Research
Phosphorus cycling in deciduous forest soil differs between
stands dominated by ecto- and arbuscular mycorrhizal trees
Anna Rosling1, Meghan G. Midgley2, Tanya Cheeke2,3, Hector Urbina1, Petra Fransson3 and Richard P. Phillips2
1
Department of Evolutionary Biology, EBC, Uppsala University, Uppsala, Sweden; 2Biology Department, Indiana University, Bloomington, IN 47405, USA; 3Department of Forest Mycology
and Plant Pathology, Swedish University of Agricultural Sciences, Uppsala, Sweden
Summary
Author for correspondence:
Anna Rosling
Tel: +46 18 471 6444
Email: [email protected]
Received: 15 August 2015
Accepted: 16 September 2015
New Phytologist (2016) 209: 1184–1195
doi: 10.1111/nph.13720
Key words: deciduous forest,
ectomycorrhizal (ECM) fungi, mycorrhiza,
phosphorus (P) cycling, soil.
Although much is known about how trees and their associated microbes influence nitrogen
cycling in temperate forest soils, less is known about biotic controls over phosphorus (P)
cycling. Given that mycorrhizal fungi are instrumental for P acquisition and that the two dominant associations – arbuscular mycorrhizal (AM) and ectomycorrhizal (ECM) fungi – possess
different strategies for acquiring P, we hypothesized that P cycling would differ in stands
dominated by trees associated with AM vs ECM fungi.
We quantified soil solution P, microbial biomass P, and sequentially extracted inorganic and
organic P pools from May to November in plots dominated by trees forming either AM or
ECM associations in south-central Indiana, USA.
Overall, fungal communities in AM and ECM plots were functionally different and soils
exhibited fundamental differences in P cycling. Organic forms of P were more available in
ECM plots than in AM plots. Yet inorganic P decreased and organic P accumulated over the
growing season in both ECM and AM plots, resulting in increasingly P-limited microbial
biomass.
Collectively, our results suggest that P cycling in hardwood forests is strongly influenced by
biotic processes in soil and that these are driven by plant-associated fungal communities.
Introduction
Soil age and climate play a central role in determining the extent
to which nitrogen (N) and phosphorus (P) limit biotic processes
in forests (Elser et al., 2007; Vitousek et al., 2010; Harpole et al.,
2011). In pedogenically ‘young’ soils (e.g. those in temperate and
boreal regions), N generally limits forest productivity (Finzi,
2009; Weand et al., 2010; Groffman & Fisk, 2011), whereas P
generally limits productivity in highly weathered tropical soils
(Walker & Syers, 1976; Turner et al., 2007) due to the formation
of recalcitrant inorganic and organic P forms, erosion and leaching losses (Walker & Syers, 1976; Turner et al., 2007). However,
in unglaciated temperate forest soils, such as those found in the
Midwestern USA, plant and microbial functions may be Plimited. For example, P addition decreased P-releasing enzymatic
activity in southern Ohio (DeForest et al., 2012). In addition,
this region has high rates of N deposition, which may further
exaggerate P-limitation (Jonard et al., 2015). Given the vital role
of temperate forests in the land carbon (C) sink (Xiao et al.,
2011) and the emerging view that both N and P constrain this
sink globally (Cleveland et al., 2013), an improved understanding
of P dynamics in temperate forests is warranted.
Phosphorus cycling in soil is known to be under both biological and geochemical control, with biological processes generally
found to dominate in topsoil where root and microbial density is
1184 New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
high (Wood et al., 1984; Achat et al., 2012). Phosphorus is available for uptake by plants and microbes only in its ionic forms
(Plassard & Dell, 2010), and the low mobility of P ions in soil
results in sharp depletion zones around roots (Lewis & Quirk,
1967). Forest ecosystems can remain productive despite chronically low soil P availability, because microbes can enrich this pool
by decomposing P-rich organic matter and weathering P-rich
minerals (Johnson et al., 2003; Zhao et al., 2009), Hence, the soil
microbial biomass, which can account for up to 78% of the total
biomass P in temperate forest soils (Turner et al., 2013), may be
the primary driver of P dynamics in these soils (Richardson &
Simpson, 2011).
The majority of plants form symbiotic associations with mycorrhizal fungi to acquire P (Plassard & Dell, 2010; Jansa et al.,
2011). Increased C allocation to roots and their mycorrhizal partners is a way for plants to alleviate P deficiency (Wallander &
Nylund, 1992; van der Heijden, 2001; Kiers et al., 2011). The
two types of mycorrhizal fungi that associate with trees, ectomycorrhizal (ECM) fungi and arbuscular mycorrhizal (AM) fungi,
both significantly increase P acquisition to plants, primarily by
increasing the surface area for P uptake by colonizing soil beyond
the root depletion zone (Jones et al., 1998; Smith & Read, 2008;
Smith et al., 2011). ECM and AM fungi differ in what substrates
from which they can access P. ECM fungi produce enzymes to
hydrolyze organic P forms in soil (H€aussling & Marschner,
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
1989; Read & Perez-Moreno, 2003), whereas AM fungi produce
these enzymes to a much lesser extent (Joner et al., 2000; Koide
& Kabir, 2000). ECM fungi also produce low molecular weight
organic acids and chelators (Plassard & Dell, 2010) that have
been thought to play a major role in the weathering of minerals
(Hoffland et al., 2004). Interestingly, such organic weathering
agents are now proposed to play a central role in the destabilization and subsequent decomposition of organic matter in soil
(Clarholm et al., 2015). Although there seem to be less substantial differences between AM- and ECM-dominated forests in
mineral weathering than previously assumed (Koele et al., 2014),
the extent to which soils dominated by ECM- or AM-forming
trees vary in their P cycling characteristics is still unknown.
In this study, we compared forest plots dominated either by
AM- or ECM-forming trees, but underlain by the same parent
materials, to assess the degree to which mycorrhizal association
influences P cycling. In a previous investigation, we reported that
the upper surface soils of AM-dominated plots have higher pH,
lower organic matter content and lower acidphosphatase activity
than ECM-dominated plots at this site (Phillips et al., 2013).
Hence, we hypothesized that P would be cycled mostly in inorganic forms in soils dominated by AM trees, but in organic forms
in soils dominated by ECM trees. Further, because these forests
are underlain by ‘unglaciated’ soils that have low P availability
(DeForest et al., 2012), we hypothesized that biotic processes
would drive seasonal P dynamics in this forest.
Materials and Methods
Research 1185
Field sampling
In each plot, four litter traps (40 cm diameter, 1-mm mesh) were
placed at the corners of an internal 12 9 12 m square. Trap content was collected biweekly from May to November 2011. In the
laboratory, litter was air-dried for 4 d at room temperature
(21°C) and pooled by plot. For each plot, leaf litter was sorted by
species and weighed.
Soil was collected from each plot during the first week of May,
July, September and November in 2012. In each plot, four soil
cores were collected following a stratified random sampling
design by taking one core in each of four quadrants of the internal 15 9 15 m area. We sampled the top 15 cm of the soil (including the organic layer) with a 5-cm-diameter stainless steel
corer. Cores were transported to the lab in their plastic liners and
processed within 24 h of sampling by separating cores into the
top 0–5 cm and bottom 5–15 cm, and pooling cores from a given
plot. Soils were sieved using a 2-mm mesh to homogenize the soil
and remove rocks and roots. Five subsamples of each soil sample
were immediately stored at 4, 20, 80°C, freeze- or air-dried
until analysis (see later).
Soil solution P concentration data were collected at 13 time
points from 1 May until mid-November 2012. The anion
exchange membrane (AEM) method developed by Shaw &
DeForest (2013) was used to estimate in situ soil solution P concentrations (orthophosphate ion) throughout the growing season,
detailed in Method S1. Soil solution P was calculated as average
lg P per strip.
Site description
Aboveground characteristics
This research was conducted at Indiana University’s Moores
Creek Research and Teaching Preserve, a c. 80-yr-old temperate
deciduous forest in south-central Indiana, USA (39°050 N,
86°270 W) with elevation ranging from 165 to 230 m asl. Soils
are thin, unglaciated Inceptisols derived from siltstone, shale and,
to a lesser extent, limestone. The forest naturally regenerated following the abandonment of farming c. 1930. The climate is
humid continental with mean annual precipitation of 1200 mm
and mean annual temperature of 11.6°C.
