. The second half was carefully divided into 1 cm depth increments. A small portion of these segments were then smeared on a piece of paper to qualitatively determine the presence/absence of charcoal within the increment (Jones and others 2013). When the charcoal layer was found, the increment containing the charcoal and at least 2 increments on both sides of this increment were chosen for further processing and radiocarbon analysis. Moss macrofossils were carefully removed with a dissecting scope from the chosen increments in the black spruce SOLs and a bulk soil sample (excluding roots) was also obtained. The birch SOL did not contain any charcoal or mosses, so only a root-free bulk soil sample was analyzed. Cellulose was extracted from the macrofossil samples (Gaudinski and others 2005) and all macrofossil and bulk soil samples were converted to graphite at 650oC with an iron catalyst in a hydrogen atmosphere (Vogel and others 1987). Graphite samples were then sent to the UC Irvine W.M. Keck Carbon Cycle Accelerator Mass Spectrometry Laboratory for analysis.Values obtained for the bulk and macrofossil samples from a given depth increment in spruce stands differed little and here we only report the bulk values. Also, we were unable to keep the black spruce SOL core from block C intact and therefore it was excluded from the analysis.

Ion exchange resins were rinsed with deionized water after returning to the lab and refrigerated until analysis.In preparation for extraction, resin bags were rinsed for 20 seconds with nano-pure water to remove all remaining soil particles. Each bag was placed in atube with 30 mL of a mixture of 0.1 MHCl and 2Msodium chloride and shaken for 1 hr on a shaker table. Extractant was then drip-filtered through a Whatman GF/A filter and frozen until further analysis. Phosphate was measured colorimetrically within 1 week of extraction using a spectrophotometer microplate reader (PowerWave XS Microplate Reader, Bio-Tek Instruments, Inc., Winooski, VT) using the ascorbic acid molybdenum-blue method (Murphy and Riley 1962). The remaining sample was frozen until analysis of NH4+and NO3-using a segmented flow autoanalyzer (Astoria-Pacific, Inc., Clackamas, OR) at the University of Florida.

Soil C and N fluxes

To estimate soil carbon dioxide (CO2)respiration(90-day) and potential net N mineralization and nitrification (30- and 90-day), an approximately 3 cm x 3 cm intact piece of each SOL horizon was removed from each intact core so as to include the entire vertical length of the horizon. Subsamples were then placed in a 32 oz. mason jar with glass beads underlying perforated aluminum foil on the bottom of the jar. For each horizon, the subsamples originating from the 3 field sampling locations within a given plot were put into a single jar, creating 1 composite sample per plot, per horizon. For each 10 cm mineral soil core, a small portion of soil was removed using a spatula vertically along the length of the core. Once in jars, deionized water was added to each sample to approximate field capacity and samples were placed in the dark at 15oC. Soil CO2 respiration was measured using an automated soil incubation system equipped with an infrared gas analyzer (Li-820, Licor, Inc, Lincoln, NE) (Brachoand others unpublished). To estimate 14CO2, CO2 was first removed from each jar by scrubbing with magnesium perchlorate and soda lime for 5 min. Jars were then left at 15oC for 3-5 days to allow CO2 to accumulate. The headspace air was then pulled through a zeolite molecular sieve trap (Alltech 13X, Alltech Associates, Deerfield, IL) to absorb the CO2, baked, reduced to graphite (Hicks Pries and others 2013), then sent to the UC Irvine W.M. Keck Carbon Cycle Accelerator Mass Spectrometry Laboratory for final analysis.

REFERENCES

Alexander HD, Mack MC, Goetz S, Beck PSA, Belshe EF. 2012. Implications of increased deciduous cover on stand structure and aboveground carbon pools of Alaskan boreal forests. Ecosphere 3: 1-21.

Bond-Lamberty B, Wang C, Gower ST. 2003. The use of multiple measurement techniques to refine estimates of conifer needle geometry. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 33: 101-105.

Brown JK. 1974. Handbook for inventorying downed woody material. Ogden, Utah, USA: USDA Forest Service, Intermountain Forest and Range Experiment Station.

Gaudinski JB, Dawson TE, Quideau S, Schuur EAG, Roden JS, Trumbore SE, Sandquist DR, Oh SW, Wasylishen RE. 2005. Comparative analysis of cellulose preparation techniques for use with 13C, 14C, and 18O isotopic measurements. Analytical Chemistry 77: 7212-7224.

Hicks Pries CE, Schuur EAG, Crummer KG. 2013. Thawing permafrost increases old soil and autotrophic respiration in tundra: Partitioning ecosystem respiration using δ13C and Δ14C. Global Change Biology 19: 649-661.

Jones BM, Breen AL, Gaglioti BV, Mann DH, Rocha AV, Grosse G, Arp CD, Kunz ML, Walker DA. 2013. Identification of unrecognized tundra fire events on the north slope of Alaska. Journal of Geophysical Research-Biogeosciences 118: 1334-1344.

Kuo S. 1996. Methods of soil analysis. Bartels JM editor. Methods of soil analysis part 3- chemical methods. Madison: Soil Science Society of America, p893-894.

Mack MC, Treseder KK, Manies KL, Harden JW, Schuur EAG, Vogel JG, Randerson JT, Chapin FS, III. 2008. Recovery of aboveground plant biomass and productivity after fire in mesic and dry black spruce forests of interior alaska. Ecosystems 11: 209-225.

Murphy J, Riley JP. 1962. A modified single solution method for determination of phosphate in natural waters. Analytica Chimica Acta 26: 31-&.

Nalder IA, Wein RW, Alexander ME, deGroot WJ. 1997. Physical properties of dead and downed round-wood fuels in the boreal forests of Alberta and Northwest Territories. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 27: 1513-1517.

Robertson GP, Coleman DC, Bledsoe CS, Sollins P editors. 1999. Standard soil methods for long-term ecological research. New York: Oxford University Press.

Ter-Mikaelian MT, Colombo SJ, Chen JX. 2008. Amount of downed woody debris and its prediction using stand characteristics in boreal and mixedwood forests of Ontario, Canada. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38: 2189-2197.

Vogel JS, Southon JR, Nelson DE. 1987. Catalyst and binder effects in the use of filamentous graphite for AMS. Nuclear Instruments & Methods in Physics Research Section B-Beam Interactions with Materials and Atoms 29: 50-56.

Appendix2. ForestStand Characteristics for Each Spatially Interspersed Block of Black Spruce andPaper Birch across the Study Site

All values are mean (SE), n= 5 plots per block.

Trees ≥ 1.4 m in height are included in the rows labeled DBH (diameter at breast height) and trees < 1.4 m tall are included in the BD (basal diameter) category.

Other species includes all measure trees that were not the focus species studied within the given plot.

Appendix3.TotalElemental Concentrations for Mineral Soils (0-100 cm) for Studied Black Spruce and Alaska Paper Birch Stands

Mean (SE), n=3 samples per species, per depth increment.Cells containing ND indicate no data available.