ArticlePDF Available

Ectomycorrhizal and saprotrophic fungal diversity are linked to different tree community attributes in a field-based tree experiment

Authors:

Abstract and Figures

Exploring the link between above- and belowground biodiversity has been a major theme of recent ecological research, due in large part to the increasingly well-recognized role that soil microorganisms play in driving plant community processes. In this study, we utilized a field-based tree experiment in Minnesota, USA to assess the effect of changes in plant species richness and phylogenetic diversity on the richness and composition of both ectomycorrhizal and saprotrophic fungal communities. We found that ectomycorrhizal fungal species richness was significantly positively influenced by increasing plant phylogenetic diversity, while saprotrophic fungal species richness was significantly affected by plant leaf nitrogen content, specific root length, and standing biomass. The increasing ectomycorrhizal fungal richness associated with increasing plant phylogenetic diversity was driven by the combined presence of ectomycorrhizal fungal specialists in plots with both gymnosperm and angiosperm hosts. Although the species composition of both the ectomycorrhizal and saprotrophic fungal communities changed significantly in response to changes in plant species composition, the effect was much greater for ectomycorrhizal fungi. In addition, ectomycorrhizal but not saprotrophic fungal species composition was significantly influenced by both plant phylum (angiosperm, gymnosperm, both) and origin (Europe, America, both). While the phylum effect was driven by differences in ectomycorrhizal fungal community composition, the origin effect was driven by differences in community heterogeneity. Taken together, this study emphasizes the guild-specific nature of plant-associated effects on soil fungal communities and provides a mechanistic basis for the positive link between plant phylogenetic diversity and ectomycorrhizal fungal richness. This article is protected by copyright. All rights reserved.
Content may be subject to copyright.
For Review Only
Ectomycorrhizal and sa
protrophic fungal diversity are linked
to different tree community attributes in a field-based tree
experiment
Journal:
Molecular Ecology
Manuscript ID
MEC-15-1506.R2
Manuscript Type:
Original Article
Date Submitted by the Author:
n/a
Complete List of Authors:
Nguyen, Nhu; University of Minnesota Twin Cities, Plant Biology
Williams, Laura; University of Minnesota Twin Cities, Ecology, Evolution,
and Behavior
Stefanski, Artur; University of Minnesota Twin Cities, Department of Forest
Resources
Vincent, John; University of Minnesota, Department of Plant Biology
Cavender-Bares, Jeannine; University of Minnesota Twin Cities, Ecology,
Evolution, and Behavior
Messier, Christian; Universite du Quebec a Montreal, Department of
Biological Sciences
Paquette, Alain; Universite du Quebec a Montreal, Department of Biological
Sciences
Gravel, Dominique; Universite de Sherbrooke, Department of Biology
Reich, Peter; University of Minnesota Twin Cities, Department of Forest
Resources
Kennedy, Peter; University of Minnesota Twin Cities, Plant Biology &
Ecology, Evolution, and Behavior
Keywords:
Community Ecology, Fungi, Species Interactions, DNA Barcoding
Note: The following files were submitted by the author for peer review, but cannot be converted to
PDF. You must view these files (e.g. movies) online.
TableS3.ECM.AllModels.csv
TableS4.SAP.AllModels.csv
Molecular Ecology
For Review Only
1
Ectomycorrhizal and saprotrophic fungal diversity are linked to different tree community 1
attributes in a field-based tree experiment 2
3
4
5
Nhu H. Nguyen
1
, Laura Williams
2
, John B. Vincent
1
, Artur Stefanski
3
, Jeannine 6
Cavender-Bares
2
, Christian Messier
4
, Alain Paquette
4
, Dominique Gravel
5
, Peter B. 7
Reich
3
, Peter G. Kennedy
1,2
8
9
10
11
1. Department of Plant Biology, University of Minnesota, St. Paul, MN, USA 12
2. Department of Ecology, Evolution, and Behavior, University of Minnesota, St. Paul, 13
MN, USA 14
3. Department of Forest Resources, University of Minnesota, St. Paul, MN, USA 15
4. Department of Biological Sciences, University of Quebec, Montreal, Montreal, 16
Quebec, Canada 17
5. Department of Biology, University of Sherbrooke, Sherbrooke, Quebec, Canada 18
19
20
21
Word Count: 6908 (excluding abstract and references) 22
Tables: 2 23
Figures: 3 24
Supplementary Information: Table S1-S8, Figures S1-S8. 25
Keywords: fungal guild, plant phylogenetic diversity, host specificity, geographic origin, 26
microbial richness 27
Running Title: Soil fungal responses to tree diversity 28
Corresponding Author: Peter Kennedy, 1479 Gortner Ave., St. Paul, MN, USA. Email: 29
kennedyp@umn.edu, Phone: 612-624-8519, Fax: 612-625-1738. 30
Page 1 of 52 Molecular Ecology
For Review Only
2
Abstract 31
Exploring the link between above- and belowground biodiversity has been a major theme 32
of recent ecological research, due in large part to the increasingly well-recognized role 33
that soil microorganisms play in driving plant community processes. In this study, we 34
utilized a field-based tree experiment in Minnesota, USA to assess the effect of changes 35
in plant species richness and phylogenetic diversity on the richness and composition of 36
both ectomycorrhizal and saprotrophic fungal communities. We found that 37
ectomycorrhizal fungal species richness was significantly positively influenced by 38
increasing plant phylogenetic diversity, while saprotrophic fungal species richness was 39
significantly affected by plant leaf nitrogen content, specific root length, and standing 40
biomass. The increasing ectomycorrhizal fungal richness associated with increasing plant 41
phylogenetic diversity was driven by the combined presence of ectomycorrhizal fungal 42
specialists in plots with both gymnosperm and angiosperm hosts. Although the species 43
composition of both the ectomycorrhizal and saprotrophic fungal communities changed 44
significantly in response to changes in plant species composition, the effect was much 45
greater for ectomycorrhizal fungi. In addition, ectomycorrhizal but not saprotrophic 46
fungal species composition was significantly influenced by both plant phylum 47
(angiosperm, gymnosperm, both) and origin (Europe, America, both). While the phylum 48
effect was driven by differences in ectomycorrhizal fungal community composition, the 49
origin effect was driven by differences in community heterogeneity. Taken together, this 50
study emphasizes the guild-specific nature of plant-associated effects on soil fungal 51
communities and provides a mechanistic basis for the positive link between plant 52
phylogenetic diversity and ectomycorrhizal fungal richness. 53
Page 2 of 52Molecular Ecology
For Review Only
3
Introduction 54
Understanding the link between above- and belowground biodiversity has been a major 55
theme of recent ecological research, due in large part to the increasingly well-recognized 56
role that soil microorganisms play in driving plant community productivity (Van der 57
Heijden et al. 2008), vegetational succession dynamics (Van der Putten et al. 1993, Nara 58
and Hogetsu 2004), and ecosystem functioning (Bardgett and Van der Putten 2014). It is 59
now widely accepted that strong feedbacks exist between plant and soil microbial 60
communities, although the direction and intensity of those feedbacks are spatially (De 61
Deyn and Van der Putten 2005), temporally (Eisenhauer et al. 2011), and 62
environmentally context-dependent (Smith and Reynolds 2015). Along with better 63
understanding the basic functioning of ecosystems, determining the extent to which 64
above- and belowground communities are coupled also has significant conservation 65
implications (Wardle 2006, Wagg et al. 2014). 66
67
One of the most fundamental aspects of the link between plant and soil microbial 68
communities is the extent to which above- and belowground species richness are 69
interdependent. Many of the theoretical treatments of this topic suggest that the 70
relationship between plant and soil microbial richness should be positive due to greater 71
heterogeneity in resources and environmental conditions as plant richness increases 72
(Hooper et al. 2005, Wardle et al. 2006). Despite a well-established conceptual 73
foundation, and evidence for a positive relationship between plant richness and above-74
ground food web richness (Haddad et al. 2009), the effect of plant richness on soil 75
microbial richness remains equivocal. A number of field-based surveys indicate that 76
Page 3 of 52 Molecular Ecology
For Review Only
4
bacterial species richness is actually poorly predicted by plant species richness (Fierer 77
and Jackson 2006, Bryant et al. 2008, Prober et al. 2014), with environmental variables 78
such as soil pH being more significant ecological drivers (Lauber et al. 2009). For soil 79
fungi, which may be more strongly influenced by plant effects on soil due to their growth 80
at larger spatial scales (Urbanova et al. 2015), the evidence is currently mixed. Tedersoo 81
et al. (2014a) found that abiotic factors, particularly mean annual precipitation, were 82
significantly more important than plant species richness in determining global hotspots of 83
soil fungal richness. Similarly, Prober et al. (2014) found no consistent relationship 84
between plant and soil fungal richness at 25 grassland sites located on four continents. In 85
contrast, Gao et al. (2013) found that ectomycorrhizal fungal richness, a dominant fungal 86
guild in many forests globally, increased significantly with increasing plant species 87
richness in both temperate and tropical forests. At local spatial scales, positive 88
correlations between plant and soil fungal richness has been observed in both forest (Peay 89
et al. 2013, Tedersoo et al. 2015) and grassland ecosystems (Pellisimer et al. 2014). 90
91
An inherent challenge of field surveys of plant-soil microbial richness relationships is the 92
correlation of species richness with other environmental variables. For example, Landis 93
et al. (2004) found that arbuscular mycorrhizal fungal richness was positively associated 94
with plant species richness, but both also increased with greater nitrogen availability and 95
changes in soil texture. Disentangling this relationship can often be achieved with 96
experimental plantings of plant species, but those can also have shortcomings, 97
particularly if monocultures are not present to account for sampling effects (Huston 1997, 98
Tedersoo et al. 2014b). In the field-based studies with the proper set of experimental 99
Page 4 of 52Molecular Ecology
For Review Only
5
treatments, there has been general support for a strong relationship between plant and soil 100
fungal communities. Both Burrows and Pfleger (2002) and Hiiselu et al. (2014) found 101
that increasing plant species richness resulted in a significant increase in the richness of 102
arbuscular mycorrhizal fungal communities. Working in the same study system, Antonika 103
et al. (2011) found arbuscular mycorrhizal fungal richness was higher in mono- and 104
polyculture plots, but LeBlanc et al. (2015) recently found the reversed pattern for the 105
bulk soil fungal community. Greenhouse studies explicitly manipulating plant and soil 106
fungal community richness have also typically found a strong relationship (Van der 107
Heijden et al. 1998, Vogelsang et al. 2006). Despite this growing consensus nearly all of 108
the experimental studies examining this relationship have focused on herbaceous plant 109
communities, particularly grasses. For tree-based ecosystems, Tedersoo et al. (2015) 110
recently documented a strong relationship between tree and ectomycorrhizal fungal 111
richness at one experimental site (Finland) but no relationship at another (Estonia). Given 112
the divergence in response between sites, determining the nature of this relationship in 113
other experimental forest settings is important for making broader ecological predictions. 114
115
Despite the greater study of the links between plant and microbial community richness 116
and composition, the mechanisms underlying these trends remain largely unresolved. 117
Some research suggests that the positive relationships between plant and soil fungal 118
community richness are not driven by the number of plant species, but rather by the 119
increased productivity in plots with higher plant species richness (Waldrop et al. 2006). A 120
recent study in which plant biomass (a traditional measure of productivity) was carefully 121
measured belowground, however, found that increases in belowground plant biomass 122
Page 5 of 52 Molecular Ecology
For Review Only
6
(and species richness) were actually negatively correlated with arbuscular mycorrhizal 123
fungal community richness (Hiiesalu et al. 2014). Those results suggest the effect of plant 124
productivity on fungal richness may be system- and measurement-type dependent. An 125
intuitive but less well-discussed alternative factor driving positive relationships between 126
plant and soil fungal community richness is differential host specificity, which would 127
result in higher fungal richness in plots with more plant species. Support for the role of 128
host specificity in increasing fungal richness has been shown in observational 129
ectomycorrhizal fungal community studies (Kernaghan et al. 2003, Ishida et al. 2007, 130
Tedersoo et al. 2013, Gao et al. 2013), as well as experimental studies of fungal 131
pathogens (Rottstock et al. 2014), although the absence of monoculture plots have often 132