In 2011, 14 plots (20 9 20 m) were identified to represent
stands dominated by trees associated with either AM or ECM
fungi. AM plots contained a mixture of sugar maple (Acer
saccharum), tulip poplar (Liriodendron tulipifera) and sassafras
(Sassafras albidum). ECM plots contained pignut hickory
(Carya glabra), American beech (Fagus grandifolia), white oak
(Quercus alba), red oak (Q. rubra) and black oak (Q. velutina).
Plots were assigned to a mycorrhizal association when trees
from that mycorrhizal association (AM or ECM) contributed
to > 80% of the basal area of the plot (n = 7). All plots were
located in similar landscape positions and contained more than
one tree species from the dominant mycorrhizal association.
Slope position and aspect were not correlated with dominant
mycorrhizal association (Supporting Information Fig. S1). All
sampling was performed in a 15 9 15 m internal plot to avoid
edge effects.
In order to estimate the annual litter P pool, we measured the P
content of the three dominant AM and ECM leaf litters (AM:
A. saccharum, L. tulipifera and S. albidum; ECM: C. glabra,
Q. alba and Q. rubra). For each species, two to three fallen leaves
were randomly selected from three plots where those trees were
most abundant. Leaves were ground using a mortar and pestle,
and 0.2 g was digested in nitric and perchoric acid (Sommers &
Nelson, 1972). Because there were no significant differences
among tree species within a mycorrhizal association, we used an
average leaf litter P for AM-forming tree species
(0.49 0.06 mg P g1)
and
ECM-forming
trees
(0.45 0.03 mg P g1) multiplied by the weight of each litter
type in each plot to estimate litter P in mg m2. Further, an estimate of annual tree P requirements per plot was derived from the
litter P content based on available data from temperate forests
(Johnson et al., 2003). Johnson et al. (2003) estimate that P in litter accounts for c. 60% of the annual P requirement of trees, the
other parts being wood increment and root growth. Hence, plot
average litter P m2 was divided by 0.6 to estimate annual tree
P requirement m2.
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Soil characteristics
A field-moist sample of 1–2 g of soil was weighed and dried at
105°C for 1 h for gravimetric soil moisture (SM) determination.
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
New
Phytologist
1186 Research
Organic matter (OM) content was determined by ashing dry soils
at 450°C for 16 h. Weight loss after ashing was used to calculate
the percentage OM. Air-dried soil samples were ground, and
total C and N concentrations were determined by dry combustion on a Costech ECS4010 Elemental Analyzer (Costech Analytical, Valencia, CA, USA). Soil pH was determined on 5 g of
DW-equivalent soil suspended in 40 ml of 0.01 M CaCl2 solution. The suspensions were shaken for 1 h and vortexed immediately before analysis. Soil pH was measured using a bench top
electrode Ag/AgCl pH meter (VWR, Radnor, PA, USA).
Sequential P extraction
Immediately following sample preparation, 5 g of field-moist soil
was subjected to a partial Hedley sequential P fractionation
(Hedley et al., 1982a,b). We extracted and quantified inorganic P
(Pi) in the following four fractions: resin P, bicarbonate P,
hydroxide P and acid P (Cross & Schlesinger, 1995). As bicarbonate P and hydroxide P also contain organic P (Po), we also
quantified Po in these fractions. Resin P and bicarbonate P are
considered to extract more bioavailable P pools than the more
complex P forms extracted by hydroxide P (secondary P minerals) and acid P (calcium bound P minerals) (Cross & Schlesinger,
1995; Yang & Post, 2011). We did not perform complete digests
of the soil samples and hence have no estimate of residual P.
Thus, the sum of the extracted P fractions – that is, total Pi plus
total Po – does not reflect total soil P in these soils. Details on
the sequential P extraction procedure are given in Methods S2.
Enzyme activities
The potential extracellular enzyme activities of acid phosphomonoesterase (AP) and phosphodiesterase (PD) were analyzed using
previously published methods (Saiya-Cork et al., 2002), detailed
in the Methods S3. Measured acid phosphatase activity provides
a good estimate of extracellular enzymatic activity in soil because
it is known to be unaffected by cell lysis (Blankinship et al.,
2014) and intracellular phosphatases mostly have an alkaline pH
optima (Plassard et al., 2011). Because freezing is expected to
cause the least alteration of soil enzymatic activity during storage
(Burns et al., 2013), soils were stored at 80°C before analysis
(< 1 month).
Soil microbial biomass
Soil microbial biomass was extracted from soils stored at 20°C
by fumigating c. 10 g of soil for 4 d in a desiccator with an
ethanol-free chloroform atmosphere (Brookes et al., 1982).
Organic C was extracted from 10 g of fumigated and unfumigated soils with 50 ml of 0.5 M K2SO4 (Vance et al., 1987).
Extracts were stored at 20°C until analyzed. Total organic C in
extracts was quantified via high temperature oxidation on a Shimadzu TOC analyzer. Microbial biomass C (MBC) was calculated as the difference in extracted C before and after fumigation
and corrected assuming an extraction efficiency of 45% (Jenkinson et al., 2004). Total ergosterol in soil was extracted from
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
freeze-dried soils according to Methods S4 and used as an indicator of biomass of Ascomycete and Basidiomycete fungi in soil
(Olsson et al., 2003).
Similarly to fresh soils, resin P and bicarbonate P were
extracted from 5 g of fumigated soil according to the procedure
described for sequential P extraction, detailed in Methods S2.
Released microbial biomass P (rMBP) was calculated as the difference in extracted Pi before and after fumigation. We corrected
microbial biomass P (MBP) to account for differences in P sorption among plots: MBP = rMBP/(1 sorption) (Brookes et al.,
1982). Phosphorus sorption capacity was determined for soil
samples collected in May 2012 and assumed to be constant over
time. From each sample, subsamples of 1.2–1.3 g of field-moist
soil were equilibrated in 25 ml 1 mM KH2PO4 (+P) or 25 ml
ddH2O (noP) by horizontal shaking for 16 h (Giesler et al.,
2002). Soil slurries were filtered through prerinsed Whatman
No. 1 filter papers and extracts were stored at 20°C until analyzed. Sorption was calculated as the percentage of added P
remaining in +P solutions minus P released in the noP solutions
(Giesler et al., 2002). We calculated the molar ratio of C : P in
the microbial biomass (MBC : P) after correcting MBP by assuming an extraction efficiency of 40% (Brookes et al., 1984). However, to avoid overestimating the size of the MBP pool, we did
not otherwise correct for extraction efficiency.
Molecular characterization of soil fungal community
Four plots were randomly selected from each mycorrhizal association and total soil DNA was extracted from freeze-dried soil samples (0–5 cm) from May and September. Extractions were
performed in duplicate from 0.25 g of soil using the XpeditionTM
Soil/Fecal DNA miniprep kit (Zymo Research, Irvine, CA,
USA). The fungal barcode region ITS2 rRNA (Schoch et al.,
2012) was amplified using gITS7 (Ihrmark et al., 2012) and
ITS4 m modified from ITS4 to account for mismatches against
the
class
Archaeorhizomycetes
(50 -TCCTC[C/G][G/C]
0
CTTATTGATATGC-3 ). Life technology fusion primers were
constructed on the forward primer for multiplexing. PCR was
performed with 10–20 ng DNA template, 1 9 SSoAdvancedTM
Universal SYBR® Green Supermix (Bio-Rad) and 0.8 nM of each
primer on a CFR96 TouchTM Real/Time PCR Detection system
(Bio-Rad). Following 10 min of pre-denaturation at 95°C, we
ran 27 cycles of 1 min at 95°C, 45 s at 50, 54 and 58°C (performed in parallel reactions), and 50 s at 72°C, followed by
3 min at 72°C. Three annealing temperatures were used to
reduce primer bias (Schmidt et al., 2013) and monitoring the
runs by real-time PCR ensured that amplification remained linear. All six reactions from each soil sample were combined and
purified using the ZR-96 DNA Clean & ConcentratorTM-5
(Zymo Research). The fungal community sequence analysis is
detailed in Methods S5.
