Soil Warming Plus Nitrogen Addition Experiment at Harvard Forest since 2006

Climate warming and N deposition are occurring on a global scale with unknown long-term effects on soil microbial communities and the biogeochemical processes they perform. Few studies have examined the interactive effects of elevated temperatures and N additions on soil microbial community structure and function. The overall objective of this study is to investigate whether warming and N additions restructure microbial communities and alter the response of soil C pools to these two stressors. A related study is examining the interactive effects of warming and N additions on plant and ant diversity. This research is being carried out at the Soil Warming x Nitrogen Addition Study at the Harvard Forest which includes four treatments: control, warming (heating to 5 deg C above ambient), warming x N, and N additions only (addition of 50 kg N/ha/yr). Soil respiration measurements have been made monthly since the beginning of the experiment in 2006. In 2010 and 2011, two different methods were compared: static chamber measurements and instantaneous field IRGA assessments. Soil samples (~0-10 cm) have been sampled annually for total C and N, N mineralization, and microbial community composition. Most recently, soils were collected in October 2011 from across the entire profile (0-50 cm) to access potential changes in soil C and N pools with depth. First, 20 x 20 cm forest floor samples were collected. Mineral soils were then collected in 10 cm depth increments to ~50 cm. Samples are currently being analyzed for total C and N, microbial biomass and community composition and fungal gene expression (transcriptomics). Additionally, long-term incubations are being conducted to measure labile and recalcitrant C fractions. Additional soil physical (texture) and chemical (pH, inorganic N) are being measured. Field season measurements of soil respiration indicate that both warming and N additions continue to stimulate CO2 flux, with warming treatments having a stronger effect on respiration than N additions alone. Where warming x N occur together, warming appears to moderate the negative effects of N additions alone on soil respiration. Microbial biomass estimates show the greatest biomass within the warming x N treatment. Declines in microbial biomass with warming or N additions alone support similar findings from the same assessment in 2007 and 2009. Forest floor mass show small declines under warmed conditions. Soil C storage in the mineral soil showed modest changes with warming (-10% difference compared to control). Genomic and transcriptomic pipelines optimized in the Frey Lab are being used to measure the diversity, composition, and function (via gene expression) of the active fungal community in response to warming and N additions. Recent transcriptomic work has demonstrated that chronic N fertilization results in a decrease in expression of several transcripts of fungal laccase and glycoside hydrolases which are enzymes involved in lignocellulose degradation. Overall, our recent results suggest that anthropogenic stressors and seasonal changes continue to interact to affect soil microbial communities and biogeochemical cycles. Plant roots are primary drivers of soil organic matter dynamics, mediating belowground carbon (C) inputs, stabilization, and losses. Yet, how global changes such as rising temperatures and altered nitrogen (N) availability interact to affect these dynamics has rarely been tested empirically in the field. Here, we quantify how inputs to soil organic matter from fine root production, root exudates, and root-associated fungi respond to long-term (16 years) soil warming (+5°C), nitrogen (N) enrichment (+5 g N m⁻² yr⁻¹), and their combination in a temperate hardwood forest. Warming alone reduced root-derived C inputs by 21% and increased microbial respiration by 46%, resulting in a net soil C loss of 135 g C m⁻² yr⁻¹. In contrast, N enrichment increased root-derived SOC accumulation by 47% and reduced root respiration by 40%, contributing to a near neutral soil C balance. When combined, warming × N addition increased root-derived SOC fourfold (from 70 to 281 g C m⁻² yr⁻¹), fully offsetting warming-induced C losses and maintaining soil C stocks at control levels. Root-derived SOC accumulation was positively related to fine root production (r² = 0.42) and to maple:oak exudate ratios (r² = 0.31), highlighting species-specific control over C stabilization. These findings demonstrate that interacting global change factors can have balancing effects on root C allocation and microbial losses, highlighting soil N availability as a critical control determining whether warming accelerates SOC depletion or stabilizes new root-derived C.

This dataset is released to the public under Creative Commons CC0 1.0 (No Rights Reserved). Please keep the dataset creators informed of any plans to use the dataset. Consultation with the original investigators is strongly encouraged. Publications and data products that make use of the dataset should include proper acknowledgement.

