ALBERT

Description: Increasing concentrations of dissolved organic carbon (DOC) have been identified in many freshwater systems over the last three decades. While studies have generally nominated changes in temperature and atmospheric deposition as determinants of DOC concentrations, there is still much uncertainty concerning net effects on DOC concentrations and export dynamics. We leveraged long-term datasets (1988-2013) from a first-order tributary of Lake Superior to understand changes in DOC concentrations and exports from the watershed, particularly through snowmelt dynamics. We observed an increase in DOC concentrations of approximately 0.12 mg C L -1 year -1 , which were coincident with regional warming and declines in sulfate deposition. DOC fluxes were most strongly related to snowmelt dynamics, with peak snow water equivalencies generally being lower and less variable in the 21 st century, compared with measurements taken in the 1980s and 1990s. As a result of the countervailing changes in DOC concentrations and runoff patterns, we observed no significant trends in DOC export through time. Based on these trends, any future changes in climate that lessen the dominance of the spring freshet on annual runoff dynamics would decrease DOC export in snowmelt dominated systems, but identifying the mechanism of increasing DOC concentrations is needed to fully understand the future effects on DOC export. Together these findings illustrate complex interactions between climate and atmospheric deposition in carbon cycle processes, and highlight the importance of long-term monitoring efforts for understanding the consequences of a changing climate.

Description: We measured organic layer (OL) recovery and carbon stocks in dead woody debris and soil pools a decade following wildfire in black spruce forests of interior Alaska. Previous study at these research plots has shown the strong role landscape position plays in governing the proportion of OL consumed during fire, and post-fire revegetation. Here, we show that landscape position likely influences fire dynamics in these stands through changes in mineral soil texture. The content of fine textured materials in underlying mineral soils was positively related to OL depths measured one and ten years post-fire, and there was an interaction between soil texture and elevation in governing OL consumption, and OL recovery a decade following fire. OL depths a decade post-fire were 2 cm greater than one year post-fire, with a range of 19 cm of accumulation to 9 cm of subsidence. Subsidence was inversely related to the percentage of fine textures within the parent material. The most influential factor determining the accumulation of soil organic carbon stocks a decade following wildfire was the interaction between landscape position and the presence of fine textured soil. As such, parent material texture interacted with biological processes to govern the recovery of organic soils.

Description: The following analysis was performed using a Costech Elemental Combustion System 4010 connected to a Thermo Finnigan ConfloIII Interface and Deltaplus Continuous Flow-Stable Isotpe Ratio Mass Spectrometer. IAEA, USGS, and NIST certified isotopic standards are run at the beginning of each analysis. One certified standard is also run at the end of the analysis to check for stability of the calibration. These standards are recognized internationally, and are used to calibrate the CO2 and N2 reference gases which are analyzed in conjunction with every individual sample. Values are reported on the VPDB scale for d13C and the Atm. air scale for d15N. An internal standard is usually run every 10 samples. The precision of the certified isotopic standards are typically 0.2 to 0.5 per mil, so the best analytical precision we report is +/- 0.25 per mil for d13C and +/- 0.5 per mil for d15N. Therefore, representations of isotopic values in resulting graphs and text should be to one digit beyond the decimal point. Please contact Jennfer Eikenberry with any questions. Please also refer to the Run ID for questions about specific samples.

Description: Peat was stored in a refrigerator or freezer (samples processed at later date) until processing (storage time appoximately 3-8 months). A weight weight was obtained of the full sample, then it was cut in half, weighed again, dried completely at 60°C and a dry weight was recorded.Peat cubes (10x10x10 cm) were obtained at 5 depths, from 25 different locationswith depths of 0-10 cm, 10-20 cm, 20-30 cm, 50-60 cm, 70-80 cm