Calculations and statistical analyses
Principal components analysis (PCA) of soil parameters was conducted in PC-ORD version 5.33 software (McCune & Mefford,
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
2006). Before PCA analysis, we used correlations (Pearson’s r) to
evaluate which variables covaried (MINITAB 16.0; Minitab Inc.,
State College, PA, USA) in order to subsequently remove those
from the analysis. The enzyme activities AP and PD showed a
significant positive correlation, as did Pi content in different fractions (P < 0.001). We also performed a number of PCA runs,
including selected individual P fractions and enzyme activities in
different combinations, to ensure the robustness of the analyses.
Here, we present the results based on total Pi, total Po, P sorption, MBC : P, enzyme activity for AP, ergosterol, SM, OM content and pH; all other combinations were very similar in
outcome. Ordination results are presented in joint plots. Correlations between soil variables and ordination axes using PC-ORD
default settings are reported as r2 values and used as indicators for
the importance of variables in structuring the ordination.
We used the multi-response permutation procedure (MRPP),
a nonparametric procedure in PC-ORD for testing the hypothesis of no difference between two or more groups (McCune &
Grace, 2002), in order to test for the effects of mycorrhizal association, soil layer and time. MRPP provides P-values as well as Avalues that measure ‘effect sizes,’ representing homogeneity
within the group compared with that expected randomly. For
instance, perfect homogeneity in the group gives A = 1, whereas
A values between 0 and 1 indicate that heterogeneity between the
groups is greater than that expected by chance.
Mixed linear model analyses were conducted including either
one or two factors depending on the tested parameter (see later),
followed by pairwise comparisons using Student’s t-test or
Tukey’s t-test, respectively. These statistical analyses were performed in JMP 10 (SAS Institute Inc., Cary, NC, USA). Basal
area, litter biomass, leaf litter P content and average seasonal
in situ soil solution P were analyzed using mycorrhizal association
as a fixed effect. Leaf litter P content was analyzed with tree
species as a fixed effect. Soil moisture, OM content, C : N ratio,
pH, P sorption, total ergosterol, MBC, MBP, MBC : P, and Pi
and Po content in all extracted fractions were analyzed using
mycorrhizal association, soil layer and their interaction as fixed
effects, and plot as a random effect. The same analysis was conducted for P content of all fractions and enzymatic activity normalized by ergosterol. In situ soil solution P from all 13 sampling
times between 1 May and 15 November, SM and P content in all
extracted fractions in the top 0–5 cm soil at the four sampling
times were analyzed using mycorrhizal association, time and their
interaction as fixed effects, and plot as a random effect. Regression analysis was used to identify correlations amongst measured
soil characteristics. Distributions of parameters were checked for
normality and transformed to normal distributions when needed
using log (total Po, MBP, MBC, soil solution P per strip, AP,
PD and total enzyme activity) and square root (resin P, bicarbonate P, bicarbonate Po, hydroxide P, hydroxide Po, acid P, ergosterol and OM) transformations.
In order to relate aboveground P pool sizes (litter P and estimated annual P requirement of trees; see ‘Aboveground
characteristics’) to belowground P fractions, P content in
extracted soil fractions was converted to pool sizes for the top
15 cm soil (kg ha1). Soil weight per surface area was
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Research 1187
approximated using measurements from 33 plots in three adjacent forest sites: 51.7 15.5 and 133.6 17.2 kg m2 for the 0–
5 and 5–15 cm soil layers, respectively. No correlation was
detected between soil weight and proportion of ECM forming
trees in the plots. The P contents of extracted fractions were multiplied by the average soil weight of the soil layer and an estimate
of each P pool down to 15 cm depth was produced by addition
of values for 0–5 cm and 5–15 cm for each plot. Only plots that
had measurements from both soil layers were used in further
analysis. Mixed linear models with mycorrhizal association, sampling time and their interaction as fixed effects, and plot as a random effect, were used to analyze the estimated P pool sizes.
Results
Plot level effects of mycorrhizal association
Based on abiotic soil characteristics, total extracted organic and
inorganic P content, MBC : P, ergosterol and acid phosphatase
activity, AM plots separated from ECM plots in multivariate
space (Fig. 1). The first three PCA axes accounted for 70% of the
variation. MRPP analyses revealed significant differences between
mycorrhizal associations (A = 0.017, P = 0.007), soil layers
(A = 0.066, P < 0.001) and over time (A = 0.21, P < 0.001). The
separation was mainly driven by differences in ergosterol (axis 1;
r2 = 0.83), OM content (axis 1; r2 = 0.77) and enzyme activity for AP (axis 1; r2 = 0.71), and partly driven by pH (axis 2;
r2 = 0.51), P sorption (axis 2; r2 = 0.45) and MBC : P (axis 2;
r2 = 0.44) (Fig. 1). Mycorrhizal association separated mainly
along ordination axis 1, whereas soil layers separated mainly
along axis 2 (Fig. S2). The layer differences were greater in ECM
plots than in AM plots (Fig. S2).
Mycorrhizal association significantly affected most soil characteristics (Table 1) and extracted P fractions (Table S1). Percentage OM (P < 0.001), soil C : N ratio (P < 0.001) and P sorption
(P < 0.001) were significantly higher in ECM plots compared
with AM plots, whereas pH was significantly lower in ECM compared with AM plots (P < 0.001). Microbial biomass C
(P = 0.008) and total ergosterol (P < 0.001) were greater in ECM
plots compared with AM plots, but the opposite was true for
MBP which was highest in AM plots (P = 0.034). Soil moisture
was not affected by mycorrhizal association (P = 0.543). Many
variables also exhibited interactive effects between mycorrhizal
association and soil layer (Tables 1, S1). For example, the molar
ratio C : P in microbial biomass was significantly higher in ECM
plots as a result of high values in the deeper soil layer. Soil moisture, OM content, C : N ratio, MBC and total ergosterol were
greater in the 0–5 cm layer than in the 5–15 cm layer (all
P < 0.001). Soil layer also affected all extracted P fractions with
significantly less P in the 5–15 cm soil layer compared with the
top 5 cm of soil. Only pH was not significantly affected by soil
layer (P = 0.256), but there was a significant interaction between
mycorrhizal association and soil layer for pH (P = 0.001).
In contrast to belowground characteristics, AM and ECM
plots were similar in aboveground characteristics, both with
respect to basal area (P = 0.271) and leaf litter biomass
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
New
Phytologist
1188 Research
5
P sorption
4
MBC:P
3
2
Tot.Po
AP
–12
–10
–8
SM
–4
–6
0
–2
2
Ergo.
–1
OM
–2
Total Pi
–3
ECM
AM
pH
–4
–6
PCA axis 2 (21%)
(P = 0.102) (Table 1). Litter from both AM- and ECM-forming
tree species was found in all plots, but plot-specific litter dominated. No significant difference in P content of the litter was
detected among the six tree species (P = 0.681) or between the
two categories AM or ECM (P = 0.563). On average,
632 42 mg P m2 was delivered to the forest floor by litter.
AM plots received marginally more P from litter compared with
ECM plots (P = 0.061) (Table 1).