The experiment consists of four treatments (control, heated, heated +N, and +N only) with six replicates per treatment in a completely randomized design. The warming and N fertilization treatments were initiated in August 2006. Like the Soil Warming Study at Prospect Hill, average soil temperatures in the heated plots are continuously elevated 5 deg C above ambient by the use of buried heating cables placed at 10 cm depth in the soil and spaced 20 cm apart. The heating cables are controlled by a data logger that monitors thermistors every 10 min. Plots automatically turn on and off to maintain a 5 deg C temperature difference between the heated and control plots. The N addition plots (heated +N, +N only) are fertilized following the protocol of the Chronic Nitrogen Addition Study. An aqueous solution of NH4NO3 is applied at a rate equivalent to the low N plots at the chronic N study (5 g m-2 yr-1). Fertilizer is applied in equal monthly doses during the growing season (Apr-Oct). The unfertilized control plots and heated only plots are sprayed with water as a moisture control. Soil CO2 flux is measured by placing static chambers over pre-installed collars for 15 minutes, and taking headspace samples from these chambers at five minute intervals. Sampling occurs between 1100 and 1300 h. In the winter months, collars were dug from underneath the snow and allowed to equilibrate with the atmosphere for one hour prior to sampling. Gas samples are stored in air-tight, plastic syringes until they are transported back to the University of New Hampshire, and are analyzed using an infrared gas analyzer. The change in CO2 concentrations over the 15 minute sampling period was used to calculate net CO2 flux rates in units of mg CO2-C m-2 h-1. Plot temperatures are measured by thermistors buried to 5 cm depth in both heated and adjacent, unheated plots. Heated plots contain five thermistors each, while unheated plots contain two. The thermistors are connected to a Campbell Scientific CR10 datalogger, which reads them every 15 minutes and records the average plot temperature every hour. For the ant study, two pitfall traps were placed within each plot. For 2006-2007, a pitfall trap consisted of one ten ounce cup placed flush with the ground. Approximately 20 ml of soapy water was put in each pitfall trap and trap contents were collected after 48 hours. For 2008-2009, a pitfall trap consisted of a 50 mL centrifuge placed in a PVC sleeve and filled with ~25 mL of ethanol with trap contents collected after 48 hours. Pitfalls were set to trap ants on days with clear, sunny weather and no rain. Ant sampling by S. Record occurred on: 28-30 June 2006, 11-13 July 2007, and 11-12 July 2009. M. Kaspari sampled ants in July 2008. Useful characteristics for distinguishing A. picea from A. rudis are still being worked out (and we used to call them all A. rudis), so this distinction will not be meaningful across years. To sample plants, S. Record counted all stems of every plant in each 3m x 3m plot and identified them to species when possible. Plant sampling dates were: 15-27 July 2006, 31 July 2007-16 Aug 2007, 2-13 Aug 2008, and 11-15 July 2009. Fine roots (less than 2 mm diameter) were separated from fresh soil using forceps. Briefly, soil samples were placed onto a clean 2 mm sieve. Roots were picked gently from the soil, separating live root tissues into either absorptive (1st, 2nd, and 3rd order) or transport root (4th and 5th order) functional types based on branching order. Absorptive roots were further divided into tree species (maple or oak) based on color, morphology, and presence of EM or AM associations. Specifically, oak roots (Quercus spp.) were separated based on their darker color, root branching, and number of root tips. In contrast, maple roots (AM-associated) were identified based on lighter-color and larger-diameter characteristics. A stereomicroscope was used to aid the visualization and sorting of fine roots and confirm mycorrhizal status. Roots were picked in 5-minute intervals, recording the mass returned after each interval until a period of stabilized diminishing returns was reached. Dead roots were identified and discarded based on their dark discoloration, poor adhesion between the stele and cortex, and brittleness. Fine roots from each functional pool (absorptive vs. transport) and species (oak vs maple) combination were cleaned with deionized water to remove soil and organic debris, dried at 60°C, and weighed. Absorptive roots were finely ground using a stainless-steel ball mill grinder, and N concentrations were determined using a Perkin Elmer 2400 Series II CHN Elemental Analyzer. During root processing, a random subsample of oak (EM-associated) absorptive fine roots from the organic horizon was collected, washed gently with D.I. water, and stored in 2x CTAB solution (100 mM Tris-HCl (pH 8.0), 1.4 M NaCl, 20 mM EDTA, 2% cetyltrimethylammonium bromide). EM colonized root tips were visually morphotyped based on color, ramification, and hyphal color (Agerer, 2001). At least 100 root tips were counted per sample, and the number of root tips colonized was divided by the total number of root tips counted to determine EM colonization rate. Starting in spring 2021, fine roots were collected the first week of each month during the growing season (May-Oct) from the organic horizon of each plot to measure root respiration using a standard approach (Burton et al., 2008; Paradiso et al., 2019). This process involved removing the litter layer until one maple and one oak absorptive root system were identified. Roots (at least 10 cm long) were collected within 2-5 cm of the soil surface, severed, immediately transported to an onsite field facility, and confirmed for species identity. Roots were then carefully