Description: In 2010, we harvested 24 intact 1-m-3 peat monoliths from an oligotrophic peatland in Meadowlands, MN USA (N 47.07278°, W 92.73167°) and subsequently installed them in the Mesocosm facility at the USDA Forest Service Northern Research Station Forestry Sciences Laboratory in Houghton, MI USA (N 47.11469°, W 88.54787°). Beginning in 2011, we initiated a full-factorial randomized complete block manipulation of water table levels and vegetation types. We manipulated water tables to reflect either “high” or “low” levels using a 45-year water table record from the Marcell Experimental Forest (USDA Forest Service) near the peat harvest location. The vegetation type treatments included three levels: unmanipulated (“Control”), all Ericaceae removed (“Sedge”), or all sedges removed (“Ericaceous”). To initiate treatments, we removed all aboveground and, when possible, belowground plant material. In 2014, 8 bins were selected for radiocarbon analysis of respired CO2 and dissolved inorganic carbon (DIC). Two mesocosms were chosen for each treatment combination (e.g., two bins with 'Sedge' and 'low' water table treatments) for each manipulated vegetation type. Bins were randomly selected from the stratified design across all experimental blocks. To collect respired CO2, we placed the 100cm x 100cm x 40cm silicone-sealed opaque chambers used for CH4 measurements over the mesocosm bins. We taped the chamber to create an airtight seal to the rim of the mesocosm. The chamber was in a closed loop with an IRGA (described above) to monitor headspace CO2 levels. To scrub the chamber headspace of ambient air, we employed another closed loop with a pump (3.74 m3 hr-1; 115 VAC oil-less, Cole Parmer #EW-07061-60) connected to a 5.1 diameter column containing 400g fresh Soda Lime (Indicating Type 4-8 Mesh Baker Analyzed ACS #3448-05-cs). We circulated chamber air through the Soda Lime to remove atmospheric CO2 for two hours, which brought the headspace CO2 down from ambient to approximately 130 ppm. At the end of the scrubbing period, we recorded the CO2 concentration, disconnected the pump lines, and allowed CO2 to accumulate in the chamber until a concentration of approximately eight times (〉1000ppm) initial scrubbed concentration was achieved in the static chamber. A purpose-built air capture apparatus (6 1 volume) consisting of an airtight stainless-steel cylinder outfitted with an internal battery-operated CPU fan was also connected in a closed loop with the flux chamber. We verified the CO2 concentration within the apparatus, and then the attached cold trap was lowered into liquid nitrogen (N) to trap CO2; the CPU fan situated above the trap ensured a heterogeneous pool of CO2 within the apparatus. After 10 minutes we sealed the trap and transferred it in a dewer filled with liquid N to the USDA Northern Research Station radiocarbon facility on site and immediately connected to a vacuum line for CO2 purification (removal of non-condensables) and transfer to evacuated glass tubes to be sealed for conversion to graphite. Purified CO2 was then graphitized with an iron powder catalyst (99.99%) in a method modified from Vogel et al. (1984). Graphite targets were sent to the Lawrence Livermore National Laboratory in Berkeley, CA USA for analysis on an accelerator mass spectrometer. The 13C was calculated on splits of the same sample based on the permil ratio of 13C/12C relative to the international standard Vienna Peedee belemnite. Radiocarbon values were normalized to a 13C of -25‰ to account for mass-dependent fractionation and 2014 was the year of measurement for decay corrections. To obtain DI14C, 60ml of porewater from 20, 40, and 70cm below peat surface from the same bins as selected for respiration 14CO2 were collected after flushing and placed in pre-evacuated 125ml glass serum bottles via septa. The collections were made through pre-installed micropiezometers made of ultra‐high‐density polyethylene casing with inner Teflon tubing, where each depth represented 10cm centered around the three depths stated above. After shaking, 10ml of headspace air was removed, purified, graphitized and analyzed for radiocarbon following the same procedure described above.

Description: Globally important carbon (C) stores in northern peatlands are vulnerable to oxidation in a changing climate. A growing body of literature draws attention to the importance of dissolved organic matter (DOM) in governing anaerobic metabolism in organic soil, but exactly how the reduction-oxidation (redox) activities of DOM, and particularly the phenolic fraction, are likely to change in an altered climate remain unclear. We used large mesocosms in the PEATcosm experiment to assess changes in peatland DOM and redox potential in response to experimental manipulations of water table (WT) position and plant functional groups (PFGs). WT position and PFGs interacted in their effects on redox potential and quantity and quality of DOM. Phenolics were generally of higher molecular weight and more oxidized with sedges in lowered water tables. Altered DOM character included changes in dissolved nitrogen (N), with higher N:phenolics with higher E4:E6 (absorbance ratio = 465:665) DOM in the lowered WT and sedge PFG treatments. Conversely, biomolecular assignments to free amino-sugars were largely absent from low WT treatments. Drainage resulted in the creation of unique N compounds which were more condensed (lower H:C), that changed with depth and PFG. The accumulation of oxidized compounds with low WT and in sedge rhizospheres could be very important pools of electron acceptors beneath the water table, and their mechanisms of formation are discussed. This work suggests the effects of changes in vegetation communities can be as great as WT position in directly and interactively mediating peat redox environment and the redox-activity of DOM.

Description: Peatlands contain enormous carbon stocks, but the stability of this carbon is variable. Peatlands can vary in tree cover from completely open to forested with associated differences in peat quality. Peat quality, or potential for mineralization, is a contributing factor affecting how the carbon balance of peatland ecosystems could change with climate or land use changes. We compared the peat quality of open peatlands dominated by Sphagnum mosses to forested, or silvic, peatlands dominated by black spruce and tamarack or northern white cedar to quantify the effects of different carbon sources on peat quality. We used Fourier-transform infrared spectrometry (FTIR) to analyze peat properties throughout the depth profile of 30 peat cores across the hemi-boreal Upper Great Lakes region. We found that tree cover was associated with differences in both surficial and deep peat quality. Silvic peat had lower peat quality than Sphagnum peat as shown by FTIR indices. Sphagnum peat also had significantly higher peat quality at the surface compared to at depth. However, silvic peat showed no significant difference with depth in any indices. Our results indicate that the dominant plant functional type is a strong driver of peat quality as we identified key differences between silvic and Sphagnum peatlands. These relatively local differences are similar in magnitude to those found across biomes comparing tropical swamps to boreal Sphagnum peatlands. This implies that the dominant plant functional type (e.g. tree, shrub, graminoid, or moss) may be more important to peat quality than species identity - or even latitude - in peatlands.

Description: We measured organic-layer (OL) recovery and carbon stocks in dead woody debris a decade after wildfire in black spruce (Picea mariana (Mill.) B.S.P.) forests of interior Alaska. Previous study at these research plots has shown the strong role that landscape position plays in governing the proportion of OL consumed during fire and revegetation after fire. Here, we show that landscape position likely influences fire dynamics in these stands through changes in mineral soil texture. The content of fine-textured materials in underlying mineral soils was positively related to OL depths measured 1 and 10 years after fire, and there was an interaction between soil texture and elevation in governing OL consumption and OL recovery a decade following fire. OL depths 10 years after fire were 2 cm greater than 1 year after fire, with a range of 19 cm of accumulation to 9 cm of subsidence. Subsidence was inversely related to the percentage of fine textures within the parent material. The most influential factor determining the accumulation of OL carbon stocks a decade following wildfire was the interaction between landscape position and the presence of fine-textured soil. As such, parent material texture interacted with biological processes to govern the recovery of soil organic layers.