Effects of mycorrhizal association on soil enzymatic activity
and microbial communities
Soil phosphatase activities were significantly affected by mycorrhizal association and soil layer with higher activity in ECM plots
compared with AM plots (P < 0.001 for both): 1.5 times higher
for AP and 2.3 times higher for DP (Table 1). Total soil phosphatase activity (i.e. the sum of AP and PD) was more strongly
correlated with total extracted ergosterol (r2 = 0.61, Fig. S3a)
than with MBC (Fig. S3b). In ECM plots, there was a weak correlation between total phosphatase activity and MBC (r2 = 0.43),
whereas no correlation was found in AM plots (r2 = 0.04)
(Fig. S3b). The correlation between total phosphatase and ergosterol was conserved within ECM plots but no correlation
between phosphatase activity and ergosterol was found for AM
plots (r2 = 0.08) (data not shown). There was a strong positive
correlation between ergosterol and MBC in ECM plots
(r2 = 0.86) indicating that fungi in the Dikarya constitute a major
fraction of microbial biomass in these plots (Fig. S4). The correlation between ergosterol and MBC was weaker in AM plots
(r2 = 0.54) (Fig. S4). Despite these patterns, P pool content and
enzymatic activity remained significantly different between plots
dominated by different mycorrhizal associations after
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
4
6
PCA axis 1
(36%)
Fig. 1 Principal components analysis (PCA)
ordination plot of total organic and inorganic
soil phosphorus (P) pools (total Po and total
Pi, respectively), P sorption, molar ratio of
C : P in the microbial biomass (MBC : P),
enzyme activity of acid
phosphomonoesterase (AP), ergosterol
(Ergo.), soil moisture (SM), organic matter
content (OM) and pH, showing the
difference in soil chemistry between
arbuscular mycorrhizal (AM)- and
ectomycorrhizal (ECM)-dominated plots.
Data from both 0–5 and 5–15 cm soil layers
are included and seasonal averages of these
variables are also presented in Table 1 and
Supporting Information Table S1. Closed and
open circles represent AM and ECM
mycorrhizal association, respectively. The
multi-response permutation procedure
(MRPP) showed significant (P = 0.0072)
separation between AM and ECM plots. The
vector length indicates the relative
importance of the variables.
normalization by the amount of fungi in the Dikarya, estimated
by extracted soil ergosterol (Table S2). Thus, observed differences
in the P content of extracted soil fractions are not driven only by
the biomass of fungi in the Dikarya.
Arbuscular and ectomycorrhizal plots harbored distinct soil
fungal communities in topsoil. Mycorrhizal association explained
37% of the observed variance in community structure, whereas
sampling time explained 10% (Fig. S5). Fungal richness was significant higher (Adonis, r2 = 0.34, P < 0.001) in the AM plots
compared with the ECM plots (Fig. 2a). The characterized fungal
communities were dominated by taxa belonging to the Dikarya
in both AM and ECM plots, and AM fungal communities in
these soils thus remain unknown. Ascomycetes were by far the
most abundant phyla, constituting c. 80% of the rarefied fungal
sequence reads, except for ECM plots sampled in May at which
time Ascomycetes contributed to only 50% of the captured fungal community (Fig. 2b). Instead ECM plots sampled in May
were dominated by fungal taxa that have a known ECM lifestyle
and these constituted 40% of the observed fungal abundance of
these samples (Fig. 2c). The relative abundance of ECM fungi
was much lower in September than in May but still higher in
ECM plots compared with AM plots (Table S3).
Effects of mycorrhizal association on P pool sizes
Above- and belowground P pools were scaled to kg P ha1
(Table S4) to allow for a comparison of average annual pool sizes
(Fig. 3). This comparison indicated that annual P requirements
of the trees correspond to approximately one-third of the annual
average P in soil microbial biomass. Furthermore, the sum of seasonal average Hedley available pools (resin P, bicarbonate Pi and
Po) exceeds the annual P requirement of the trees biomass by two
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
Research 1189
Table 1 Average values SE, expressed as percentage or per kg of DW
soil for arbuscular mycorrhizal (AM)- and ectomycorrhizal (ECM)dominated plots
1400
(a)
1200
1000
Aboveground
characteristics
AM
ECM
Basal area (m2 ha1)
Leaf litter (g m2)
Total litter P (mg m2)
126 16a
1509 150a
716 70a
105 9a
1224 74a
560 36a
Soil characteristics
SM %L
OM %M,L,MxL
C : NM,L
pHM, MxL
P sorption %M
Microbial biomass
Tot ergo
mg kg1 M,L,MxL
MBC g kg1
soilM,L
MBC : P M,L,MxL
Enzyme activities
AP
activityM,L,MxL
PD
activityM,L,MxL
800
600
400
200
0
0–5 cm
5–15 cm
0–5 cm
5–15 cm
28 2ab
8.0 0.3b
13 0.3c
5.0 0.1a
24 3b
23 1bc
4.6 0.1c
11 0.4d
4.6 0.1b
29 4bc
30 2a
15.2 1.2a
21 0.9a
4.0 0.1c
38 1ab
20 1c
4.4 0.2c
17 0.6b
4.2 0.1c
42 3a
1.5 0.2b
0.5 0.1d
6.3 0.6a
0.9 0.1c
40%
1.3 0.11b
0.5 0.06c
2.0 0.18a
0.6 0.05c
20%
22 3b
35 9b
24 4b
63 12a
0%
Observed
100%
29 1.7
25 1.5
57 3
33 2
4.6 0.8b
2.1 0.1c
14 1.6a
2.9 0.3bc
bc
c
a
(b)
80%
60%
May
September
b
Microbial biomass C : P is expressed as a molar ratio (MBC : P). Acid
phosphatase activity (AP) and phosphodiesterase (PD) are expressed in
arbitrary activity units based on fluorescence in soil enzyme assay. Soil
characteristics are further separated by soil depth. SM, soil moisture; OM,
organic matter; Tot ergo, total ergosterol. Different letters indicate that
values are significantly different based on Tukey t-test. MMain effect of
mycorrhizal association. LMain effect of soil layer. M9LInteraction effect for
mycorrhizal association 9 soil layer. For each P pool, different letters indicate that values are significantly different based on Student’s t-test (plot
characteristics) and Tukey t-test (soil characteristics).
to five times (Fig. 3). AM plots had significantly more extractable
P
(229 12 kg ha1)
compared
with
ECM
plots
(187 16 kg ha1) (P = 0.019). The difference was due to significantly more total extractable Po in the AM plots (P = 0.027).
However, the ECM plots had significantly more available Po
compared with AM plots (P < 0.001). The opposite was true for
the less available complex P fractions, where AM plots had significantly more Po compared with ECM plots (P < 0.001). Overall,
Po made up 73 3% of the available P in ECM plots compared
with 65 4% in AM plots, whereas Po made up 63 3% of the
less-available P in ECM plots compared with 79 1% in AM
plots.
Seasonal effects on soil P dynamics
In both AM and ECM plots, there was a significant effect of sampling time on total Pi (P < 0.001) (Fig. S6a) and total Po
(P < 0.001) (Fig. S6b), as well as all individual P fractions, except
bicarbonate Po (Table S4). On the one hand, we observed a
steady decrease in total extractable Pi over the season from an
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Chao1
50%
(c)
40%
30%
20%
10%
0%
May
September
Fig. 2 Top-soil fungal communities in arbuscular mycorrhizal (AM; dark
grey) vs ectomycorrhizal (ECM; light grey) plots sampled in May and
September. (a) Observed and estimated (Chao 1) fungal richness. (b)
Average relative abundance of Ascomycota as percentage of total fungal
community. (c) Average relative abundance of ECM forming fungi as
percentage of total fungal community. Data are based on rarefied
sequence reads and presented as average SE.
average of 56 4 mg P kg1 soil in May to 24 2 mg P kg1
soil in November (Fig. 4a). Organic P, on the other hand,
increased sharply between July and September, from an early season average of 59 5 mg P kg1 soil to 126 11 mg P kg1 soil
in September (Fig. 4a). Total Po was also significantly affected by
mycorrhizal association (P = 0.027), but there was no significant
interaction effect between mycorrhizal association and sampling
time (Fig. S6b).
Over the growing season, there was a significant increase in
MBC : P ratios from an average of 25 in May, a ratio c. 50 in July
and September, and a final high of 140 recorded in November.
MBC : P was also higher in ECM plots compared with AM plots
(P = 0.066) (Fig. S6c). Changes in MBC : P were driven both by
increases in MBC and decreases in MBP (Fig. 4b). Microbial
biomass C was significantly affected by mycorrhizal association
(P = 0.003), season (P < 0.001), and their interaction
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
New
Phytologist
1190 Research
9
12
(a) ECM dominated plots
(b) AM domianted plots
Litter P
6 ± 0.4
Litter P
7 ± 0.7
Complex Po
3 ± 0.5
Microbial
Biomass P
Solut.