confounded this interpretation (Tedersoo et al. 2014a, but see Gao et al. 2014). Work in 133
plant-fungal pathogen systems also indicates that specificity-richness relationships may 134
be strongly mediated by host phylogenetic relationships (Gilbert and Webb 2007), which 135
would suggest that plots containing more phylogenetically distant plant species should 136
have higher fungal richness than those with the same number of more closely related 137
plant hosts. Indirect support for this effect comes from the recent study of global fungal 138
richness patterns, where ectomycorrhizal fungal richness peaked in temperate forests, 139
which contained higher host phylogenetic diversity than ectomycorrhizal-dominated 140
tropical forests (Tedersoo et al. 2014b). 141
142
The recent surge in research on links between above- and belowground species 143
communities has been facilitated by the application of molecular techniques, which have 144
allowed ecologists to more accurately quantify microbial community richness and 145
Page 6 of 52Molecular Ecology
For Review Only
7
composition (Peay et al. 2008, Caporaso et al. 2011). Importantly, current methods allow 146
researchers to quantify entire microbial communities rather just individual functional 147
guilds (Nguyen et al. 2016). Differences among guilds in their response to plant richness 148
might be expected, particularly between biotrophic guilds that directly interact with 149
plants and saprotrophic guilds that interact with plants indirectly (Urbanova et al. 2015). 150
Support for a guild-level difference was shown by Peay et al. (2013) and at the Finnish 151
site in Tedersoo et al. (2015), where the richness of symbiotic soil fungal guilds were 152
positively associated with higher plant richness, but the richness of saprotrophic fungi 153
was not. However, both Prober et al. (2014) and Pellissier et al. (2014) found contrasting 154
results, with the richness of both mutualistic and total soil fungal communities not being 155
significantly affected by plant richness. 156
157
In this study, we utilized a field-based tree experiment to test the link between plant and 158
soil fungal community richness and composition. The design of the experimental site we 159
sampled, which includes both mono- and different kinds of mixed-species plots, allowed 160
us to account directly for sampling effects, while testing the effect of plant species 161
richness, phylogenetic diversity, and range of additional biotic and abiotic parameters on 162
soil fungal richness and composition. In addition, the experiment includes pairings of tree 163
species from multiple tree genera, in which one tree species was native to North America 164
while the other was native to Europe. This phylogenetically controlled design also 165
allowed for an explicit testing of the effect of plant origin on soil fungal community 166
richness and composition. Based on previous observational studies, we hypothesized that 167
plant and soil fungal community richness would be positively correlated, with the 168
Page 7 of 52 Molecular Ecology
For Review Only
8
greatest effects being apparent in mixed-host species plots with higher plant phylogenetic 169
diversity. If fungal host specificity was the important mechanism driving this 170
relationship, we also expected to see many fungal species present only in the mono- and 171
mixed-species plots containing their specific host species. With regard to the effects of 172
plant origin, we anticipated higher fungal richness in the plots with North American hosts 173
compared to those from Europe, due a greater likelihood of compatible symbionts being 174
present. Finally, similar to fungal species richness, we expected that patterns of fungal 175
community composition would also be significantly influenced by both plant 176
phylogenetic diversity and plant origin. Given the potential for differential responses to 177
plant species richness and composition among fungal guilds, we examined the 178
aforementioned predictions for both ectomycorrhizal and saprotrophic fungal 179
communities. 180
181
Methods 182
Experiment Design and Environmental Sampling 183
This study conducted in the high-density tree-diversity experiment established in 184
Cloquet, Minnesota, USA as part of the International Diversity Experiment Network with 185
Trees (IDENT, Tobner et al. 2013). In the spring of 2010, 9408 tree seedlings were 186
planted at the Cloquet IDENT site (Fig. S1). Six common North American temperate-187
boreal tree species were planted (Acer saccharum, Betula papyrifera, Larix laricina, 188
Picea glauca, Pinus strobus, Quercus rubra) along with a congener of each species 189
originating from Europe (A. platanoides, B. pendula, L. decidua, Picea abies, Pinus 190
sylvestris, and Q. robur). Trees were planted in a grid pattern with 0.4 m spacing among 191
Page 8 of 52Molecular Ecology
For Review Only
9
seedlings to form plots 2.8 x 2.8 m in size containing 49 trees. Plots were spaced one m 192
apart. Here we focused on a subset of the plots in the experiment: those plots where each 193
species was growing in monoculture and four six-species mixtures. The latter contained 194
either 1) six North American species, 2) six European species, 2) six gymnosperms, or 2) 195
six angiosperms, respectively. Each of these plots was replicated four times within the 196
experiment in a randomized block design. 197
198
We characterized the tree communities and environmental conditions of each plot in 199
several ways, using the core area containing 25 trees (to minimize edge effects). We 200
calculated plot plant species richness by counting the number of plant species present. 201
We used the phylogenetic species variability metric (PSV; Helmus et al. 2007) to 202
quantify plot plant phylogenetic diversity (calculated in R using the Picante package). 203
Unlike other phylogenetic diversity metrics such as Faith’s PD, PSV is independent of 204
species richness(Pearse et al. 2014). To estimate standing biomass in the plots, we 205
measured the stem basal diameter (D) and stem height (H) for each tree at the end of the 206
2013 growing season. These measurements along with wood density (WD) values 207
obtained from literature (Chave et al. 2009) were used to calculate standing biomass (D
2
x 208
H x WD). To assess specific root length in the plots, we calculated a community-209
weighted mean using in situ measurements. In June 2014, we sampled trees in each 210
monoculture plot growing in the outside row. For each tree, we analyzed scanned images 211
of three to five sub-samples of first to third order roots (following Pregitzer et al. 2002) 212
using WinRhizo (Regent Instruments Inc., Quebec, Canada). Specific root lengths for 213
each species were calculated as the average of measurements taken on six trees and the 214
Page 9 of 52 Molecular Ecology
For Review Only
10
community-weighted specific root length was calculated by weighting species’ mean 215
values by the proportion of trees on a plot that belonged to each species. 216
217
We assessed plot leaf nitrogen concentration using the same community-weighted 218
approach as for specific root length. For this measure, one fully-expanded, sun-exposed 219
leaf (or five fascicles for Pinus and approx. 20 needles for Picea and Larix spp.) was 220
collected in June 2013 from each of five trees chosen at random in each monoculture plot 221
and pooled to give one sample per species per block. Each sample was dried at 65°C, 222
finely ground and analyzed for total nitrogen and carbon at the University of Nebraska, 223
Lincoln, using a Costech ECS 4010 element analyzer. We estimated plot leaf area index 224
(LAI) during peak biomass (the last two weeks of August 2013) using a line intercept 225
method. Fifteen vertical lines were located randomly on each plot. For each line, the 226
species identity was recorded for each leaf or gymnosperm branchlet intercepted, and the 227
angle of each leaf or branchlet intercepted was measured to the nearest 15 degrees. A 228
subsample of gymnosperm branchlets was destructively sampled to estimate the ratio of 229
interceptable to total leaf area (Thomas and Winner 2000). For all of the aforementioned 230
measurements, we calculated plot values in two ways, once including all of the tree 231
species on the plot and again excluding the two species that were not ectomycorrhizal 232
hosts (i.e., A. saccharum and A. platanoides). 233
234
To assess the physical environment, we estimated plot surface soil moisture (% 235
volumetric water content) by averaging six measurements from the start of June and then 236
two-week intervals from the start of July to mid-August 2013. Measurements were taken 237
Page 10 of 52Molecular Ecology
For Review Only
11
to 12 cm depth at four locations per plot using time domain reflectometry with a 238
FieldScout TDR 300 (standard soil type; Spectrum Technologies Inc., Illinois, USA). We 239
determined plot soil pH from the soil samples collected for fungal molecular analyses 240
using a water-based laboratory method (Soil Survey Staff 2004). We estimated the 241
standing pools of two forms of inorganic, plant-available nitrogen (nitrate and 242
ammonium) at the end of June 2013. Two soil cores (2.54 cm diameter and 15 cm deep) 243
were collected from random locations on each plot using a PVC corer and pooled to give 244
one sample per plot. Samples were collected, sieved, homogenized and extracted with 2 245
M KCl within 48 h. Extracts were analyzed for nitrate and ammonium using colorimetric 246
assays (Hood-Nowotny et al. 2010). 247
248
Fungal Sampling and Molecular Identification 249
On 28 October 2013, we collected soil cores from each of the four replicate plots of the 250
16 tree species treatments (N = 64). We first removed the surface litter layer and then 251
took four individual 2.5 x 10 cm deep cores located 0.8 m in from each plot edge (all 252
cores were taken with surface-sterilized PVC cores to prevent any cross contamination). 253
The four cores were combined into a single plastic bag and immediately stored on ice. 254
We returned the soil samples to the laboratory the same day and processed them for 255
molecular analyses within 48 hours of field collection. 256
Soils were homogenized by shaking and then sieved through a 2 mm surface-sterilized 257
mesh (to remove roots and other plant material). DNA was extracted from 10 g of soil per 258
sample using the PowerMax Soil DNA Isolation Kit (MO Bio, Carlsbad, CA, USA) 259
according to the manufacturer’s instructions. The ITS1 rDNA subunit was amplified 260
Page 11 of 52 Molecular Ecology
For Review Only
12
using a barcoded fungal-specific ITS1F-ITS2 primer set (Smith and Peay 2014). For each 261
sample, we ran three individual PCR reactions at different annealing temperatures (55, 262
56.5, 58°C), which were pooled to minimize the possibility of annealing temperature 263
biases. Along with the experimental samples, we included a series of field, lab extraction, 264
and PCR negative controls as well as a mock community as a positive control (Nguyen et 265
al. 2015). Amplified products were magnetically cleaned using the Agencourt arbuscular 266
mycorrhizalPure XP kit (Beckman Coulter, Brea, CA, USA) and quantified using the 267
Qubit dsDNA HS Fluorometer (Life Technologies, Carlsbad, CA, USA) according to 268
manufacturer’s instructions. The final library was sequenced at the University of 269
Minnesota Genomics Center using the 250 bp paired-end MiSeq Illumina platform. Raw 270
sequence data and associated metadata were deposited in MG-RAST under project 271
number 13302. 272
Raw sequences were demultiplexed, quality filtered using Phred = 20, and culled of 273
sequences less than 125 bp, with any ambiguous bases, or a homopolymer run of >8 bp 274
using QIIME v 1.8 (Caporaso et al. 2010) and MOTHUR v 1.33.3 (Schloss et al. 2009). 275
In this particular run, the reverse reads had poor quality and pairing them with forward 276
reads much reduced the final number of sequences (Nguyen et al. 2015). Therefore, we 277
chose to use only the forward reads for all of the final analyses. Following the guidelines 278
discussed in Nguyen et al. (2015), we employed a multi-step operational taxonomic unit 279
(OTU) picking strategy in QIIME by first clustering with reference USEARCH 280
(including reference and de novo chimera checking) at a 95% sequence similarity 281
followed by denovo UCLUST at 95% sequence similarity. We found that these criteria 282
best recovered the mock community and therefore applied them to the rest of the 283
Page 12 of 52Molecular Ecology
For Review Only
13
samples. The UNITE v6 database (Kõljalg et al. 2013; http://unite.ut.ee/repository.php) 284
was used in chimera checking and OTU clustering. OTUs were identified using NCBI 285
BLAST+ v2.2.29 (Altschul et al. 1990) and a custom database with sequences from the 286
UNITE v6 and vouchered mushrooms collected from the field site. Since sequences with 287
a low length/qlen ratio were typically non-fungal or contained a very short fungal 288
segment, we kept only sequences had a length/qlen 0.845 from BLAST. OTUs that 289
occurred in the negative controls were removed by subtracting the number of sequences 290
of each OTU present in the negative controls from the sequence abundance of that OTU 291
in the experimental samples. To account for differences in sequence reads among 292
samples, all samples were rarefied to 30,300 sequences, which represented the lowest 293
total in one of our 64 samples. In preliminary data analyses, we compared this approach 294
to that used by Tedersoo et al. (2015), which was based on residual richness after 295
