Site selection
We established 60 plots distributed across eight distinct forest sites
in northern New Hampshire and Vermont, USA (Figure 1). Plots were 10 m
in diameter (78.54 m2 ) and each plot was
dominated (> 50% basal area) by one of six AM- or ECM-associated
tree species: AM species included white ash (Fraxinus
americana L.), sugar maple (Acer
saccharum Marsh.), and northern white cedar
(Thuja occidentalis L.), and the ECM-associated
tree species included yellow birch (Betula
alleghaniensis Britton), American beech (Fagus
grandifolia Ehrh.), and eastern hemlock (Tsuga
canadensis L. Carrière). Though a plot was considered
dominated by each target species when the proportion of that species’
basal area exceeded 50% of the total basal area, the target species
comprised more than 90% of the total basal area in 57% of plots.
Because tree species tend to grow in particular soil types that may
vary in the properties that affect MAOM formation, we sampled multiple
species within close geographical range at most sites and ensured that
each species was sampled at multiple sites to capture the range of
soil conditions wherein each species can be found. Further, we took
care to identify sites with little or no documented human disturbance
in the last fifty years (i.e. conservation areas, experimental forest,
and privately-owned land with known history) to minimize differences
in soil properties based on past land use.
Soil collection
All soil was sampled during the snow-free seasons in 2017, 2018, or
2019. At the time of establishment, we collected one soil sample from
the organic horizon and one from the uppermost mineral horizon at the
center of each plot. Organic (O) horizon soil was collected from the
top of the O horizon, immediately below the loose leaf litter. Mineral
soil was collected from the upper 10 cm of mineral soil except when an
E horizon was present. In such cases, we collected soil from the 10 cm
immediately below the E horizon. Samples were sieved to remove rocks
and roots > 2 mm in diameter and placed in a freezer at -80 ℃
within a week of collection. Samples were stored for 6 months to 1
year, until sequential density fractionation.
Sequential Density Fractionation
Each mineral soil sample was separated into three distinct fractions
based on density and occlusion: the “free” light fraction (hereafter
fPOM) and the occluded light fraction (oPOM), with densities < 1.7
g/mL, and the remaining heavy fraction, which included
mineral-associated organic matter (MAOM) with a density > 1.7 g/mL.
These operationally-defined soil fractions can be used to distinguish
soil organic matter that is relatively accessible to microbial
decomposers (fPOM), from that which is isolated by physical barriers
like soil aggregates (oPOM), and bound to dense soil minerals (MAOM;
<ulink url="https://paperpile.com/c/IRXefK/i42t">Sollins et al.
2006)</ulink>.
First, samples of mineral soil from each site were dried at 60 ℃ until
they reached a constant mass. Approximately 20 g of each dry sample
was then placed in a 175-mL centrifuge tube and submerged in a
solution of sodium polytungstate (SPT) with a density of 1.7 g/mL,
which was determined to be ideal for isolating mineral particles from
organic matter in forest soils from our study region (Hicks
Pries,
unpublished data). The fPOM was
isolated by aspirating all floating material after initial submersion
in SPT followed by one hour of centrifugation at 3600 rpm. To disrupt
aggregates and release physically occluded organic material, we then
placed each centrifuge tube on a shaker table for 5 minutes and
subjected the soil solution to two minutes of sonication at 20 kHz
with a benchtop sonifier (Branson Ultrasonics, Danbury, CT, USA). The
floating material (oPOM) was then aspirated, leaving only the densest
fraction containing minerals and their associated organic matter.
Aspirated materials in the light fractions were isolated by vacuum
filtration and rinsed three times with ultrapure water to wash away
residual SPT. The remaining dense material was rinsed to remove all
SPT by agitating the material in ultra-pure water followed by
centrifugation at 3500 rpm for 20 minutes. The supernatant was then
discarded, and this process was repeated for a total of three rinses.
All three soil fractions were then oven-dried at 60 ℃. All
fractionated mineral soil samples and bulk samples of soil from both
the mineral and organic horizons were analyzed for total C and N
concentrations by mass (Carlo-Erba Instruments, Wigan, UK).
Soil texture analysis
To characterize underlying soil properties that influence MAOM
formation, we measured soil sand, silt, and clay content from a subset
of 3 mineral horizon samples at each site using the hydrometer
method. Briefly, we calculated the percentage of soil mass as clay
(particles < 2 µm), silt (2.0 – 50 µm), and sand (50 - 2000 µm)
after suspending a homogenized sample of each soil in a 5% sodium
hexametaphosphate solution and recording the solution density at
specified time intervals
<ulink url="https://paperpile.com/c/IRXefK/n87r">(Ashworth et al.
2001)</ulink>.
Ashworth J, Keyes D, Kirk R, Lessard R (2001) Standard procedure in
the hydrometer method for particle size analysis. Commun Soil Sci
Plant Anal 32:633–642
Sollins P, Swanston C, Kleber M, et al (2006) Organic C and N
stabilization in a forest soil: Evidence from sequential density
fractionation. Soil Biol Biochem 38:3313–3324