3 ± 0.4
Complex Po
Microbial
Biomass P
Solut.
25 ± 3
29 ± 4
4 ± 0.4
4 ± 0.4
Exchang.
Pi
Exchang.
Pi
Avail. Po
Avail. Po
26 ± 2
14 ± 2
Complex Pi
Complex Pi
52 ± 4
108 ± 15
56 ± 6
159 ± 12
Fig. 3 Growth seasonal average phosphorus (P) pools scaled to illustrate relative pools sizes in kg P ha1 down to 15 cm soil depth. Complex and available
organic soil P (Po) pools are grey to illustrate that these were significantly different between (a) ectomycorrhizal (ECM)- and (b) arbuscular mycorrhizal
(AM)-dominated stands. Biotic pools (annual tree requirement and microbial biomass) are illustrated as circles. Yearly tree P requirements are estimated,
derived from the litter P. Uptake to trees is assumed to be predominantly through the microbial biomass P pool. Resin P pool was closely correlated to soil
solution P and is thus labeled solute (Solut.). Exchangeable (Exchang.) inorganic soil P (Pi) and available (Avail.) Po represent bicarbonate extracted pools.
For less available P forms, (Complex) Po represents the hydroxide extractable fraction and Complex Pi represents the sum of hydroxide and acid
extractable Pi. Arrows illustrate the proposed main fluxes in the two stand types over the growth season. Values, except for tree P requirements, are the
seasonal averages SE.
140
a
(a)
a
mg P kg–1 soil
120
100
80
b
b
60
40
a
a
20
b
b
00
May
a
Sept
Nov
a
70
MBC mg kg–1 soil
1200
60
1000
a
b
a
800
40
c
600
400
200
50
30
20
b
b
10
0
0
May
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
July
Sept
Nov
MBP mg kg–1 soil
1400 (b)
July
Fig. 4 Seasonal dynamics of (a) total
inorganic phosphorus (P) fractions (Pi; light
grey) and total organic P (Po; dark grey) and
(b) corrected microbial biomass carbon
(MBC) (black) and P (MBP; grey). Values are
weighted average across soil layer with
SEM as error bars. Different letters (above
Po and MBP, and below Pi and MBC)
indicate significantly different values
between sampling times within each pool.
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
Research 1191
Discussion
We examined P dynamics in soils in central hardwood forests of
southern Indiana where a rich assemblage of tree species forming
symbiotic associations with predominantly arbuscular (AM) or
ectomycorrhizal (ECM) fungi co-occur on soils developed from
the same unglaciated parent material. Accumulating evidence
suggests that the mycorrhizal associations of dominant tree
*
(a)
100
Soil solution P (µg per strip)
July (c. 16 mm on 7–8 July). The peak in ECM plots was > 10
times higher than the peak in AM plots: 89 24.4 compared
with 8.3 2.9 mg P m2. Among all measured P fractions in the
top 5 cm, resin P, MBP and hydroxide Po (Fig. 5b–d) were the
only P fractions significantly affected by mycorrhizal association
and time; significant differences between AM and ECM plots
were observed in all of these in July, the time of lowest soil moisture. Finally, there was no correlation between soil pH and resin
P (r2 = 0.006) across the season.
35
80
30
60
25
40
*
20
15
10
0
1-May 15-May 1-Jun 15-Jun 1-Jul
12
Resin Pi (mg P kg–1 soil)
20
% Soil moisture
(P = 0.017); MBC was higher in ECM plots compared with AM
plots in July and September (Fig. S6d). For both mycorrhizal
associations, MBC fluctuated over the season with high values
recorded in May and November (Fig. 4b). Microbial biomass P,
however, decreased over the season (Fig. 4b) and was not significantly affected by mycorrhizal association (P = 0.973, Fig. S6e).
The summer of 2012 was dry with limited rainfall from midMay to mid-August. As a result, soil moisture changed significantly over the growing season with the lowest values (14%)
recorded in July (Fig. 5a). In situ soil solution P was significantly
affected by mycorrhizal association (P < 0.001) and time
(P < 0.001), and a significant interaction effect was detected
(P = 0.001; Fig. 5a). Seasonal average in situ soil solution P was
three times higher in ECM plots at 18 3.2 lg P per strip compared with AM plots at 6.2 0.9 lg P per strip. This difference
was the result of an extended peak with significantly higher soil
solution P detected in ECM plots compared with AM plots following the first rain, which began to alleviate the drought in early
(b)
15-Jul 1-Aug 15-Aug 1-Sep 15-Sep 1-Oct 15-Oct 1-Nov 15-Nov
a
10
8
b
6
b
b
4
bc
2
cd
d
d
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Hydroxide Po (mg P kg–1 soil)
Fig. 5 Seasonal dynamics of phosphorus (P)
fractions with significant differences between
arbuscular mycorrhizal (AM) stands (dark
grey) and ectomycorrhizal (ECM) stands
(light grey): (a) illustrates in situ soil solution
P (lg/ strip) collected in 2-wk intervals and
percentage soil moisture (dashed line); (b)
readily available inorganic soil P (Pi)
measured as resin Pi; (c) microbial biomass P
(MBP) and (d) less available organic soil P
(Po) measured as hydroxide(-extractable) Po.
Average values are presented SE. In (a),
asterisks indicate time points when AM and
ECM stands are significantly different; in (b–
d) different letters indicate significant
differences based on Tukey’s pairwise
comparison.
Corrected MBP (mg P kg–1 soil)
0
(c)
80
60
a
ab
abc
abc
bcd
40
bcd
cd
d
20
0
150
(d)
a
ab
ab
120
abc
90
bc
c
60
c
d
30
0
May
July
Sept
Nov
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
1192 Research
species have distinct effects on ecosystem-level C and N cycling
(Phillips et al., 2013). For instance, differences in the enzymatic
capacities of AM and ECM fungi have been suggested to be the
best predictor of decomposition and C storage in soil at both
ecosystem and global scales (Averill et al., 2014). Our study provides evidence that these associations also result in plot-level differences in P dynamics, and that microbial biomass and
community compositions drives these dynamics.
In accordance with the recently proposed framework of distinct mycorrhizal-associated nutrient economies (Phillips et al.,
2013), we hypothesized that in plots where trees associate with
AM fungi, soil P would cycle predominantly in inorganic forms,
but that in plots dominated by ECM forming trees, soil P would
cycle mostly in organic forms of P. Overall, our findings confirm
these expectations. Significantly higher soil phosphatase activity
and higher Po availability in ECM plots support the hypothesis
that these systems cycle organic P substrates to a greater extent
than AM plots. In AM plots, organic P was instead found to
accumulate in the less-available hydroxide-extractable pool. The
sum of all inorganic P pools, however, was not significantly
affected by mycorrhizal association.
Our results are in line with the emerging view that P cycling in
forest soil systems is controlled by the soil microbial biomass
rather than by the annual uptake of trees (Cole et al., 1978; Yang
& Post, 2011). As in earlier reports from sites presumed to be Plimited, we found that Hedley available pools exceed tree requirements for P (Johnson et al., 2003; Yang & Post, 2011). This is,
in part, because resorption and internal storage of P within
woody plants decrease their dependency on de novo P acquisition
(Rennenberg & Herschbach, 2013). Furthermore, in our study
the seasonal average microbial biomass P (MBP) was about three
times bigger than the estimated P requirements of the trees. Similarly, microbial biomass has been shown to constitute 68–78% of
the total biomass P in other temperate forests (Turner et al.,
2013). Overall, the sum of the two biotic pools – seasonal average
MBP and estimated annual P requirements of the canopy trees –
exceeded the size of the available P pool, indicating that the
microbial biomass drove the observed P dynamics. Seasonal averages provide a conservative estimate of available pool sizes
because these pools are expected to have a high turnover rate of
days to weeks (Shaw & DeForest, 2013). The trend was stronger
for AM plots and whether these are more P-limited than the
ECM plots due to the lower ability of their microbial biomass to
utilize organic P sources, or whether AM plots have a higher
turnover rate of MBP, warrant further study.