accounting for sequence reads and OTU richness. The results were similar, suggesting 296
that our rarefaction approach did not generate any notable bias (McMurdie and Holmes 297
2014). The final raw community matrices, including taxonomic identification for each 298
OTU (hereafter referred to as species), are provided in Tables S1 (ectomycorrhizal fungi) 299
and S2 (saprotrophic fungi). 300
Statistical Analyses 301
While the use of high throughput sequencing has become a standard way to study soil 302
fungal communities, there is growing recognition that multiple quality control steps are 303
needed to ensure ecological accurate analyses (Lindahl et al. 2013, Nguyen et al. 2015, 304
Song et al. 2015). To address the issue of false positives in our data (i.e. sequences 305
present in samples due to spurious biological sources and/or sequencing artifacts such as 306
Page 13 of 52 Molecular Ecology
For Review Only
14
tag switching (Carlsen et al. 2012)), we began by examining the single-species Acer 307
plots, which do not host ectomycorrhizal fungi. Over 70% (842/1176) of the sample-by-308
species combinations had no ectomycorrhizal fungal sequences, but 95 of the 147 309
ectomycorrhizal species were detected at least once in the Acer-only plots. Despite this 310
relatively high frequency, the number of ectomycorrhizal fungal sequences per species 311
was 1.5 orders of magnitude lower than plots containing ectomycorrhizal host trees 312
(Acer-only zero-excluded mean = 23 reads, ectomycorrhizal host zero-excluded mean = 313
409 reads). Given the differences in sequence abundance, we believe the ectomycorrhizal 314
fungal sequences present in the Acer-only plots were not likely due to living mycelia, but 315
rather the product of ectomycorrhizal fungal spores and/or tag switching (in a separate 316
study looking at mushrooms at the Cloquet IDENT site, we noticed almost no fruiting 317
between plots, suggesting root growth between plots at the time of sampling was not a 318
significant issue). Because ectomycorrhizal spores/tag switching would also be present in 319
plots with ectomycorrhizal host vegetation, determining the number of sequence reads 320
that indicate an ectomycorrhizal fungal species is actively growing in the plot rather than 321
simply present as spores or via tag switching is essential for accurately assessing plant-322
associated patterns of fungal richness (also true for saprotrophic fungi as well). 323
Specifically, the inclusion of these false positives would lead to overestimates of plant 324
association for fungal species (i.e. an absence of specificity). Since our DNA-based 325
approach did not allow us to molecularly parse these two source pools (see Van der Linde 326
and Haller (2013) for an RNA-based alternative regarding spores), we employed a 327
simulation approach instead. 328
329
Page 14 of 52Molecular Ecology
For Review Only
15
We examined the ectomycorrhizal sequence abundance data from the 40 single-species 330
ectomycorrhizal host plots. Looking at the sequence abundances of known 331
ectomycorrhizal fungal specialists in our dataset (i.e. species only associated with certain 332
host genera (Molina et al. 1992)), values above 21 sequences per sample appeared to 333
clearly represent active growth rather than presence as spores or switched tags (i.e. in the 334
majority of the samples where specialist taxa were present with their specific hosts, the 335
number of sequence reads was at or well above 21 sequences). In a small number of 336
cases, however, values below 21 also looked biologically real (i.e. a specialist was 337
present in only the plots it should have been, but with an abundance below 21). As such, 338
we simulated the number of plots each ectomycorrhizal fungal species was present in 339
using a rolling cut-off that increased from 1 to 21 in two sequence increments (Fig. S2, R 340
code available upon request). Comparing the distributions of the 10 increment steps, we 341
found that when samples with 1 and 3 sequence reads were included, the number of plots 342
that ectomycorrhizal fungal species were present in was significantly different from the 343
majority of the other cut-offs. Since 5 sequences was the lowest cut-off that was not 344
significantly different from 21 sequences in our simulation, we decided to use that value 345
as a conservative threshold for taxon inclusion in the species richness analyses (this value 346
also matches the recommendations of Lindahl et al. (2013)). Because saprotrophic fungi 347
have similar spore dispersal strategies (i.e. many produce mushrooms like 348
ectomycorrhizal fungi) and would likely have the same issues with tag switching, we 349
applied the same 5 sequence read cut-off to the richness analyses for that dataset as well. 350
351
Page 15 of 52 Molecular Ecology
For Review Only
16
To assess the effects of different continuous biotic and abiotic factors on ectomycorrhizal 352
and saprotrophic fungal species richness, we used a linear modeling approach. Due to 353
shortcomings associated stepwise model selection (Whittingham et al. 2004), we used 354
information theory model selection in the ‘MuMIn’ package in R (R Core Team 2014). 355
The models assessing ectomycorrhizal and saprotrophic fungal richness were identical in 356
terms of type and number of predictor variables, but differed in replication due to the lack 357
of ectomycorrhizal associations with Acer trees. As a result, we did not include the Acer-358
only plots in the ectomycorrhizal model and when counting plant species richness, for 359
example, the number of species was 4 in the ectomycorrhizal fungal model (excluding 360
both Acer species) and 6 in the saprotrophic fungal model (including both Acer species) 361
in the 6 angiosperm treatment plots. One Quercus rubra plot in which soil moisture was a 362
significant outlier from all other plots was also excluded. 363
364
For both the ectomycorrhizal and saprotrophic richness models, we included 9 predictor 365
variables: plant species richness, plant phylogenetic diversity, plant standing biomass, 366
plant leaf area index (LAI), plant specific root length, plant leaf nitrogen, soil water 367
content, soil nitrogen, and soil pH (see Table 1 and above for calculations/units). Due to 368
differences in the ranges of each variable, all were first standardized (using the ‘scale’ 369
function). We then used the lm function to fit a basic linear model for each response 370
variable (i.e. ectomycorrhizal or saprotrophic richness) and determined that there were no 371
significant issues with multi-collinearity among variables using the ‘fmsbpackage (all 372
VIC scores < 2.4). Using the ‘dredge’ function in MuMIn, we then generated a full set of 373
models for each response variable (ectomycorrhizal or saprotrophic richness) that varied 374
Page 16 of 52Molecular Ecology
For Review Only
17
in inclusion of the nine predictor variables (Tables S3, S4). Those models were then 375
ranked and all those within four AIC
c
units of the best fitting model (i.e. with the lowest 376
AIC
c
score) were further compared using model averaging. (We also assessed those 377
within a 95% confidence interval of the best fitting model and obtained the same results 378
(Tables S5, S6)). The beta coefficients, standard errors, confidence intervals, 379
significance, and relative importance values reported in Table 1 are based on the top 38 380
and 12 models for ectomycorrhizal and saprotrophic richness, respectively (additional 381
details on each of the models with the lowest AICc score for each fungal guild are 382
presented in Tables S7, S8). We also explored the effect of adding interaction terms to 383
each model, but none significantly improved the global AIC
c
scores, so were not 384
included. 385
386
To assess the effect of the three categorical variables on ectomycorrhizal and 387
saprotrophic fungal richness, we also examined sampling block (A, B, C, & D), host 388
origin (Europe & America), and host phylum (Gymnosperm, Angiosperm, Both). We 389
used a three-way ANOVA for each fungal group, with no interaction terms (due to loss of 390
degrees of freedom). For consistency between the continuous and categorical variable 391
analyses, the one Quercus rubra plot in which soil moisture was a significant outlier from 392
all other plots remained excluded. Both ANOVAs were confirmed to meet model 393
assumptions by checking residual and Q-Q plots. Differences among of treatment means 394
for significant variables were determined using Tukey HSD tests. 395
396
Page 17 of 52 Molecular Ecology
For Review Only
18
We also assessed the effects of plant species composition, plant phylum, and plant origin 397
on changes in fungal species composition for both guilds. We first compared the 398
relationship between plant and fungal species composition for ectomycorrhizal and 399
saprotrophic fungi using Mantel tests. These analyses were run using both presence-400
absence and abundance-based data. Since the results were functionally equivalent (data 401
not shown), only the abundance-based data is presented. Prior to the Mantel tests, the 402
fungal data were Hellinger transformed and dissimilarities were calculated using the 403
Bray-Curtis metric. We additionally assessed the effect of changes in plant species 404
composition using the Raup-Crick metric (Chase et al. 2011), which better accounts for 405
the changes in alpha diversity across our experimental design (i.e. 1 vs. 4, 5, or 6 plant 406
species in the ectomycorrhizal and saprotrophic fungal analyses). The effect of plant 407
phylum and plant origin on both ectomycorrhizal fungal and saprotrophic fungal species 408
composition were compared using PERMANOVA using complete randomization. To 409
determine whether significant differences were based on centroid location or dispersion, 410
we used the ‘betadisper’ function. Variation in composition among plots was visualized 411
for each guild (ectomycorrhizal vs. saprotroph) and variable (plant phylum vs. plant 412
origin) using non-metric multidimensional scaling (NMDS) on Bray-Curtis dissimilarity 413
matrices. All assessments of the relationships between plant and fungal composition were 414
run in the Vegan package in R. 415
416
Results 417
From the total of 3,508,787 sequences remaining after bioinformatics filtering and quality 418
control, 2,842,254 (81%) belonged to ectomycorrhizal fungal species and 666,533 (19%) 419
Page 18 of 52Molecular Ecology
For Review Only
19
belonged to saprotrophic fungal species. Across the two guilds, we detected a total of 457 420
fungal species in the final rarefied dataset; 147 were ectomycorrhizal and 310 were 421
saprotrophic. The average number of sequences per plot was 15,107 for ectomycorrhizal 422
and 6,415 saprotrophic fungal species, respectively. Species accumulation curves showed 423
that plots were well saturated for both guilds (Fig. S3), with an average richness of 18 424
ectomycorrhizal fungal species and 38 saprotrophic fungal species. 425
426
In the statistical models assessing the effects of continuous biotic and abiotic factors on 427
fungal richness, the outcomes differed by guild (Table 1). Ectomycorrhizal fungal 428
richness was significantly positively associated with host phylogenetic diversity, but not 429
with any other factors (Fig. S4). Host phylogenetic diversity dominated the lowest-AIC 430
scoring models, having a relative importance value of 93% (the next closest was host LAI 431
at 48%). Unlike ectomycorrhizal fungal richness, saprotrophic fungal richness was 432
significantly positively influenced by plant leaf nitrogen content and negatively 433
influenced by plant standing biomass and plant specific root length (Fig. S5). Among 434
these variables, plant leaf nitrogen content, specific root length, and standing biomass all 435
had relative importance values of 100% in the lowest-AIC scoring models. 436
437
In the statistical models assessing the effects of plant phylum, plant origin, and block on 438
fungal richness, results were consistent between fungal guilds except for plant phylum. 439
Neither block nor plant origin had a significant effect on ectomycorrhizal or saprotrophic 440
fungal richness (Fig. S6). In contrast, plant phylum had a significant effect on 441
ectomycorrhizal fungal richness (F
2,47
= 92.29, P < 0.004), but not saprotrophic fungal 442
Page 19 of 52 Molecular Ecology
For Review Only
20
richness (Fig. 1). Species richness did not differ between the gymnosperm-only and 443
angiosperm-only plots for ectomycorrhizal fungi, but was significantly higher in the plots 444
containing both gymnosperm and angiosperm host trees (Tukey HSD test, P < 0.05). 445
Although the total number of ectomycorrhizal fungal species present in plots with both 446
host types was higher, there were no ectomycorrhizal species unique to those plots (Table 447
2). Instead, many of the species present in either the gymnosperm-only or the 448
angiosperm-only plots were also present in the mixed host plots. 449
450
Comparing fungal species composition across plots, Bray-Curtis dissimilarities among 451
both the ectomycorrhizal and saprotrophic fungal guilds were significantly correlated 452
with Bray-Curtis as well as Raup-Crick dissimilarities in plant species composition 453
(Ectomycorrhizal fungal Bray-Curtis/Plant Bray-Curtis Mantel test: r = 0.409, P <0.001; 454
Ectomycorrhizal fungal Bray-Curtis/Plant Raup-Crick Mantel test: r = 0.423, P <0.001; 455
Saprotrophic fungal Bray-Curtis/Plant Bray-Curtis Mantel test: r = 0.123, P <0.001; 456
Saprotrophic fungal Bray-Curtis/Plant Raup-Crick Mantel test: r = 0.13, P <0.001). In 457
both ectomycorrhizal and saprotrophic fungal guilds, fungal species composition was 458
