Plant mycorrhizal associations influence the accumulation and persistence of soil organic matter and could therefore shape ecosystem biogeochemical responses to global changes that are altering forest composition. For instance, arbuscular mycorrhizal (AM) tree dominance is increasing in temperate forests, and ericoid mycorrhizal (ErM) shrubs can respond positively to canopy disturbances. Yet how shifts in the co-occurrence of trees and shrubs with different mycorrhizal associations will affect soil organic matter pools remains largely unknown. We examine the effects of ErM shrubs on soil carbon and nitrogen stocks and indicators of microbial activity at different depths across gradients of AM versus ectomycorrhizal (EcM) tree dominance in three temperate forest sites. We find that ErM shrubs strongly modulate tree mycorrhizal dominance effects. In surface soils, ErM shrubs increase particulate organic matter accumulation and weaken the positive relationship between soil organic matter stocks and indicators of microbial activity. These effects are strongest under AM trees that lack fungal symbionts that can degrade organic matter. In subsurface soil organic matter pools, by contrast, tree mycorrhizal dominance effects are stronger than those of ErM shrubs. Ectomycorrhizal tree dominance has a negative influence on particulate and mineral-associated soil organic matter pools, and these effects are stronger for nitrogen than for carbon stocks. Our findings suggest that increasing co-occurrence of ErM shrubs and AM trees will enhance particulate organic matter accumulation in surface soils by suppressing microbial activity while having little influence on mineral-associated organic matter in subsurface soils. Our study highlights the importance of considering interactions between co-occurring plant mycorrhizal types, as well as their depth-dependent effects, for projecting changes in soil carbon and nitrogen stocks in response to compositional shifts in temperate forests driven by disturbances and global change.
We worked in a 3,213-ha second-growth, mixed-hardwood forest in Connecticut, USA (41°57’ N, 72°07’ W). We established 18 10-m radius plots, each containing a pair of 1-m radius subplots (n =36), evenly arrayed across three forest stands that contained areas of both high AM and high EcM tree relative basal area as well as a patchy distribution of the ErM shrub Kalmia latifolia.
Within each of the 18 plots, we established paired 1-m radius subplots with and without K. latifolia in the understory ( “+/- ErM subplot”) within 2 m of the center of the 10-m radius plot. In each 1-m radius subplot, we measured trees ≥1 cm diameter at breast height (DBH; 1.37 m). We also measured DBH of all trees ≥20 cm DBH within 10 m and trees ≥5 cm DBH within 5 m of plot center. We calculated the percentage of EcM tree basal area out of total basal area, scaled to m2 ha-1.
In June 2021, we collected and pooled two soil samples for each of three depths within the 36 paired subplots (i.e. 18 +ErM and 18 -ErM subplots). The three depths included: (1) the Oa horizon (depth varied depending on the thickness of the horizon); (2) the top 10 cm of the A horizon, beginning at the base of the Oa horizon; and (3) a second, contiguous A horizon sample that reached a cumulative sampling depth of 30 cm, inclusive of the depth of the Oa horizon. For the organic layer, we removed the litter layer (i.e. the Oi and Oe horizons) and collected and pooled two 25 by 25-cm areas of the Oa horizon using a square template. For the mineral layers, we collected two contiguous depth increments from the A horizon within the footprint of the 25 by 25-cm areas using a 5.08-cm diameter hammer corer. In each instance, we recorded the exact sampling depth. Two subplots did not have an Oa horizon, so we collected a total of 106 samples (3 sites × 6 plots × 2 subplots × 3 depths − 2 Oa samples). Soils were stored at 4°C prior to their analysis.
To prepare the soil samples for analysis, we weighed and homogenized each sample, air dried a representative subsample of non-sieved soil, and passed the remaining field-moist sample through a 4-mm sieve. Using the non-sieved subsample, we estimated the mass and volume of roots and stones and calculated soil bulk density values. For total soil organic matter (SOM) content, we heated samples at 550°C for 12-h in a muffle furnace and calculated loss on ignition.