Higher concentrations of available Po in ECM plots than in
AM plots appear to be driven by higher enzymatic activities of
fungi in ECM plots due to differences in fungal community composition. High phosphatase activity of ECM fungi (Plassard &
Dell, 2010), the ability of ECM fungi to capture P directly from
saprotrophic fungi (Lindahl et al., 1999), and the recently proposed pathway of direct Po uptake by ECM fungi (Rennenberg
& Herschbach, 2013; Becquer et al., 2014) highlight the importance of Po for ECM fungi and their associated host trees. Across
all plots, total phosphatase activity was strongly correlated with
total ergosterol, but less so with MBC, supporting the idea that
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
New
Phytologist
fungi in the Dikarya are predominantly involved in P release
from soil organic matter. However, ergosterol alone did not
explain soil enzymatic activity or P content in measured P pools
because these remained significantly different when normalized
by ergosterol. The composition of fungal taxa and taxonomic
rank also differed between AM and ECM plots, and the capacities to exude extracellular phosphatases varies significantly among
species in the Dikarya, even among ECM fungi (Nygren &
Rosling, 2009). Finally, plant roots, bacteria and saprotrophic
fungi may all contribute to non-negligible extracellular phosphatase activity and cycling of organic P in AM plots. Thus, community compositional differences may explain observed
differences in phosphatase activities and Po availability between
AM and ECM plots.
In contrast to our expectations, we found no differences in
total Pi in AM and ECM plots and higher in situ soil solution Pi
in ECM plots than in AM plots. We had expected that the higher
pH of AM soils would result in higher Pi availability in AM plots
compared with ECM plots. However, the lack of correlation
between soil pH and the available Pi extracted in the Resin fraction across the season suggests that biological processes, rather
than chemical processes, control P availability in forest top soils
(Wood et al., 1984). Furthermore, higher in situ soil solution Pi
in ECM plots was driven by drought-alleviating rains in early
July that resulted in a soil solution P peak that was significantly
higher in ECM plots compared with AM plots. This peak may
be explained by high phosphatase activities in ECM plots that
hydrolyzed hydroxide Po and resulted in high resin Pi and bicarbonate Po pools in ECM plot soils. The accumulated resin Pi
pool was then detected as a soil solution P peak as soon as
drought conditions no longer prevented the mobility of hydrolysis products. Lysis of microbial biomass following the rain could
also have contributed to the soil solution P peak, as MBP was significantly higher in ECM plots compared with AM plots before
the rains. Low average soil solution P concentrations in AM
plots, particularity the low peaks following the droughtalleviating rains in July, and undetectable differences in total Pi
between AM and ECM plots strongly support the idea that the
AM pathway of P acquisition is based primarily on their highaffinity uptake system and fast translocation to plants (Smith
et al., 2011). Rapid P uptake may explain the consistently low
levels of in situ soil solution P in AM plot soils.
In addition to differences between AM and ECM plots, we
observed a seasonal decline in total Pi and an increase in total Po
across all plots. The total Pi decline is in accordance with established knowledge that trees and their mycorrhizal symbionts
assimilate inorganic P from the soil solution, which is in equilibrium with absorbed and complexed forms of Pi (Frossard et al.,
2011). In addition, we found that both hydroxide and acid Pi
pools varied significantly over the season, suggesting that these
pools contribute to seasonal Pi cycling. Although phosphatase
enzymes in soil hydrolyze Po, de novo formation of Po appeared
to exceed the decomposition of organic P substrates during the
growing season. Given that the increase in total Po happened
before litter fall, we conclude that the buildup of Po was likely to
be a product of rhizodeposition (e.g. sloughed cells) and turnover
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
of the microbial biomass (Richardson & Simpson, 2011). During
the winter season, decomposition as well as abiotic processes have
the potential to restore inorganic P levels by transforming accumulated organic P. Although seasonal changes in P pools could
reflect changes in the extractability of organic material (Chen
et al., 2003), significant differences in soil conditions (i.e. pH,
OM and P sorption) between mycorrhizal associations suggest
that the seasonal decrease in Pi and increase in Po are real changes
and are not an artifact of the extraction process. We conclude
that over the growing season, inorganic P substrates were assimilated and subsequently transformed to organic P substrates by
plants and microbes in both AM and ECM plots.
In addition to changes in total Pi and Po, we found that the
microbial biomass became increasingly P-limited over the season,
particularly at the end of the growing season. This contradicts the
idea that soil microbial biomass has a constant C : P ratio (Cleveland & Liptzin, 2007). This stoichiometric change is likely linked
to a seasonal shift from mycorrhizal to saprotrophic life-strategies
of the microbial community. During the growing season, when
mycorrhizal fungi mediate plant-derived C flux to the soil system,
mycorrhizal fungi rapidly assimilate and transfer P to their host
trees (Plassard & Dell, 2010). Trees that associate with mycorrhizal fungi experience P-limitation only if they fail to translocate
C to their fungal symbionts (Fellbaum et al., 2014). Towards the
end of the growing season, as leaves senesce and the belowground
C flow to roots decreases, the microbial biomass is expected to
shift from a predominantly root symbiotic community to one
dominated by free-living C-limited decomposers, as we observed
in the fungal community of ECM stands. Taken together, the
observed seasonal changes in total Pi, Po and MBP suggest that
both AM and ECM stands become more P-limited over the
course of the growing season.
Our study provides evidence that tree mycorrhizal associations
affect P cycling in hardwood forest soils. Common garden experiments have previously demonstrated that tree species significantly
affect soil nutrient cycling and decomposition activity within
30 yr of planting (Hobbie et al., 2007; Vesterdal et al., 2012). In
naturally generated forest stands, the causal link between vegetation and soil P cycling cannot be explicitly tested. However, the
studied plots have been vegetated for over 80 yr, have the same
parental material and experience no systematic climate differences. Hence, it is reasonable to assume that differences in soil P
cycling are determined by the current stand dynamics rather than
conditions prevailing before establishment. Our results partially
support the mycorrhizal-associated nutrient economy framework
(Phillips et al., 2013) with ECM plots having more organic P in
available forms compared with AM plots where more complex
organic P accumulated. However, seasonal buildup of organic P
in soil resulted in an increasingly P-limited microbial biomass in
both systems. These seasonal differences were larger than those
observed between mycorrhizal associations. Soil microbial
biomass was the largest biotic P pool, and we propose that
changes in seasonal tree derived C alters microbial community
composition and, subsequently, the cycling of P in soil. In the
future, it will be important to further investigate MBP turnover
rates, the importance of C availability for P cycling and the
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Research 1193
seasonal composition of the microbial community, to better
understand how P cycling is controlled in hardwood forest soils
and how it may be affected by changing environmental conditions.
Acknowledgements
This project was supported by grants from the Swedish research
councils FORMAS Grant 2008-1410 and VR Grant no. 20123950; the US Department of Energy-Office of Biological and
Environmental Research-Terrestrial Ecosystem Science Program;
and the US National Science Foundation (DEB, Ecosystem
Studies; no. 1153401). Additional support came from the Magn
Bergvalls and Carl Tryggers foundation. We thank Zach Brown,
Tyler Pietrykowski and Daniel Lehman for field and lab assistance, and Edward Brzostek for valuable discussions and laboratory assistance. The manuscript has been greatly improved by
valuable feedback from Ben Turner, Reiner Giesler, Marianne
Clarholm and three anonymous referees.
Author contributions
A.R., M.G.M. and R.P.P. planned and designed the study. A.R.
and M.G.M. conducted fieldwork and soil analysis. T.C. performed ergosterol analysis and H.U. performed molecular characterization of fungal communities. A.R., M.G.M. and P.F.
analyzed the data and A.R., M.G.M. and R.P.P. wrote the
manuscript with considerable input from the other authors.
References
Achat DL, Augusto L, Bakker MR, Gallet-Budynek A, Morel C. 2012.
Microbial processes controlling P availability in forest spodosols as
affected by soil depth and soil properties. Soil Biology and Biochemistry
44: 39–48.