more similar in plots containing more similar plant species, however, the amount of 459
change in ectomycorrhizal fungal species composition per change in plant species 460
composition was notably higher than for saprotrophic fungi (Fig. 2). The two guilds also 461
differed in the response of species composition to both plant origin and plant phylum. 462
Ectomycorrhizal fungal species composition differed significantly by plant phylum 463
(PERMANOVA: F
2,50
= 4.82, P < 0.001, r
2
= 0.153), with the gymnosperm and 464
angiosperm plots showing the greatest divergence in composition and plots containing 465
Page 20 of 52Molecular Ecology
For Review Only
21
both host types being intermediate (Fig. 3). Betadisper analysis of host phylum effects 466
were not significant (ANOVA: F
2,52
= 2.452, P = 0.096), indicating this was not due to 467
differences in dispersion among groups. Ectomycorrhizal fungal species composition also 468
differed significantly by plant origin (PERMANOVA: F
2,50
= 1.69, P = 0.021, r
2
= 0.053), 469
with plots containing for American and European plant species showing the most 470
divergence and plots containing both being intermediate. However, the significant 471
betadisper analysis (ANOVA: F
2,52
= 4.805, P = 0.013) suggested the significant 472
PERMANOVA difference was caused by the greater dispersion in the plots with only 473
European or American plant species compared to those with both. In contrast, 474
saprotrophic fungal species composition did not respond significantly to either variable 475
(Plant phylum PERMANOVA: F
2,58
= 0.91, P = 0.646, r
2
= 0.029, Plant origin 476
PERMANOVA: F
2,58
= 1.16, P = 0.197, r
2
= 0.038). 477
478
Discussion 479
Our results are the first to experimentally validate that the significant positive 480
relationships between plant phylogenetic diversity and ectomycorrhizal fungal species 481
richness and composition are not confounded by sampling effects (Ishida et al. 2007, 482
Tedersoo et al. 2012, Tedersoo et al. 2013, Gao et al. 2013, Tedersoo et al. 2014a, Gao et 483
al. 2014). The absence of a similar response for soil saprotrophic fungi supports the 484
growing consensus that the nature of the relationship between plant diversity and soil 485
fungal richness depends significantly on fungal guild (Peay et al. 2013, Tedersoo et al. 486
2014a, Tedersoo et a. 2015). Interestingly, we found no significant effect of increasing 487
plant species richness on fungal species richness for either ectomycorrhizal or 488
Page 21 of 52 Molecular Ecology
For Review Only
22
saprotrophic fungi. While these results are consistent with the findings of Prober et al. 489
(2014), other studies have found a significantly positive relationship between plant 490
richness and either mycorrhizal or saprotrophic fungal guilds (Hiiesalu et al. 2013, Peay 491
et al. 2013, Pellissier et al. 2014, Tedersoo et al. 2015). While many factors may 492
contribute the difference between our results and those in other studies (e.g. differences 493
in regional species pools, examination of different aged hosts, focusing on only on soil 494
rather than root fungal communities), one that we think is particularly notable is the 495
relatively low richness of plants present in the Cloquet IDENT plots. The maximum plot 496
plant richness at the study site was just six species, which is an order of magnitude lower 497
than all but one of the aforementioned studies. This difference suggests the effect of plant 498
species richness on soil fungal richness may be relatively small and that studies 499
incorporating larger plant species gradients (>20 species) are needed to clearly observe 500
the cumulative nature of this relationship. 501
502
In addition to demonstrating a link between plant phylogenetic diversity and 503
ectomycorrhizal fungal richness and composition, our results may also shed light on the 504
mechanism by which plant and ectomycorrhizal fungal richness are connected. It appears 505
the influence of plant phylogenetic diversity on ectomycorrhizal fungal richness is driven 506
by host specificity rather than any chemical or physiological synergy among different 507
host trees. Had the co-presence of gymnosperms and angiosperms uniquely facilitated the 508
growth of certain ectomycorrhizal fungal species, we would have expected to see a set of 509
species only present in the mixed-host plots. However, we found no ectomycorrhizal 510
fungal species unique to the 6 North American and 6 European plots (which contained 511
Page 22 of 52Molecular Ecology
For Review Only
23
both gymnosperm and angiosperm hosts) (Table 2). Instead, we found many of the 512
ectomycorrhizal fungal species present in gymnosperm-only plots but not angiosperm-513
only plots (and vice versa) were also present in the mixed-host plots. While we did not 514
sample roots and therefore cannot by certain of the presence of functional mycorrhizas 515
between specialist species and their known hosts in mixed-host plots, this latter pattern is 516
consistent with a specificity effect, i.e. only in the presence of particular hosts are certain 517
ectomycorrhizal fungal species also present. Examining the identity of species further 518
confirmed this pattern, with known ectomycorrhizal fungal host specialists such as 519
Suillus spp. and Rhizopogon roseolus on gymnosperm hosts and Leccinum spp. on 520
angiosperm hosts only occurring in plots where particular hosts were present. While the 521
specificity patterns were strong at inter-phylum level (gymnosperm vs. angiosperm), we 522
found little evidence that intra-phylum (within gymnosperm or within angiosperm) 523
specificity was a significant factor driving ectomycorrhizal fungal richness patterns. 524
Specifically, we did not see a significant increase in ectomycorrhizal fungal species 525
richness in the 6 gymnosperm or 6 angiosperm plots, compared to the monoculture plots 526
of each type, respectively (Fig. S7). Although our power to test intra-phylum effects was 527
weaker than those between phyla, these results suggest ectomycorrhizal fungal specificity 528
effects may be most consistently apparent at higher (e.g. order and phylum) than lower 529
(e.g. species and genus) plant taxonomic levels (Ishida et al. 2007, Tedersoo et al. 2013, 530
Gao et al. 2013). 531
532
Despite being more than twice as species rich as the ectomycorrhizal fungal 533
communities, the saprotrophic fungal communities had less than half the number of 534
Page 23 of 52 Molecular Ecology
For Review Only
24
sequences in the plots. This suggests that ectomycorrhizal fungi dominated the soil fungal 535
communities in the Cloquet IDENT plots, which is not surprising given the high host 536
density present. The highly guild-dependent results that we observed in fungal species 537
composition parallel the results of Peay et al. (2013), who found that symbiotic fungal 538
community richness responded more strongly to plant community changes than that of 539
saprotrophic fungi. While differential responses between these fungal guilds seems likely 540
given to the varying nature of their carbon sources, we believe the nature of our study site 541
may also have been an important contributing factor. The Cloquet IDENT site that we 542
sampled is a relatively young forest setting (5 years old at the time of sampling) and, 543
although some of the trees were well over 2 m tall, there has been limited soil organic 544
matter accumulation in the plots (the soil cores were also sieved, so larger coarse organic 545
material that may have harbored more saprotrophic fungi was also removed). In older 546
forest sites in Europe, both Urbanova et al. (2015) and Tedersoo et al. (2015) found that 547
the richness and composition of soil fungal communities was strongly coupled to 548
presence of particular host trees. Their findings suggest that rather than soil saprotrophic 549
fungal communities being unaffected by plant phylogenetic diversity, greater differences 550
in the saprotrophic fungal community are likely to develop with time, as the differential 551
effects of host on litter chemistry and quality accumulate. Along with differences in 552
responses related to plant composition, the saprotrophic fungal communities in our study 553
also had greater overall homogeneity (i.e. consistently lower dissimilarity in all plots) 554
than the ectomycorrhizal fungal communities. Although the reasons for this pattern are 555
not clear, they do not appear to be related to differences in species richness between the 556
Page 24 of 52Molecular Ecology
For Review Only
25
two fungal guilds, as Peay et al. (2013) found a similar pattern between mycorrhizal and 557
saprotrophic fungi despite having nearly identical species richness. 558
559
The positive correlation between saprotrophic fungal richness and host leaf nitrogen can 560
be interpreted in multiple ways. It is possible that more nitrogen-rich litter may support a 561
greater number of saprotrophic fungal species by decreasing competition for this 562
resource. However, it has been widely shown that soil saprotrophic fungi are more carbon 563
than nitrogen limited (Dighton 2003), which does not support this interpretation. Reich 564
(2012) showed that leaf nitrogen content is strongly correlated with site productivity in 565
temperate and boreal forests, likely due to the nitrogen present in photosynthetic 566
enzymes. Given this pattern, it is also possible that the increase in saprotrophic fungal 567
richness with leaf nitrogen content that we observed may be due to greater carbon inputs, 568
either through greater rhizosphere C deposition or increased leaf litter (Broekling et al. 569
2009). The negative correlation between soil saprotrophic richness and specific root 570
length was surprising, given that leaf nitrogen content and specific root length were 571
themselves positively correlated (Fig. S8). While we did not collect any data on fungal 572
biomass in these plots, it is possible that greater specific root length may have favored the 573
growth of ectomycorrhizal fungal species by providing additional root tips for 574
colonization. Since previous work has shown that ectomycorrhizal fungi can suppress the 575
growth of saprotrophic fungi (Gadgil and Gadgil 1975, but see Shaw et al. 1995), greater 576
abundances of ectomycorrhizal fungi may be responsible for the negative relationship 577
between specific root length and saprotrophic fungal richness. Clearly, additional studies 578
Page 25 of 52 Molecular Ecology
For Review Only
26
are needed to better understand the mechanisms driving saprotrophic fungal richness at 579
local spatial scales. 580
581
Plant origin did not have strong effects on ectomycorrhizal fungal species richness, but it 582
did significantly affect ectomycorrhizal fungal species composition. While this difference 583
appeared to largely associated by less variation in species composition across plots with 584
both American and European hosts compared to those with only American or only 585
European hosts, we believe a close inspection of members of the genera Rhizopogon, 586
Suillus, and Leccinum provides additional insight. These genera contain a number of 587
species with known geographic distributions, both in North America and Europe. For 588
example, the species Suillus grevillei is known to occur on Larix hosts on both continents 589
and was present in similar abundance in the L. laricina and L. decidua plots. Two other 590
Suillus species, S. pictus and S. intermedius, are known to occur only in North America 591
and were both only present in plots containing the North American host species, Pinus 592
strobus. In contrast, Rhizopogon roseolus, a species of known European origin, was only 593
present in plots containing the European host species, Pinus sylvestris. The two Leccinum 594
species also showed similar differences, with the European species L. schistophilum 595
being only present on the European Betula pendula, and the other species, L. scabrum, 596
while known from both Europe and North America (Den Bakker et al. 2007), being only 597
present on North American Betula species, B. papyifera. Collectively, these plot 598
occurrence patterns suggest that specificity due to host origin may also contribute the 599
observed divergences in ectomycorrhizal fungal species composition across plots. At the 600
same time, because the congeneric species occurred with different hosts, this did not 601
Page 26 of 52Molecular Ecology
For Review Only
27
strongly affect plot-level ectomycorrhizal fungal richness (which is consistent with the 602
absence of a significant host origin effect on ectomycorrhizal richness). Although more 603
sampling in studies employing similar designs is needed to determine the strength and 604
consistency of these patterns, our results suggest that host origin along with phylogenetic 605
diversity can have important effects of ectomycorrhizal fungal community structure. 606
607
Conclusions 608
Taken together, our results provide an important experimental test of the nature of plant 609
and soil fungal community relationships. Our findings that host plant phylogenetic 610
diversity is a significant driver of ectomycorrhizal fungal richness and composition is 611
consistent with other observation-based studies and suggests that the global peak in 612
ectomycorrhizal fungal richness at temperate latitudes may be driven by the parallel 613
increase in host plant phylogenetic diversity (Tedersoo et al. 2012, 2014). As other 614
important variables related to ectomycorrhizal fungal richness (e.g. soil organic matter 615
content and host density) are also high in temperate zones, future experimental tests will 616