cleaned to remove organic debris, and respiration rates were determined within 15 minutes of collection. Previous work on fine roots has indicated that rates of root physiological traits attenuate after 2 hours (Bloom and Caldwell, 1988), and specifically, root respiration in temperate forests has been shown to remain stable for up to 4 hours following removal from soil (Burton et al., 2002). To measure root respiration, absorptive roots were individually inserted into a sealed glass chamber (175 ml) connected using polyethylene tubing (19 ml) to a LICOR 8100a field infrared gas analyzer (LI-COR Biogeosciences) that measured CO2 concentrations every second for a 5-minute interval. Each interval contained a 15-second chamber pre-purge and a 15-second chamber post-purge. We focused our measurements on absorptive fine roots to standardize the process between tree species (Freschet and Roumet, 2017; McCormack et al., 2015). Upon returning to the laboratory, roots were stored at 4°C for up to 1 week before being dried at 60°C for 48 hours for determination of dry mass. We calculated specific root respiration by fitting a linear model to the relationship between time and CO2 concentration (ppm) within the glass chamber using Soil Flux Pro (Hoffman et al., 2019) and then converting the slope of this linear model to CO2 respiration flux in nmol CO2 s-1. We divided root respiration measurements by dry root biomass to quantify mass-specific root respiration (nmol CO2 g-1 s-1). Mass-specific root respiration was scaled to the ecosystem level (hereafter “ecosystem root respiration”) as the product of specific fine-root respiration (nmol CO2 g-1 s-1) and absorptive root biomass (g m-2), expressed as nmol CO2 g-1 m-2 s-1 for each tree species (Jarvi and Burton, 2020; Melillo et al., 2011; Tunison et al., 2024). We quantified new belowground carbon inputs using ingrowth cores filled with C4-derived soil (distinct δ¹³C) installed in replicated SWaN treatment plots from May 2021 to Oct 2022. Cores included (i) a root + hyphae compartment installed to 10 cm depth, (ii) a hyphae-only insert (roots excluded by fine mesh) installed to 7 cm depth, and (iii) paired exclusion controls to estimate background δ¹³C and non-ingrowth contributions. After harvest, soils were processed and analyzed for total C and δ¹³C. Root- and fungal-derived SOC were calculated from the shift in soil δ¹³C relative to controls using a two-endmember mixing model and converted to areal C accumulation using soil C concentration, bulk density, and core dimensions. Annualized rates were calculated by dividing by two growing seasons. Live fine roots recovered from root + hyphae cores were separated, assigned to Acer (AM-associated) or Quercus (EM-associated) based on morphology, dried, weighed, and converted to C using elemental analysis. Production rates were annualized over the two-year deployment and scaled by core area. The δ¹³C values (‰) used for species inputs in isotope-based calculations were treatment-specific: Control—Acer −30.23, Quercus −28.60; N addition—Acer −27.95, Quercus −28.02; Heated—Acer −28.92, Quercus −28.06; Heated×N addition—Acer −31.07, Quercus −28.54. Exudates were collected in five campaigns (May-Aug 2023) by incubating intact absorptive roots in C-free solution, including root-free blanks. Dissolved organic carbon was measured by TOC analysis (NPOC), blank-corrected, normalized to root dry mass and incubation time, and scaled to ecosystem rates using plot-level absorptive root biomass estimates.

Citations Bloom, A. J., and Caldwell, R. M. (1988). Root Excision Decreases Nutrient Absorption and Gas Fluxes 1. Plant Physiology, 87(4), 794–796. Burton, A. J., Melillo, J. M., and Frey, S. D. (2008). Adjustment of forest ecosystem root respiration as temperature warms. Journal of Integrative Plant Biology, 50(11), 1467–1483. https://doi.org/10.1111/j.1744-7909.2008.00750.x Burton, A. J., Pregitzer, K., Ruess, R., Hendrick, R., and Allen, M. (2002). Root respiration in North American forests: Effects of nitrogen concentration and temperature across biomes. Oecologia, 131(4), 559–568. https://doi.org/10.1007/s00442-002-0931-7 Freschet, G. T., and Roumet, C. (2017). Sampling roots to capture plant and soil functions. Functional Ecology, 31(8), 1506–1518. https://doi.org/10.1111/1365-2435.12883 Jarvi, M. P., and Burton, A. J. (2020). Root respiration and biomass responses to experimental soil warming vary with root diameter and soil depth. Plant and Soil, 451(1), 435–446. https://doi.org/10.1007/s11104-020-04540-1 McCormack, L., Dickie, I. A., Eissenstat, D. M., Fahey, T. J., Fernandez, C. W., Guo, D., Helmisaari, H.-S., Hobbie, E. A., Iversen, C. M., Jackson, R. B., Leppälammi‐Kujansuu, J., Norby, R. J., Phillips, R. P., Pregitzer, K. S., Pritchard, S. G., Rewald, B., and Zadworny, M. (2015). Redefining fine roots improves understanding of below-ground contributions to terrestrial biosphere processes. New Phytologist, 207(3), 505–518. https://doi.org/10.1111/nph.13363 Melillo, J. M., Butler, S., Johnson, J., Mohan, J., Steudler, P., Lux, H., Burrows, E., Bowles, F., Smith, R., Scott, L., Vario, C., Hill, T., Burton, A., Zhou, Y.-M., and Tang, J. (2011). Soil warming, carbon–nitrogen interactions, and forest carbon budgets. Proceedings of the National Academy of Sciences, 108(23), 9508–9512. https://doi.org/10.1073/pnas.1018189108 Muratore, T., Knorr, M., Simpson, M., Stephens, R., Phillips, R., and Frey, S. (In press). Response of root respiration to warming and nitrogen addition depends on tree species. Global Change Biology. Paradiso, E., Jevon, F., and Matthes, J. (2019). Fine root respiration is more strongly correlated with root traits than tree species identity. Ecosphere, 10(11), e02944. https://doi.org/10.1002/ecs2.2944