Averill C, Turner BL, Finzi AC. 2014. Mycorrhiza-mediated competition between
plants and decomposers drives soil carbon storage. Nature 505: 543–545.
Becquer A, Trap J, Irshad U, Ali MA, Claude P. 2014. From soil to plant, the
journey of P through trophic relationships and ectomycorrhizal association.
Frontiers in Plant Science 5: 548.
Blankinship JC, Becerra CA, Schaeffer SM, Schimel JP. 2014. Separating
cellular metabolism from exoenzyme activity in soil organic matter
decomposition. Soil Biology and Biochemistry 71: 68–75.
Brookes PC, Powlson DS, Jenkinson DS. 1982. Measurement of microbial
biomass phosphorus in soil. Soil Biology and Biochemistry 14: 319–329.
Brookes PC, Powlson DS, Jenkinson DS. 1984. Phosphorus in the soil microbial
biomass. Soil Biology and Biochemistry 16: 169–175.
Burns RG, DeForest JL, Marxsen J, Sinsabaugh RL, Stromberger ME,
Wallenstein MD, Weintraub MN, Zoppini A. 2013. Soil enzymes in a
changing environment: current knowledge and future directions. Soil Biology
and Biochemistry 58: 216–234.
Chen CR, Condron LM, Davis MR, Sherlock RR. 2003. Seasonal changes in
soil phosphorus and associated microbial properties under adjacent grassland
and forest in New Zealand. Forest Ecology and Management 177: 539–557.
Clarholm M, Skyllberg U, Rosling A. 2015. Organic acid induced release of
nutrients from metal-stabilized soil organic matter – the unbutton model. Soil
Biology and Biochemistry 84: 168–176.
Cleveland C, Liptzin D. 2007. C : N : P stoichiometry in soil: is there a “Redfield
ratio” for the microbial biomass? Biogeochemistry 85: 235–252.
Cleveland CC, Houlton BZ, Smith WK, Marklein AR, Reed SC, Parton W,
Del Grosso SJ, Running SW. 2013. Patterns of new versus recycled primary
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
1194 Research
production in the terrestrial biosphere. Proceedings of the National Academy of
Sciences, USA 110: 12 733–12 737.
Cole CV, Innis G, Stewart JWB. 1978. Simulation of phosphorus cycling in
semiarid grasslands. In: Innis G, ed. Grassland simulation model. New York,
NY, USA: Springer, 205–230.
Cross AF, Schlesinger WH. 1995. A literature review and evaluation of the.
Hedley fractionation: applications to the biogeochemical cycle of soil
phosphorus in natural ecosystems. Geoderma 64: 197–214.
DeForest J, Smemo K, Burke D, Elliott H, Becker J. 2012. Soil microbial
responses to elevated phosphorus and pH in acidic temperate deciduous forests.
Biogeochemistry 109: 189–202.
Elser JJ, Bracken MES, Cleland EE, Gruner DS, Harpole WS, Hillebrand H,
Ngai JT, Seabloom EW, Shurin JB, Smith JE. 2007. Global analysis of
nitrogen and phosphorus limitation of primary producers in freshwater, marine
and terrestrial ecosystems. Ecology Letters 10: 1135–1142.
Fellbaum CR, Mensah JA, Cloos AJ, Strahan GE, Pfeffer PE, Kiers ET, B€
ucking
H. 2014. Fungal nutrient allocation in common mycorrhizal networks is
regulated by the carbon source strength of individual host plants. New
Phytologist 203: 646–656.
Finzi A. 2009. Decades of atmospheric deposition have not resulted in
widespread phosphorus limitation or saturation of tree demand for nitrogen in
southern New England. Biogeochemistry 92: 217–229.
Frossard E, Achat D, Bernasconi S, B€
unemann E, Fardeau J-C, Jansa J, Morel
C, Rabeharisoa L, Randriamanantsoa L, Sinaj S et al. 2011. The use of tracers
to investigate phosphate cycling in soil–plant systems. In: B€
unemann E,
Oberson A, Frossard E, eds. Phosphorus in action. Berlin, Germany: Springer,
59–91.
Giesler R, Petersson T, H€ogberg P. 2002. Phosphorus limitation in boreal
forests: effects of aluminum and iron accumulation in the humus layer.
Ecosystems 5: 300–314.
Groffman PM, Fisk MC. 2011. Phosphate additions have no effect on microbial
biomass and activity in a northern hardwood forest. Soil Biology and
Biochemistry 43: 2441–2449.
Harpole WS, Ngai JT, Cleland EE, Seabloom EW, Borer ET, Bracken MES,
Elser JJ, Gruner DS, Hillebrand H, Shurin JB et al. 2011. Nutrient colimitation of primary producer communities. Ecology Letters 14: 852–862.
H€a ussling M, Marschner H. 1989. Organic and inorganic soil phosphates and
acid phosphatase activity in the rhizosphere of 80-year-old Norway spruce
[Picea abies (L.) Karst.] trees. Biology and Fertility of Soils 8: 128–133.
Hedley MJ, Stewart JWB, Chauhan BS. 1982a. Changes in inorganic and
organic soil phosphorus fractions induced by cultivation practices and by
laboratory incubations. Soil Science Society of America Journal 46: 970–976.
Hedley MJ, White RE, Nye PH. 1982b. Plant-Induced changes in the
rhizosphere of rape (Brassica napus var. emerald) seedlings. New Phytologist 91:
45–56.
van der Heijden EW. 2001. Differential benefits of arbuscular mycorrhizal and
ectomycorrhizal infection of Salix repens. Mycorrhiza 10: 185–193.
Hobbie SE, Ogdahl M, Chorover J, Chadwick OA, Oleksyn J, Zytkowiak R,
Reich PB. 2007. Tree species effects on soil organic matter dynamics: the role
of soil cation composition. Ecosystems 10: 999–1018.
Hoffland E, Kuyper TW, Wallander H, Plassard C, Gorbushina AA,
Haselwandter K, Holmstr€om S, Landeweert R, Lundstr€om US, Rosling A.
2004. The role of fungi in weathering. Frontiers in Ecology and the Environment
2: 258–264.
Ihrmark K, Bodeker IT, Cruz-Martinez K, Friberg H, Kubartova A, Schenck J,
Strid Y, Stenlid J, Brandstrom-Durling M, Clemmensen KE et al. 2012. New
primers to amplify the fungal ITS2 region–evaluation by 454-sequencing of
artificial and natural communities. FEMS Microbiology Ecology 82: 666–677.
Jansa J, Finlay R, Wallander H, Smith FA, Smith S. 2011. Role of mycorrhizal
symbioses in phosphorus cycling. In: B€
unemann E, Oberson A, Frossard E,
eds. Phosphorus in action. Berlin, Germany: Springer, 137–168.
Jenkinson DS, Brookes PC, Powlson DS. 2004. Measuring soil microbial
biomass. Soil Biology and Biochemistry 36: 5–7.
Johnson A, Frizano J, Vann D. 2003. Biogeochemical implications of labile
phosphorus in forest soils determined by the Hedley fractionation procedure.
Oecologia 135: 487–499.
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com
New
Phytologist
Jonard M, F€
urst A, Verstraeten A, Thimonier A, Timmermann V, Potocic N,
Waldner P, Benham S, Hansen K, Meril€a P. 2015. Tree mineral nutrition is
deteriorating in Europe. Global Change Biology 21: 418–430.
Joner E, van Aarle I, Vosatka M. 2000. Phosphatase activity of extra-radical
arbuscular mycorrhizal hyphae: a review. Plant and Soil 226: 199–210.
Jones MD, Durall D, Tinker P. 1998. A comparison of arbuscular and
ectomycorrhizal Eucalyptus coccifera: growth response, phosphorus uptake
efficiency and external hyphal production. New Phytologist 140: 125–134.
Kiers ET, Duhamel M, Beesetty Y, Mensah JA, Franken O, Verbruggen E,
Fellbaum CR, Kowalchuk GA, Hart MM, Bago A et al. 2011. Reciprocal
rewards stabilize cooperation in the mycorrhizal symbiosis. Science 333: 880–
882.