be important in further refining our understanding of the global drivers of 617
ectomycorrhizal fungal richness (Kennedy et al. 2012). The different responses of soil 618
saprotrophic fungi suggest that while their community richness and composition is 619
influenced by plant community structure, other ecological factors are more influential on 620
non-symbiotic fungal communities. Similar differences between guilds have been seen in 621
studies at larger spatial scales (Peay et al. 2013, Tedersoo et al. 2014a, but see Prober et 622
al. 2014), suggesting the patterns we observed in this local-scale study is one that is 623
robust. Fortunately, the pairing of high-throughput sequencing, which captures all fungi 624
Page 27 of 52 Molecular Ecology
For Review Only
28
present in samples, and the better automation of parsing fungi by ecological guild 625
(Nguyen et al., in press) will greatly facilitate the parallel study of symbiotic and free-626
living fungal communities. This will allow researchers to more fully understand the 627
important ecological similarities and differences among fungal guilds, which will 628
ultimately help in better understanding their effects on ecosystem diversity and 629
functioning. 630
631
Acknowledgements 632
The authors gratefully thank the many field assistants for the set-up and maintenance of 633
the Cloquet IDENT plots. Helpful advice about analytical approaches was provided by 634
W. Pearse, J. Grossman, and A. David. Constructive comments on previous drafts were 635
kindly provided by W. Harcombe, F. Isbell, J. Schilling, members of the UMN Mycology 636
Club reading group, and three anonymous reviewers. 637
Page 28 of 52Molecular Ecology
For Review Only
29
References 638
Antoninka A, Reich PB, Johnson NC (2011) Seven years of carbon dioxide enrichment, 639
nitrogen fertilization and plant diversity influence arbuscular mycorrhizal fungi in a 640
grassland ecosystem. New Phytologist, 192, 200–214. 641
den Bakker HC, Zuccarello GC, Kuyper TW, Noordeloos ME (2007) Phylogeographic 642
patterns in Leccinum sect. Scabra and the status of the arctic-alpine species L. 643
rotundifoliae. Mycological Research, 111, 663–672. 644
Bardgett RD, van der Putten WH (2014) Belowground biodiversity and ecosystem 645
functioning. Nature, 515, 505–511. 646
Broeckling CD, Broz AK, Bergelson J, Manter DK, Vivanco JM (2008) Root exudates 647
regulate soil fungal community composition and diversity. Applied and 648
Environmental Microbiology, 74, 738–744. 649
Bryant JA, Lamanna C, Morlon H et al. (2008) Microbes on mountainsides: contrasting 650
elevational patterns of bacterial and plant diversity. PNAS, 105 Suppl, 11505–11. 651
Burrows RL, Pfleger FL (2002) Arbuscular mycorrhizal fungi respond to increasing plant 652
diversity. Canadian Journal of Botany, 80, 120–130. 653
Caporaso JG, Lauber CL, Walters WA et al. (2011) Global patterns of 16S rRNA 654
diversity at a depth of millions of sequences per sample. PNAS, 108, 4516–4522. 655
Carlsen T, Aas AB, Lindner D et al. (2012) Don’t make a mista(g)ke: Is tag switching an 656
overlooked source of error in amplicon pyrosequencing studies? Fungal Ecology, 5, 657
747–749. 658
Chase JM, Kraft NJB, Smith KG, et al. (2011) Using null models to disentangle variation 659
in community dissimiliarity from variation in alpha-diversity. Ecosphere 2, art24. 660
Page 29 of 52 Molecular Ecology
For Review Only
30
Chave J, Coomes D, Jansen S et al. (2009) Towards a worldwide wood economics 661
spectrum. Ecology Letters, 12, 351–366. 662
De Deyn GB, Van Der Putten WH (2005) Linking aboveground and belowground 663
diversity. Trends in Ecology and Evolution, 20, 625–633. 664
Eisenhauer N, Milcu A, Sabais ACW et al. (2011) Plant diversity surpasses plant 665
functional groups and plant productivity as driver of soil biota in the long term. 666
PLoS ONE, 6. 667
Fierer N, Jackson RB (2006) The diversity and biogeography of soil bacterial 668
communities. PNAS, 103, 626–31. 669
Gadgil RL, Gadgil PD (1975) Suppression of litter decomposition by mycorrhizal roots 670
of Pinus radiata. New Zealand Journal of Forestry Science, 5, 33–41. 671
Gao C, Shi N-N, Liu Y-X et al. (2013) Host plant genus-level diversity is the best 672
predictor of ectomycorrhizal fungal diversity in a Chinese subtropical forest. 673
Molecular Ecology, 22, 3403–3414. 674
Gao C, Shi N-N, Liu Y-X et al. (2014) Host plant richness explain diversity of 675
ectomycorrhizal fungi: Response to the comment of Tedersoo et al. (2014). 676
Molecular Ecology, 23, 996–999. 677
Haddad NM, Crutsinger GM, Gross K et al. (2009) Plant species loss decreases arthropod 678
diversity and shifts trophic structure. Ecology Letters, 12, 1029–1039. 679
van der Heijden MGA, Bardgett RD, Straalen NM Van (2008) The unseen majority: soil 680
microbes as drivers of plant diversity and productivity in terrestrial ecosystems. 681
Ecology Letters, 11, 296–310. 682
Hiiesalu I, Pärtel M, Davison J et al. (2014) Species richness of arbuscular mycorrhizal 683
Page 30 of 52Molecular Ecology
For Review Only
31
fungi: Associations with grassland plant richness and biomass. New Phytologist, 684
203, 233–244. 685
Hooper DU, Chapin FS, Ewel JJ et al. (2005) Effects of biodiversity on ecosystem 686
functioning: A consensus of current knowledge. Ecological Monographs, 75, 3–35. 687
Huston MA (1997) Hidden treatments in ecological experiments: re-evalutating the 688
ecosystem function of biodiverstiy. Oecologia, 110, 449–460. 689
Ishida TA, Nara K, Hogetsu T (2007) Host effects on ectomycorrhizal fungal 690
communities: insight from eight host species in mixed conifer-broadleaf forests. 691
New Phytologist, 174, 430–440. 692
Ives AR, Helmus MR (2012) Phylogenetic Metrics of Community Similarity. American 693
Naturalist, 176. 694
Kennedy PG, Matheny PB, Ryberg KM et al. (2012) Scaling up: examining the 695
macroecology of ectomycorrhizal fungi. Molecular Ecology, 21, 4151–4154. 696
Kõljalg U, Nilsson RH, Abarenkov K et al. (2013) Towards a unified paradigm for 697
sequence-based identification of fungi. Molecular Ecology, 22, 5271–7. 698
Landis FC, Gargas A, Givnish TJ (2004) Relationships among arbuscular mycorrhizal 699
fungi, vascular plants and environmental conditions in oak savannas. New 700
Phytologist, 164, 493–504. 701
Lauber CL, Hamady M, Knight R, Fierer N (2009) Pyrosequencing-Based Assessment of 702
Soil pH as a Predictor of Soil Bacterial Community Structure at the Continental 703
Scale. Applied and Environmental Microbiology, 75, 5111–5120. 704
LeBlanc N, Kinkel LL, Kistler HC (2015) Soil Fungal Communities Respond to 705
Grassland Plant Community Richness and Soil Edaphics. Microbial Ecology, 70, 706
Page 31 of 52 Molecular Ecology
For Review Only
32
188–195. 707
Lindahl BD, Nilsson RH, Tedersoo L et al. (2013) Fungal community analysis by high-708
throughput sequencing of amplified markers – a user’s guide. New Phytologist, 199, 709
288–299. 710
Van der Linde S, Haller S (2013) Obtaining a spore free fungal community composition. 711
Fungal Ecology, 6, 522–526. 712
McMurdie PJ, Holmes S (2014) Waste not, want not: why rarefying microbiome data is 713
inadmissible. PLoS Computational Biology, 10, e1003531. 714
Molina R, Massicotte HB, Trappe JM (1992) Specificity phenomena in mycorrhizal 715
symbioses: community-ecological consequences and practical implications. In: Mycorrhizal 716
Functioning: an Integrative Plant-Fungal Process (ed. M.F. Allen). Chapman & Hall, New 717
York. 718
Nara K, Hogetsu T (2004) Ectomycorrhizal fungi on established shrubs facilitate 719
subsequent seedling establishment of successional plant species. Ecology, 85, 1700–720
1707. 721
Nguyen NH, Smith DP, Peay KG, Kennedy PG (2015) Parsing ecological signal from 722
noise in next generation amplicon sequencing. New Phytologist, 205, 1389–1993. 723
Nguyen N, Song Z, Bates S et al. FUNGuild: an open annotation tool for parsing fungal 724
community datasets by ecological guild. Fungal Ecology, in press. 725
Pearse W, Cavender-Bares J, Helmus MR (2014) Metrics and Models of Community 726
Phylogenetics. In: Modern phylogenetic comparative methods and their application 727
in evolutionary biology: concepts and practice (ed Garamszegi L). Springer. 728
Peay KG, Baraloto C, Fine PV a (2013) Strong coupling of plant and fungal community 729
structure across western Amazonian rainforests. ISME Journal, 7, 1852–61. 730
Page 32 of 52Molecular Ecology
For Review Only
33
Peay KG, Kennedy PG, Bruns TD (2008) Fungal community ecology: A hybrid beast 731
with a molecular master. Bioscience, 58, 799–810. 732
Pellisier, L, Niculita-Hirzel H, Dubuis A et al. (2014) Soil fungal communities of 733
grasslands are spatially structured at a regional scale in the Alps. Molecular Ecology 734
23, 4274-4290. 735
Pregitzer KS, DeForest JL, Burton AJ et al. (2002) Fine root architecture of nine North 736
American trees. Ecological Monographs, 72, 293–309. 737
Prober SM, Leff JW, Bates ST et al. (2014) Plant diversity predicts beta but not alpha 738
diversity of soil microbes across grasslands worldwide. Ecology Letters, 18, 85–95. 739
Putten WH Van der, Dijk C Van, Peters Barbuscular mycorrhizal (1993) Plant-specific 740
soil-borne diseases contribute to succession in foredune vegetation. Nature, 741
362, 53–56. 742
Reich PB (2012) Key canopy traits drive forest productivity. Proceedings of the Royal 743
Society B: Biological Sciences, 279, 2128–2134. 744
Schloss PD, Westcott SL, Ryabin T et al. (2009) Introducing mothur: Open-Source, 745
Platform-Independent, Community-Supported Software for Describing and 746
Comparing Microbial Communities. Applied and Environmental Microbiology, 75, 747
7537–7541. 748
Shaw TM, Dighton J, Sanders FE (1995) Interactions between ectomycorrhizal and 749
saprotrophic fungi on agar and in association with seedlings of lodgepole pine 750
(Pinus contorta). Mycological Research, 99, 159–165. 751
Smith DP, Peay KG (2014) Sequence depth, not PCR replication, improves ecological 752
inference from next generation DNA sequencing. PloS ONE, 9, e90234. 753
Page 33 of 52 Molecular Ecology
For Review Only
34
Smith LM, Reynolds HL (2015) Plant-soil feedback shifts from negative to positive with 754
decreasing light in forest understory species. Ecology, 96, 2523–2532. 755
Song Z, Schlatter D, Kennedy P et al. (2015) Effort versus Reward: Preparing Samples 756
for Fungal Community Characterization in High-Throughput Sequencing Surveys of 757
Soils. Plos One, 10, e0127234. 758
Tedersoo L, Bahram M, Cajthaml T et al. (2015) Tree diversity and species identity 759
effects on soil fungi, protists and animals are context dependent. The ISME Journal, 760
1–17. 761
Tedersoo L, Bahram M, Põlme S et al. (2014a) Global diversity and geography of soil 762
fungi. Science, 346, 1–10. 763
Tedersoo L, Bahram M, Dickie IA (2014b) Does host plant richness explain diversity of 764
ectomycorrhizal fungi? Re-evaluation of Gao et al. (2013) data sets reveals sampling 765
effects. Molecular Ecology, 23, 992–995. 766
Tedersoo L, Bahram M, Toots M et al. (2012) Towards global patterns in the diversity 767
and community structure of ectomycorrhizal fungi. Molecular Ecology, 21, 4160–768
4170. 769
Tedersoo L, Mett M, Ishida TA, Bahram M (2013) Phylogenetic relationships among 770
host plants explain differences in fungal species richness and community 771
composition in ectomycorrhizal symbiosis. New Phytologist, 199, 822–831. 772
Tobner CM, Paquette A, Reich PB, Gravel D, Messier C (2014) Advancing biodiversity-773
ecosystem functioning science using high-density tree-based experiments over 774
functional diversity gradients. Oecologia, 174, 609–21. 775
Urbanová M, Šnajdr J, Baldrian P (2015) Composition of fungal and bacterial 776
Page 34 of 52Molecular Ecology
For Review Only
35
communities in forest litter and soil is largely determined by dominant trees. Soil 777
Biology and Biochemistry, 84, 53–64. 778
Van der Heijden, MGA, Bardgett RD, Van Straalen NM (2008). The unseen majority: 779
soil microbes as drivers of plant diversity and productivity in terrestrial systems. 780
Ecology Letters 11, 296-310. 781
Vogelsang KM, Reynolds HL, Bever JD (2006) Mycorrhizal fungal identity and richness 782
determine the diversity and productivity of a tallgrass prairie system. New 783
Phytologist, 172, 554–562. 784
Wagg C, Bender SF, Widmer F, van der Heijden MGA (2014) Soil biodiversity and soil 785
community composition determine ecosystem multifunctionality. PNAS, 111, 5266–786
70. 787
Wardle DA (2006) The influence of biotic interactions on soil biodiversity. Ecology 788
Letters, 9, 870–886. 789
Whittingham MJ, Stephens P a., Bradbury RB, Freckleton RP (2006) Why do we still use 790
stepwise modelling in ecology and behaviour? Journal of Animal Ecology, 75, 791
1182–1189. 792
793
Data Accessibility 794
All DNA sequence files: MG-RAST Project #: 13302 795
Sample x OTU Matrices: Supplementary Tables S1 and S2. 796
Page 35 of 52 Molecular Ecology
For Review Only
Host Phylum
ECM Fungal Richness
BAB
A
Plant Phylum
SAP Fungal Richness
Figure 1. Ectomycorrhizal (ECM) and saprotrophic (SAP) fungal species richness in plots containing gymnosperm plants
only (Gymno), angiosperm plants only (Angio), or both types of plants (Both) in the Cloquet IDENT experiment. Host and
plant phylum differ only in the exclusion of the Acer-only plots in the ECM fungal analyses. Boxes surrounding median
richness values represent the first and third quartiles, whiskers show the maximum and minimum values excluding outliers
(which are represented by dots). Phylum groups not sharing the same letter are significantly different based on a Tukey HSD
test. ANOVA model r-square = 0.21.