We used a modified substrate-induced respiration method as an indicator of active saprotrophic microbial biomass. Using autolyzed yeast extract solution as a labile C substrate, we measured rates of CO2 efflux over a 4-h incubation period with an Infra-Red Gas Analyzer and calculated the rate of C-CO2 production per unit of equivalent soil dry mass. For microbially-available C, we estimated potential CO2 production rates over a 14-d incubation period. We measured CO2 efflux over 24-h periods at days 1, 5, 8, and 14 and integrated the four measurements to calculate cumulative C-CO2 production. We estimated water holding capacity by saturating each field-moist sample with water and allowing it to drain freely for 2 h. To calculate the equivalent dry mass of field-moist samples, we measured gravimetric water content by oven-drying the samples to constant mass at 105°C.
We separated the >53 and <53 µm particle size fractions to quantify particulate (POM) and mineral-associated soil organic matter fractions. We passed air-dried samples through a 2-mm sieve and then dispersed soil aggregates by shaking ~30 g of the sieved, air-dried sample with 30 mL of sodium hexametaphosphate (NaHMP) solution for 18 h. We rinsed each sample over a 53-µm sieve with deionized water until the water passing through the sieve ran clear. We oven-dried the >53-µm fraction retained on the top of the sieve and a representative subsample of the <53-µm fraction suspended in solution at 70°C. To estimate the mass of the <53-µm fraction, we calculated the difference between the initial soil mass (105°C equivalent) and the recovered mass of the >53-µm fraction (105°C equivalent). To convert air-dried soil mass to oven-dried mass we dried a subsample of each air-dried sample at 105°C. Fractions were ground to a fine powder and analyzed for total carbon (C) and nitrogen (N) concentrations using a Costech ESC 4010 Elemental Analyzer.
We used an equivalent soil mass approach to calculate soil C, N, SOM, microbial biomass, and microbially-available C stocks in three equivalent soil mass layers as well as the sum of the three layers to estimate cumulative stocks at the subplot level. Following this approach, we report stocks to a standard soil mass and therefore allow the depth of the equivalent soil mass layers to vary depending on soil bulk density. To calculate equivalent soil mass stocks, we added or subtracted elemental stocks of the deeper soil layer to the upper soil layer in 1-mm increments until the soil mass from the upper layer is closest to that of the target soil mass. We chose reference soil masses using the median or target field sampling depth and the mean bulk density value for each of the three depth increments to make them roughly equivalent to the sampled depths. Based on this method, the organic layer had an equivalent mass of ~2.5 kg soil m-2 (median Oa depth = 2.5 cm; mean Oa bulk density = 0.10 g cm-3), the surface mineral layer had an equivalent mass of ~37 kg soil m-2 (target sampling depth = 10 cm; mean bulk density = 0.37 g cm-3), and the subsurface mineral layer had an equivalent mass of ~126 kg soil m-2 (the target sampling depth was 17.5 cm for a sample with a 2.5 cm Oa depth; mean bulk density = 0.72 g cm-3). The cumulative equivalent soil mass for the subplot-level stocks was the sum of the three layers, or ~166 kg soil m-2.
This dataset is comprised of two tabs in a single excel file. See the "README.md" file for information pertaining to the variables measured and analyzed. The "Data by soil layer" tab includes all data collected or analyzed at the resolution of the three separate soil depths (i.e. the organic, surface mineral, and subsurface mineral soil layers). The "Data by subplot" tab includes all data collected or analyzed at the resolution of the subplot, including the cumulative carbon and nitrogen stocks that are the sum of the three individual soil layers.
All equivalent soil mass (esm1) stocks were calculated using the the SimpleESM R script (Ferchaud et al. 2023) available at https://github.com/fabienferchaud/SimpleESM and described in the following reference:
Fabien Ferchaud, Florent Chlebowski, Bruno Mary. SimpleESM: R script to calculate soil organic carbon and nitrogen stocks at Equivalent Soil Mass. Institut National de Recherche pour l’Agrgriculture, l’Alimentation et l’Environnement (INRAE). 2023. ffhal-04013158f. https://hal.science/hal-04013158/document