Koele N, Dickie IA, Blum JD, Gleason JD, de Graaf L. 2014. Ecological
significance of mineral weathering in ectomycorrhizal and arbuscular
mycorrhizal ecosystems from a field-based comparison. Soil Biology and
Biochemistry 69: 63–70.
Koide RT, Kabir Z. 2000. Extraradical hyphae of the mycorrhizal fungus Glomus
intraradices can hydrolyse organic phosphate. New Phytologist 148: 511–517.
Lewis DG, Quirk JP. 1967. Phosphate diffusion in soil and uptake by plants.
Plant and Soil 26: 99–118.
Lindahl B, Stenlid JAN, Olsson S, Finlay R. 1999. Translocation of 32P between
interacting mycelia of a wood-decomposing fungus and ectomycorrhizal fungi
in microcosm systems. New Phytologist 144: 183–193.
McCune B, Grace JB. 2002. Analysis of ecological communities. MjM Software
Design, Gleneden Beach, OR, USA.
McCune B, Mefford MJ. 2006. PC-ORD. Multivariat analysis of ecological data.
Version 5.33d. MjM Software Design, Gleneden Beach, OR, USA.
Nygren C, Rosling A. 2009. Localisation of phosphomonoesterase activity in
ectomycorrhizal fungi grown on different phosphorus sources. Mycorrhiza 19:
197–204.
Olsson PA, Larsson L, Bago B, Wallander H, Van Aarle IM. 2003. Ergosterol
and fatty acids for biomass estimation of mycorrhizal fungi. New Phytologist
159: 7–10.
Phillips RP, Brzostek E, Midgley MG. 2013. The mycorrhizal-associated
nutrient economy: a new framework for predicting carbon–nutrient couplings
in temperate forests. New Phytologist 199: 41–51.
Plassard C, Dell B. 2010. Phosphorus nutrition of mycorrhizal trees. Tree
Physiology 30: 1129–1139.
Plassard C, Louche J, Ali MA, Duchemin M, Legname E, Cloutier-Hurteau B.
2011. Diversity in phosphorus mobilisation and uptake in ectomycorrhizal
fungi. Annals of Forest Science 68: 33–43.
Read DJ, Perez-Moreno J. 2003. Mycorrhizas and nutrient cycling in ecosystems
– a journey towards relevance? New Phytologist 157: 475–492.
Rennenberg H, Herschbach C. 2013. Phosphorus nutrition of woody plants:
many questions – few answers. Plant Biology 15: 785–788.
Richardson AE, Simpson RJ. 2011. Soil microorganisms mediating phosphorus
availability update on microbial phosphorus. Plant Physiology 156: 989–996.
Saiya-Cork KR, Sinsabaugh RL, Zak DR. 2002. The effects of long term
nitrogen deposition on extracellular enzyme activity in an Acer saccharum forest
soil. Soil Biology & Biochemistry 34: 1309–1315.
Schmidt PA, Balint M, Greshake B, Bandow C, Rombke J, Schmitt I. 2013.
Illumina metabarcoding of a soil fungal community. Soil Biology &
Biochemistry 65: 128–132.
Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge JL, Levesque CA, Chen
W, Fungal Barcoding Consortium. 2012. Nuclear ribosomal internal
transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi.
Proceedings of the National Academy of Sciences, USA 109: 6241–6246.
Shaw AN, DeForest JL. 2013. The cycling of readily available phosphorus in
response to elevated phosphate in acidic temperate deciduous forests. Applied
Soil Ecology 63: 88–93.
Smith SE, Jakobsen I, Grønlund M, Smith FA. 2011. Roles of arbuscular
mycorrhizas in plant phosphorus nutrition: interactions between pathways of
phosphorus uptake in arbuscular mycorrhizal roots have important
implications for understanding and manipulating plant phosphorus
acquisition. Plant Physiology 156: 1050–1057.
Smith SE, Read DJ. 2008. Mycorrhizal symbiosis. London, UK: Academic Press.
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
New
Phytologist
Sommers L, Nelson D. 1972. Determination of total phosphorus in soils: a rapid
perchloric acid digestion procedure. Soil Science Society of America Journal 36:
902–904.
Turner B, Condron L, Richardson S, Peltzer D, Allison V. 2007. Soil organic
phosphorus transformations during pedogenesis. Ecosystems 10: 1166–1181.
Turner B, Lambers H, Condron L, Cramer M, Leake J, Richardson A, Smith S.
2013. Soil microbial biomass and the fate of phosphorus during long-term
ecosystem development. Plant and Soil 367: 225–234.
Vance E, Brookes P, Jenkinson D. 1987. An extraction method for measuring
soil microbial biomass C. Soil Biology and Biochemistry 19: 703–707.
Vesterdal L, Elberling B, Christiansen JR, Callesen I, Schmidt IK. 2012. Soil
respiration and rates of soil carbon turnover differ among six common
European tree species. Forest Ecology and Management 264: 185–196.
Vitousek PM, Porder S, Houlton BZ, Chadwick OA. 2010. Terrestrial
phosphorus limitation: mechanisms, implications, and nitrogen–phosphorus
interactions. Ecological Applications 20: 5–15.
Walker TW, Syers JK. 1976. The fate of phosphorus during pedogenesis.
Geoderma 15: 1–19.
Wallander H, Nylund JE. 1992. Effects of excess nitrogen and phosphorus
starvation on the extramatrical mycelium of ectomycorrhizas of Pinus sylvestris
L. New Phytologist 120: 495–503.
Weand M, Arthur M, Lovett G, Sikora F, Weathers K. 2010. The phosphorus
status of northern hardwoods differs by species but is unaffected by nitrogen
fertilization. Biogeochemistry 97: 159–181.
Wood T, Bormann FH, Voigt GK. 1984. Phosphorus cycling in a northern
hardwood forest: biological and chemical control. Science 223: 391–393.
Xiao J, Zhuang Q, Law BE, Baldocchi DD, Chen J, Richardson AD, Melillo
JM, Davis KJ, Hollinger DY, Wharton S. 2011. Assessing net ecosystem
carbon exchange of US terrestrial ecosystems by integrating eddy covariance
flux measurements and satellite observations. Agricultural and Forest
Meteorology 151: 60–69.
Yang X, Post WM. 2011. Phosphorus transformations as a function of
pedogenesis: a synthesis of soil phosphorus data using Hedley fractionation
method. Biogeosciences 8: 2907–2916.
Zhao Q, Zeng D-H, Fan Z-P, Yu Z-Y, Hu Y-L, Zhang J. 2009. Seasonal
variations in phosphorus fractions in semiarid sandy soils under different
vegetation types. Forest Ecology and Management 258: 1376–1382.
Supporting Information
Additional supporting information may be found in the online
version of this article.
Fig. S1 Map of plot locations.
Ó 2015 The Authors
New Phytologist Ó 2015 New Phytologist Trust
Research 1195
Fig. S2 PCA showing difference between AM- and ECMdominated plots, and soil layers..
Fig. S3 Total soil enzymatic activity vs ergosterol and MBC.
Fig. S4 Microbial biomass carbon (MBC) and ergosterol.
Fig. S5 PCA of fungal community structure of topsoil from four
AM- and ECM-dominated plots.
Fig. S6 Seasonal dynamics in total Po and Pi as well as MBC and
P.
Table S1 Seasonal average soil P contents (mg P kg1 dry soil) by
mycorrhizal association and soil depth.
Table S2 Average soil P contents normalized by soil ergosterol
(mg P mg1 ergosterol).
Table S3 Average relative abundance of the ten commonest
orders of ectomycorrhizal fungi.
Table S4 Estimated seasonal average size of P pool (kg ha1).
Methods S1 In situ soil solution P.
Methods S2 Sequential P extraction.
Methods S3 Soil enzymatic activity.
Methods S4 Ergosterol extraction.
Methods S5 Fungal community sequence analysis.
Please note: Wiley Blackwell are not responsible for the content
or functionality of any supporting information supplied by the
authors. Any queries (other than missing material) should be
directed to the New Phytologist Central Office.
New Phytologist (2016) 209: 1184–1195
www.newphytologist.com