Page 36 of 52Molecular Ecology
For Review Only
0.00
0.25
0.50
0.75
1.00
00.5 0.75 1
Host Dissimilarity (Bray-Curtis)
ECM Fungal Dissimilarity (Bray-Curtis)
0.00
0.25
0.50
0.75
1.00
00.5 0.75 1
Plant Dissimilarity (Bray-Curtis)
SAP Fungal Dissimilarity (Bray-Curtis)
Figure 2. Bray-Curtis dissimilarities between plant and ectomycorrhizal (ECM) and saprotrophic (SAP) fungal
communities in the Cloquet IDENT experiment. Values were plotted on axes ranging between zero and one, where
increasing values indicate greater dissimilarity between plot communities. Due to multiple overlapping
observations, points were plotted as partially transparent, with darkest reflecting data density. Density curves were
plotted to reflect the underlying distribution of fungal community dissimilarity at each level of plant community
dissimilarity (a.k.a. violin plots). Linear regression lines were included to indicate trends in the change in fungal
dissimilarity with changes in plant community dissimilarity.
Page 37 of 52 Molecular Ecology
For Review Only
Figure 3. Non-metric multi-dimensional scaling plots of ectomycorrhizal (ECM) and saprotrophic (SAP) fungal species
composition at the Cloquet IDENT experiment. Points in ordination space represent individual plots, with colors representing
plant phylum (Angiosperm, Gymnosperm, Both) or plant origin (America, Europe, Both). Standard deviation ellipses were
projected on the ordination plot to visualize multivariate dispersion in community composition by plant phylum and plant origin.
-0.5
0.0
0.5
-1.0 -0.5 0.0 0.5
MDS1
MDS2
group
Angio
Both
Gymno
Host Phylum: ECM Fungi
-0.5
0.0
0.5
-1.0 -0.5 0.0 0.5
MDS1
MDS2
group
America
Both
Europe
Host Origin: ECM Fungi
-0.25
0.00
0.25
0.50
-0.6 -0.3 0.0 0.3 0.6
MDS1
MDS2
group
Angio
Both
Gymno
Plant Phylum: SAP Fungi
-0.25
0.00
0.25
0.50
-0.6 -0.3 0.0 0.3 0.6
MDS1
MDS2
group
America
Both
Europe
Plant Origin: SAP Fungi
Page 38 of 52Molecular Ecology
For Review Only
Table 1. Effects of nine continuous abiotic and biotic variables on ECM and saprotrophic
fungal richness in the Cloquet IDENT experiment. MAE = Model-averaged coefficient
estimate ± unconditional standard error, CI = Confidence interval (2.5%,97.5%) P =
Significance (<0.05 in bold), RVI = Relative variable importance. Because of variable
standardization ahead of modeling, the model-averaged estimates are directly
comparable and their magnitude represents a proxy for relative effect size.
Ectomycorrhizal richness model r-square = 0.20, Saprotrophic richness model r-square
= 0.31.
Variable
Parameters (n=55)
ECM fungal richness
Parameters (n=63)
SAP fungal richness
Plant species richness
Range = 1 - 6
MAE = 0.45 ± 0.93
Range = 1 - 6
MAE = 1.36 ± 1.28
# = 4 different values
CI = -1.39, 2.30
# = 2 different values
CI = -1.20, 3.92
P = 0.629
P = 0.296
RVI = 0.11
RVI = 0.36
Plant phylogenetic diversity
Range = 0 - 0.65
MAE = 1.44 ± 0.62
Range = 0 - 0.65
MAE = -0.85 ± 1.50
# = 4 different values
CI = 0.19, 2.68
# = 2 different values
CI = -3.83, 2.14
P = 0.023
P = 0.579
RVI = 0.98
RVI = 0.15
Plant standing mass
Range = 997 - 23913
MAE = -1.13 ± 0.73
Range = 317 - 23913
MAE = -3.65 ± 1.38
# = 55 different values
CI = -2.58, 0.32
# = 64 different values
CI = -6.41, -0.88
P = 0.127
P = 0.009
RVI = 0.41
RVI = 1.00
Plant leaf area index
Range = 04.7 - 5.68
MAE = 1.18 ± 0.75
Range = 04.7 - 5.68
MAE = -0.09 ± 1.79
# = 50 different values
CI = -0.32, 2.69
# = 57 different values
CI = -3.67, 3.49
P = 0.123
P = 0.960
RVI = 0.48
RVI = 0.11
Plant leaf nitrogen
Range = 0.89 - 2.52
MAE = -0.51 ± 0.79
Range = 0.89 - 2.52
MAE = 6.69 ± 1.44
# = 24 different values
CI = -2.10, 1.35
# = 21 different values
CI = 3.81, 9.58
P = 0.526
P < 0.001
RVI = 0.12
RVI = 1.00
Plant specific root length
Range = 1.01 - 154.33
MAE = 0.99 ± 0.65
Range = 20.41 - 75.33
MAE = -5.21 ± 1.30
# = 24 different values
CI = -0.39,2.19
# = 27 different values
CI = -7.77, -2.65
P = 0.172
P < 0.001
RVI = 0.38
RVI = 1.00
Soil nitrogen
Range = 0.21 - 9.42
MAE = 0.17 ± 0.59
Range = 0.21 - 9.42
MAE = -0.14 ± 1.09
# = 49 different values
CI = -1.02, 1.35
# = 55 different values
CI = -2.32, 2.04
P = 0.784
P = 0.896
RVI = 0.06
RVI = 0.11
Soil water
Range = 8.5 - 17.2
MAE = 0.67 ± 0.61
Range = 8.5 - 22.3
MAE = 0.48 ± 1.19
# = 54 different values
CI = -0.56, 1.90
# = 61 different values
CI = -1.91, 2.87
P = 0.284
P = 0.692
RVI = 0.26
RVI = 0.12
Soil pH
Range = 4.67 - 6.04
MAE = -1.58 ± 2.13
Range = 4.67 - 6.04
MAE = -1.67 ± 3.89
# = 39 different values
CI = -5.85, 2.69
# = 55 different values
CI = -9.47, 6.11
P = 0.468
P = 0.673
RVI = 0.14
RVI = 0.11
!!
!!
!!
!!
!!
!!
!!
!!
!!
Variable
Plot Level Measure
!!
!!
!!
!!
Species richness
Host (ECM) or plant (SAP) species
Phylogenetic diversity
Helmus' phylogenetic species variability
Standing mass
Aboveground biomass (g) (DBH x height x wood density), October 2013
Leaf area index
Leaf (m^2)/ground (m^2), peak biomass August 2013
Leaf nitrogen
Dry mass (%), September 2013
Specific root length
Root length (m/g) based on single-species plots, June 2014
Soil nitrogen
[NH4 + NO3] mg/kg, June 2013
Soil water
Average volumetric content (%), June-August 2013
Soil pH
Water-based pH, October 2013
Page 39 of 52 Molecular Ecology
For Review Only
Tab le 2 . Ectomycorrhizal (ECM) fungal species present in 6 angiosperm plots, 6 gymnosperm plots, or
mixed-host (i.e. both) plots in the Cloquet IDENT experiment.
Plot Type
Species
ECM Fungal Species Name Best Match
Gymno Only
3
Amphinema byssoides (SH229867), Amphinema sp. (SH229877), Suillus cavipes (SH190974)
Angio Only
8
Clavulina sp. (SH001392), Cortinarius sp. (SH010342), Laccaria sp. (SH205134), Lactarius sp. (SH003059)
Leccinum scabrum (SH197538), Leccinum schistophilum (SH197295), Tomentella sublilacina (SH004591),
Thelephora sp. (SH219948)
Gymno + Both
11
Amphinema byssoides (SH229867), Amphinema sp. (SH229877), Inocybe lanatodisca (SH215203),
Laccaria laccata (SH205145), Meliniomyces sp. (SH012500), Rhizopogon roseolus (SH001018),
Russula sp. (SH025246), Russula sp. (SH20839), Suillus sp. (SH000892), Suillus grevillei (SH19095),
Tomentella botryoides (SH202494), Tomentella sp. (SH009549)
Angio + Both
11
Cortinarius alborufescens (SH023149), Cortinarius argyrionus (SH191810), Cortinarius sp. (SH031729),
Cortinarius sp. (SH031772),Entoloma rhodopolium (SH038256), Inocybe sp. (SH200564), Laccaria sp.
(SH010997),
Laccaria laccata (SH205145), Meliniomyces sp. (SH207199), Tomentella sp. (SH219853), Tuber sp. (SH200481)
Both Only
0
Page 40 of 52Molecular Ecology
For Review Only
Block D 1" 2" 3" 4" 5" 6" 7" 8"
A ACSA-ACPL QURU-QURO ACPL-BEPE" PIGL-ACSA" BEPA-QURU" QURU" PIGL-PIST" BEPE"
B LALA" PISY-LADE" LADE-QURO" LADE" PIST-BEPA" ACSA-BEPA" BEPA" LALA-BEPE
C BEPE-QURO" PIGL-ACSA PIAB" BEPA-BEPE PIST" PIAB-PISY" PIAB-PIST PIST-LALA"
D 6 gymno ACPL-BEPA 6"NA" 6 angio LALA-QURU" LADE-QURU QURU-ACPL ACPL"
E PIGL" LALA-LADE PIGL-QURO PIGL-PIAB NAangio
EUgymno QURO-ACSA PISY" NAgymno
EUangio
F PIST-PISY PISY-BEPE ACSA" PIGL-LADE QURO" 6"EU" PIAB-ACPL" PISY-LALA
Block C 1" 2" 3" 4" 5" 6" 7" 8"
A ACPL-BEPA 6 angio PIAB-PISY" QURU-QURO QURO-ACSA NAgymno
EUangio 6"EU" QURU-ACPL
B BEPE-QURO" 6 gymno PIGL-ACSA" PISY" PIGL-LADE LALA-QURU" NAangio
EUgymno ACSA"
C LALA" PIST-LALA" PIST" 6"NA" LALA-BEPE PIAB" PISY-LALA PIAB-ACPL"
D PIST-BEPA" LADE" ACSA-BEPA" BEPA-QURU" PIGL-PIST" LADE-QURO" LADE-QURU BEPA"
E PIGL" ACPL-BEPE" BEPA-BEPE PIAB-ACSA PISY-LADE" PIST-PISY QURU" PIGL-PIAB
F LALA-LADE ACSA-ACPL PISY-BEPE BEPE" PIAB-PIST ACPL" QURO" PIGL-QURO
Block B 1" 2" 3" 4" 5" 6" 7" 8"
A ACPL-BEPA 6 angio PISY-LALA LALA" PIAB-PIST BE PE-QURO" PIGL-PIAB BEPA"
B PIST-BEPA" PIGL-ACSA" NAg ymno
EUangio LALA-LADE QURO-ACSA PISY" ACSA-ACPL PIST-LALA"
C PIST" LADE" PIAB-PISY" ACSA" LADE-QURO" PISY-BEPE ACSA-BEPA" BEPA-BEPE
D NAangio
EUgymno ACPL" 6 gymno PIGL-LADE 6"NA" QURU" LADE-QURU LALA-QURU"
E PIGL-PIST" LALA-BEPE QURU-ACPL QURU-QURO PIGL-QURU PIGL" PIAB-ACPL" QURO"
F PIAB" ACPL- BEPE" BEPE" PIAB-ACSA PIST-PISY BEPA-QURU" 6"EU" PISY-LADE"
Block A 1" 2" 3" 4" 5" 6" 7" 8"
A PIGL-PIST" BEPE-QURO" PISY-BEPE PIST-BEPA" LADE" BEPA-QURU" PISY" PIST"
B BEPA" ACSA" QURU-ACPL 6"NA" QURU" 6"EU" PIGL" PIAB-ACPL"
C ACSA-BEPA" ACPL-BEPA PIGL-ACSA" ACSA-ACPL LADE-QURU LALA" PIAB-PIST PISY-LADE"
D ACPL" NAangio
Eugymno LALA-QURU" 6 angio BEPE" LALA-BEPE NAgymno
Euangio PIGL-QURO
E PIST-LALA" PIAB-ACSA PIAB" BEPA-BEPE QURU-QURO ACPL-BEPE" QURO-ACSA PIGL-PIAB
F LALA-LADE LADE-QURO" PIGL-LADE PISY-LALA PIST-PISY 6 gymno QURO" PIAB-PISY"
6"gymno"1" 2" 3" 4" 5" 6" 7"
A" LALA" LADE" PIAB" PIGL" PISY" PIGL" PIAB"
B" PIST" PIGL" PIGL" LADE" PIAB" LADE" PIGL"
C" LADE" PIAB" PIST" PISY" PIST" PISY" PIST"
D" PIST" LADE" PISY" LADE" PIGL" LALA" LADE"
E" PIGL" PISY" LALA" PIGL" PIAB" PIGL" PIAB"
F" LALA" PIST" PIAB" LALA" PISY" LALA" PISY"
G" PIAB" LALA" PIST" PISY" LADE" PIST" LALA"
Figure S1. Cloquet IDENT study design. Within each block, plots were randomly assigned to
a specific plant species treatment (a). Within each plot, individual trees of each species were
randomly assigned to a specific planting grid location (b). Species abbreviations: ACSA =
Acer saccharum, ACPL = A. platanoides, BEPA = Betula papyrifera, BEPE = B. pendula,
LALA = Larix laricina, LADE = L. decidua, PIGL = Picea glauca, PIAB = Picea abies, PIST =
Pinus strobus, PISY = Pinus sylvestris, QURA = Quercus rubra, QURO = Q. robur. In Tables
S1 and S2, ‘Plot’ refers to location of plot within block, e.g. AA5 = Block A, Plot A5 (LADE).
(a)
(b)
Page 41 of 52 Molecular Ecology
For Review Only
Figure S2. Effects of varying sequence cut-offs on plot-level fungal species occurrence
in the ectomycorrhizal (ECM) monoculture plots of the Cloquet IDENT experiment. For
the simulation, the community matrix was transformed to presence-absence based on
each specific threshold from 1 to 21 sequences in increments of two. For example, a
ECM fungal species with four sequences in a plot would be designated as present in
that plot under the threshold of three sequences and absent from that plot under the
threshold of five sequences. After transformation at each specific threshold, the
number of plots in which each ECM fungal species was present was summed. Those
sums were plotted as density curves to show smoothed variation in the distribution
ECM fungal species occurrence in terms of the number of plots in which each ECM
fungal species was present for each cut-off threshold. A global Kruskal-Wallis test was
performed to test for differentiation between plot-level ECM fungal occurrence
distributions. Subsequently, pairwise Wilcoxon rank-sum tests (with an adjustment for
false discover rate) were performed to test for significant differences between all pairs
of distributions.
0.0
0.1
0.2
0.3
010 20 30
Number of Plots Present
Proportion of ECM Fungal Species
Cut-off
1
3
5
7
9
11
13
15
17
19
21
Page 42 of 52Molecular Ecology
For Review Only
ECM Fungal Richness
SAP Fungal Richness
Figure S3. Species accumulation curves showing sample saturation, variation in sequence abundance and depth of
sequencing of ectomycorrhizal (ECM) and saprotrophic (SAP) fungal datasets. For these analyses, species abundances of
less than five sequences per sample were converted to zeros (based on the simulation model in Figure S1). Samples did
not have the same depth in each dataset because ECM and SAP fungal guilds were parsed after the initial rarefaction step.
Page 43 of 52 Molecular Ecology
For Review Only
Figure S4. Bivariate plots of ectomycorrhizal (ECM) fungal richness and the nine
predictor variables in the ‘MuMin’ multivariate analyses.
!
123456
10 15 20 25 30 35
Host Species Richness
ECM Fungal Richness
0.0 0.1 0.2 0.3 0.4 0.5 0.6
10 15 20 25 30 35
Host PSV
ECM Fungal Richness
5000 10000 15000 20000
10 15 20 25 30 35
Host Standing Biomass
ECM Fungal Richness
12345
10 15 20 25 30 35
Host Leaf Area Index
ECM Fungal Richness
050 100 150
10 15 20 25 30 35
Host Specific Root Length
ECM Fungal Richness
1.0 1.5 2.0 2.5
10 15 20 25 30 35
Host Leaf Nitrogen
ECM Fungal Richness
02468
10 15 20 25 30 35
Soil Nitrogen
ECM Fungal Richness
10 12 14 16
10 15 20 25 30 35
Soil Moisture
ECM Fungal Richness
4.8 5.0 5.2 5.4 5.6 5.8 6.0
10 15 20 25 30 35
Soil pH
ECM Fungal Richness
Page 44 of 52Molecular Ecology
For Review Only
123456
20 30 40 50 60
Plant Species Richness
SAP Fungal Richness
0.0 0.1 0.2 0.3 0.4 0.5 0.6
20 30 40 50 60
Plant PSV
SAP Fungal Richness
05000 10000 15000 20000
20 30 40 50 60
Plant Standing Biomass
SAP Fungal Richness
12345
20 30 40 50 60
Total Leaf Area Index
SAP Fungal Richness
20 30 40 50 60 70
20 30 40 50 60
Total Specific Root Length
SAP Fungal Richness
1.0 1.5 2.0 2.5
20 30 40 50 60
Total Leaf Nitrogen
SAP Fungal Richness
02468
20 30 40 50 60
Soil Nitrogen
SAP Fungal Richness
810 12 14 16 18 20 22
20 30 40 50 60
Soil Moisture
SAP Fungal Richness
4.8 5.0 5.2 5.4 5.6 5.8 6.0
20 30 40 50 60
Soil pH
SAP Fungal Richness
Figure S5. Bivariate plots of saprotrophic (SAP) fungal richness and the nine predictor
variables in the ‘MuMin’ multivariate analyses.
!
Page 45 of 52 Molecular Ecology
For Review Only
Europe America Both
0 5 10 15 20 25 30
Host Origin
ECM Fungal Richness
0 5 10 15 20 25 30
Europe America Both
0 10 20 30 40 50 60 70
Plant Origin
SAP Fungal Richness
0 10 20 30 40 50 60
A B C D
0 10 20 30 40
Block
ECM Fungal Richness
A B C D
0 10 20 30 40 50 60 70
Block
SAP Fungal Richness
Figure S6. Ectomycorrhizal (ECM) and saprotrophic (SAP) fungal species richness by block (A,B,C,D) and plant origin
(America, Europe, Both) in the Cloquet IDENT experiment. Host and plant phylum differ only in the exclusion of the Acer-
only plots in the ECM fungal analyses. Boxes surrounding median richness values represent the first and third quartiles,
whiskers show the maximum and minimum values excluding outliers (which are represented by dots).
Page 46 of 52Molecular Ecology
For Review Only
123456
10 15 20 25 30 35
Angiosperm
Host Richness
ECM Fungal Richness
R2 = 0.0640711
p = 0.153488
123456
10 12 14 16 18 20 22
Gymnosperm
Host Richness
ECM Fungal Richness
R2 = -0.01643982
p = 0.459671
Figure S7. Ectomycorrhizal (ECM) fungal species richness in the angiosperm monoculture plots vs. the 6 angiosperm
plots or the gymnosperm monoculture plots vs. the 6 gymnosperm plots in the Cloquet IDENT experiment. See methods
for details about the specific composition of the 6 species plots.
Page 47 of 52 Molecular Ecology
For Review Only
Figure S8. Pair-wise correlation matrix of the continuous biotic and abiotic variables used
to predict ECM fungal species richness in the Cloquet IDENT experiment. SoilH2OAvg =
soil moisture, SoilN = plant available soil nitrogen, SoilpH = soil pH, HostMass =
aboveground standing biomass, HostLeafN = leaf nitrogen content, HostLAI = leaf area
index, HostSRL = specific root length, HostPSV = phylogenetic species variability, HostSR
= host species richness. See methods for details on how each variable was calculated.
Red values represent positive correlation coefficients and blue values represent negative
correlation coefficients, with significance denoted by an asterisk.
SoilH2OAvg
048
0.212
0.046
5000 20000
-0.373
**
-0.132
135
-0.042
0.091
0.0 0.3 0.6
-0.013
10 14
-0.05
048
SoilN
0.11
-0.177
-0.091
-0.078
-0.049
-0.061
-0.027
SoilpH
0
0.135
0.012
-0.064
0.136
4.8 5.4 6.0
-0.024
5000 20000
HostMass
0.6
***
0.492
***
0.027
0.113
0.104
HostLeafN
0.605
***
0.335
*
-0.077
1.0 2.0
-0.08
135
HostLAI
0.499
***
0.163
0.142
HostSRL
0.181
050 150
0.057
0.0 0.3 0.6
HostPSV
0.741
***
10 14 4.8 5.4 6.0 1.0 2.0 050 150 135
135
HostSR
Environmental Response Variables: Bivariate Correlation Matrix
Page 48 of 52Molecular Ecology
For Review Only
Table S5. Results of ectomycorrhizal fungal richness multimodel averaging comparing
two subsets of models: delta < 4 from lowest AICc score or 95% confidence interval of
AICc score. Sc = Scaled. LAI = Leaf Area Index, Mass = Standing Mass, PSV =
Phylogenetic Diversity, SRL = Specific Root Length, H20Avg = Soil Water, pH = Soil
pH, LeafN = Leaf Nitrogen, N = Soil Nitrogen, SR = Species Richness. See Table 1 for
units of variables.
Delta < 4:
95% Confidence Interval:
Page 49 of 52 Molecular Ecology
For Review Only
Table S6. Results of saprotrophic fungal richness multimodel averaging comparing two
subsets of models: delta < 4 from lowest AICc score or 95% confidence interval of AICc
score. Sc = Scaled. LAI = Leaf Area Index, Mass = Standing Mass, PSV = Phylogenetic
Diversity, SRL = Specific Root Length, H20Avg = Soil Water, pH = Soil pH, LeafN =
Leaf Nitrogen, N = Soil Nitrogen, SR = Species Richness. See Table 1 for units of
variables.
Delta < 4:
95% Confidence Interval:
Page 50 of 52Molecular Ecology
For Review Only
Table S7. Model of ectomycorrhizal fungal richness with lowest AICc score in the full
multimodel analysis using the ‘dredge’ function in MuMIn.
Page 51 of 52 Molecular Ecology
For Review Only
Table S8. Model of saprotrophic fungal richness with lowest AICc score in the full
multimodel analysis using the ‘dredge’ function in MuMIn.
Page 52 of 52Molecular Ecology
... Roots with higher SRL can grow faster because they require less carbon to construct, but this also determines the timing, quality, and quantity of host-derived resources and habitat available for soil microbes (Pérez-Jaramillo et al. 2018). Thus, SRL is known to influence the richness of root-associated saprophytic fungi (Nguyen et al. 2016) as well as colonization intensity of mycorrhizal fungi (Baon et al. 1994), and infection by bacterial endophytes has been shown to alter root carbon allocation (Henning et al. 2016). ...
... High SRL is often associated with low-fertility soils (Pérez-Jaramillo et al. 2018), so soil conditions could shape the patterns observed here. Lower SRL can also be correlated with greater intensity of infection by mycorrhizal fungi (Baon et al. 1994), as well as higher diversity of saprotrophic fungi (Nguyen et al. 2016). In addition, the possibility remains that these differences are also driven by an effect of evolutionary history on root exudates, which are key components in the perception of host roots by both beneficial and pathogenic bacteria (Bulgarelli et al. 2013). ...
Article
Rhododendron are popular ornamental plants which are well-known for forming mycorrhizal associations with ericoid fungi, but little is known about how host traits influence their microbiome more broadly. This study investigated leaf, root, rhizosphere soil, and bulk soil bacterial and fungal community structure for 12 Rhododendron species, representing four taxonomic clades with different leaf habits. Samples were collected when ephemeral hair roots colonized by ericoid mycorrhizae were absent, and microbial community structure was compared to leaf and root morphology for the same plants. Root morphology and the fungal communities of roots and rhizosphere soil were primarily structured by host ancestry. Leaf bacterial and fungal communities were even more distinct across clades than for roots or rhizosphere, and microbial communities of leaves and bulk soil were similarly structured by clade-wise differences in leaf morphology, suggesting a role for Rhododendron leaf litter in belowground microbial community structure. This work sheds new light on host traits influencing microbial community structure of ericaceous plants, showing a strong influence of ancestry, but also that different host traits drive bacterial and fungal communities across different plant compartments, suggesting future work on factors that drive similarity among close relatives in the non-ericoid microbes associating with Rhododendron.
... More diverse resources and greater spatial heterogeneity can be provided for microbial communities in more diverse stands. Therefore, fungal diversity can be expected to increase and fill a number of niches [6,7]. Nevertheless, previous studies reported inconsistent correlations between plant diversity and fungal diversity, especially at different scales [8]. ...
... The relative abundance of symbiotic fungi decreased with the increase of tree species diversity, but the difference was not significant. Different tree species diversity gradients and differences among regions and regional species pools will change the soil fungi trophic modes and functional guilds [7,21,40]. Pathogenic fungi can attack host cells and obtain nutrients from them, but they can also control the population structures of other plants, animals, fungi and harmful insects [43]. Meanwhile, Gonthier et al. [44] demonstrated that increased ectomycorrhizal fungi richness may inhibit the invasion and reproduction of pathogenic fungi richness. ...
Article
Full-text available
The majority of studies have found that an increase in tree species diversity can increase the productivity of forest stands thanks to complimentary effects with enhanced resource use efficiency or selection effects; however, it is unclear how tree species diversity affects the soil fungal community and enzyme activities in subtropical evergreen forests. In this study, we used soil high-throughput sequencing to investigate the soil fungal community structure and diversity in the central area of tree clusters in the gradient of tree species richness formed by four possible dominant tree species (Pinus massoniana Lamb., Choerospondias axillaris Roxb., Cyclobalanopsis glauca Thunb. and Lithocarpus glaber Thunb.) in subtropical evergreen broad-leaved forest. The results showed that soil organic carbon content and total nitrogen content were significantly higher in mixed tree clusters, and that soil fungal richness and diversity increased with the increase in tree species diversity (1–3 species). Soil acid phosphatase and urease activity were also enhanced with tree species diversity (p < 0.05). The relative abundance of soil symbiotic fungi (ectomycorrhizal fungi) decreased, while the relative abundance of saprotrophic fungi increased. Redundancy analysis (RDA) revealed that soil acid phosphatase activity was the main factor affecting soil fungal communities and functional guilds, and that soil water content was the main driving force behind fungal trophic modes. In subtropical forests, changes in tree species diversity have altered the soil fungal community structure and trophic modes and functions, accelerating the decomposition of organic matter, increasing nutrient cycling, and perhaps also changing the nutrient absorption of trees.
... In our study, tree litters and rice stubbles, which were returned to the fields, were two sources of SOM and N. Their heterogeneous distribution patterns could recruit different fungal communities. Furthermore, a previous study has reported that saprotrophic fungal diversity was positively correlated with leaf N content (Nguyen et al. 2016). Also, diversity of arbuscular mycorrhizal fungi was negatively related to N content (Zeng et al. 2021). ...
Article
Full-text available
Background Agroforestry is a promising approach for sustainable agriculture due efficient resource cycling and improved soil fertility. Bombax ceiba (Malvaceae), a tall tree with red flowers blooming in spring, is traditionally planted in rice fields in tropical Asia. However, the role of B. ceiba in the agroforestry systems remains unexplored. Methods We collected 81 soil samples at different distances to B. ceiba (0 m (D0), 1 m (D1), and 5 m (D5)) at a typical B. ceiba-rice agroforestry system in Hainan Island (south China) during three reproductive stages of rice- booting, heading and maturity. We assessed spatiotemporal variations of soil nutrient properties (by a soil nutrient analyzer (YT-TRX03)), and soil bacterial and fungal communities (by sequencing 16S rRNA gene and internal transcribed spacer (ITS) region, respectively). Results B. ceiba improved the soil nutrient conditions in a rice field, particularly the availability of potassium and soil organic matter. Soil microbial communities were significantly affected by the distances to B. ceiba and the reproductive stages of rice. Available potassium was the principal driver of soil bacterial diversity and structure. In contrast, fungal diversity was negatively correlated with total nitrogen, while soil organic matter was the main factor shaping fungal community structure. Conclusions. Our results show that B. ceiba has positive impacts on abiotic traits of rice-growing soils. B. ceiba can change soil microbial community structure, however, the principal soil driver varied according to microbial taxa. These findings support the ecological basis of traditional agroforestry systems prevalent in tropical Asia.
... Possibly due to the low availability of favorable CWD substrate in our study sites, the lower diameter wood fractions appeared to be significant in explaining a part of the variation in WSF (VFWD) and PF (FWD), indicating that this fraction of deadwood is an important microrefugium for wood-inhabiting fungi (Juutilainen et al., 2014;Abrego and Salcedo, 2014). Similar to ECMF, pathogenic fungi responded more strongly to tree-related parameters (the variety of tree species and the number of larger trees), confirming the earlier observation that different fungal ecological guilds respond to vegetation and abiotic factors to a different degree, with biotrophic symbionts and pathogens being more tightly linked to tree traits than saprotrophs Nguyen et al., 2016b, Kolaříková et al., 2017Prada-Salcedo et al., 2021). ...
Article
Full-text available
We report the first robust estimate of macrofungal assemblages in European mixed coniferous forest to better understand species diversity under two management strategies (managed vs. not managed) and to rate the recovery of forest reserves, transformed from formerly managed stands, to their natural state. Following extensive repeated sampling in three regions of Poland, we found 318 fungal species from different trophic groups. The two forest management strategies showed similar cumulative fungal species richness and values for Shannon, evenness, and Simpson’s dominance indices, suggesting that both management strategies did not generate factors stimulating or limiting species richness and biodiversity. In contrast, site as well as forest management strategy affected the composition of all fungal trophic groups, with site being a stronger influencing factor. Different fungal ecological groups responded differently to environmental drivers, with ectomycorrhizal symbionts and parasites being more tightly linked to tree traits than saprotrophs, which were affected mostly by climate and substrate. Indicator species associated mainly with managed forests comprised predominantly wood-inhabiting fungi. Conservation-relevant species (red-listed in Poland) were noted in both forest reserves and managed forests; however, they predominated in the former. The results indicated a rather slow process of transformation of fungal communities in forest reserves originating from previously managed forests. The major conclusion of this study is that managed forests complement forest reserves in fungal diversity conservation.
... Nextgeneration sequencing technologies are a powerful tool to examine the entire microbial community of a given sample without the need to culture microbes (Borsting & Morling, 2015). Fungal community diversity and structure in various plant tissues, including leaves, stems, roots, flowers and fruits, were investigated to understand plant-microbe interactions (Nguyen et al., 2016;. Sessa, Abreo, and Lupo (2018) revealed many known endophytic fungi and potential pathogenic fungi causing wood disease in the healthy shoots of fruit trees. ...
Article
Full-text available
In the present study, the differences in the antioxidant capability, metabolite composition and fungal diversity in peach gum with various colours were investigated. Metabolomics revealed that peach gum comprised many small-molecule metabolites (including primary and secondary metabolites), and most polyphenols (such as flavonoids and phenolic acids) showed a significantly positive relationship with the colour deepening, total phenol content and antioxidant capability. Using fungal diversity analysis, the abundance of five fungi at the genus level increased with peach gum colour deepening, and the fungi demonstrated a significantly positive relationship with two defense hormones (salicylic acid and abscisic acid) and most polyphenols (particularly flavonoids). The gummosis pathogenic fungus Botryosphaeria was among the five fungi, suggesting that peach gum colouring may reflect plant defense responses against pathogenic fungi. Additionally, the concentrations of 12 flavonoids in peach gum samples were detected based on LC-QQQ/MS, among which hesperetin, naringenin and eriodictyol were the most abundant.
... Furthermore, significant separation of EcM fungal communities were observed among different host genera (Fig. 5a, Table S3), which indicated the importance of host in determining the EcM fungal community (17,(63)(64)(65). The host specificity (63) could be a cause of this separation. ...
Article
Full-text available
Altitude and season represent two important environmental gradients that shape the structure of biome, including the heterogeneity of EcM fungi. Previous studies have separately considered the influences of altitude and season on EcM fungal communities, but the relative importance of altitude and season is still unknown.
... This was expected given fungi cannot x carbon and nitrogen and its growth mainly rely on external organic matter and nitrogenous compounds (Walker and White 2017), thus changes of soil SOC and TN may bring changes of fungal community. In addition, a previous study reported that saprotrophic fungal diversity was positively correlated with leaf nitrogen content (Nguyen et al. 2016), which suggested variations of fungal community in our study may be due to different responses of fungal guilds to soil microhabitat. ...
Preprint
Full-text available
Background Agroforestry system is one of promising directions for developing sustainable agriculture because the intercropping of crops and trees may facilitate resource cycling and avoid soil degradation. Bombax ceiba (Malvaceae), a tall tree with red flowers blooming in Spring, is traditionally planted in rice paddies in tropical Asia, while the roles of B. ceiba in the agroecosystem remain unexplored. Methods In this paper, we investigated and assessed spatiotemporal variations of soil nutrient contents and soil bacterial and fungal communities along the distance gradients to B. ceiba during three reproductive stages of rice, i.e. booting, heading, mature, in a typical Bombax-dominated rice paddy on Hainan Island, South China. Results B. ceiba in rice paddy could improve the soil nutrient conditions, particularly available K and soil organic carbons. The relative abundance of Chloroflexi and Ascomycota increased while Actinobacteria decreased along the distance gradient to B. ceiba. In addition, the relative abundance of Firmicutes was highest at harvesting stage of rice, while Acidobacteria was richer in the early reproductive stage of rice. Soil potassium content was the principal driver in shaping soil bacterial diversity and composition, while fungal community was mainly affected by soil nitrogen. Conclusions Our results provide evidences for positive influences of B. ceiba on biotic and abiotic traits of rice paddy soils and thus lend supports to the ecological basis of this tropical Asian endemic traditional Agroforestry system, which could increase resource cycling and paddy stability and have the potential to reduce carbon emission.
... At the local scale, numerous studies have reported positive correlations between mycorrhizal fungal richness and plant species richness. Manipulative studies show that plant diversity is linked to belowground AM (Kivlin & Hawkes, 2011) and EM fungal diversity Nguyen et al., 2016) and function (Eisenhauer et al., 2010). In glasshouse trials, plant performance is often optimized when plants associate with highly diverse AM fungal communities (Maherali & Klironomos, 2007), perhaps owing to complementarity in function among AM fungal clades (Powell et al., 2009). ...
Article
Full-text available
First principles predict that diversity at one trophic level often begets diversity at other levels, suggesting plant and mycorrhizal fungal diversity should be coupled. Local-scale studies have shown positive coupling between the two, but the association is less consistent when extended to larger spatial and temporal scales. These inconsistencies are likely due to divergent relationships of different mycorrhizal fungal guilds to plant diversity, scale dependency, and a lack of coordinated sampling efforts. Given that mycorrhizal fungi play a central role in plant productivity and nutrient cycling, as well as ecosystem responses to global change, an improved understanding of the coupling between plant and mycorrhizal fungal diversity across-scales will reduce uncertainties in predicting the ecosystem consequences of species gains and losses.
Article
Multi-mycotoxin contamination of barley reduces malt quality and also poses serious health risks to both humans and animals. Among the detected mycotoxins in barley, Fusarium and Alternaria toxins represent severe food safety issues due to their widespread prevalence and strong synergistic toxicity. In this study, the effects of Fusarium and Alternaria fungi and their related toxins were investigated using simulated malting process conditions for contaminated barley samples. The dynamic changes of the fungal community and their associated mycotoxins were analyzed via high-throughput sequencing and UHPLC-q-trap-MS/MS, respectively. The results showed significant differences in the fungal communities between barley grains and their associated malt samples. These differences were noted for the following: fungal community diversity, dominant flora and the fungal structures at phylum, genus, and family level. Additionally, significant changes in the levels of alternariol monomethyl ether, tentoxin, zearalenone, and 15Acetyl-Deoxynivelenone were observed (p < 0.05) during the malting process. These results provide valuable information for strategies aimed at controlling fungal infections and the presence of mycotoxins in malt, so as to ensure food safety and human health.
Article
Full-text available
Fungi represent a diverse group of organisms that play an essential role in maintaining soil health and ecosystem functioning. Plant root exudates form nutrient-rich niches that harbor specific fungal communities, or so-called rhizosphere mycobiomes. The long-term application of fertilizers supplies the soil with nutrients that may override the plant-related effects on rhizosphere fungal communities. Here, we assessed the effect of contrasting fertilization regimes on the composition, diversity, and abundance of bulk soil and rhizosphere mycobiomes of potato, white mustard, and maize under NPK (mineral fertilizers) or fresh cattle manure (organic fertilizers). Mineral and organic fertilizers led to distinct fungal communities in the rhizospheres of all studied crops, and the plant-related effects on the mycobiome were overridden by the effect of fertilization. The abundances of Ascomycota and Olpidiomycota were higher under manure, while the abundances of Basidiomycota and Monoblepharomycota increased under NPK. Manure input strongly increased fungal abundance but decreased fungal diversity and the total number of species. NPK had a slight effect on fungal diversity, but significantly increased the relative abundances of fungal phytopathogens, such as Alternaria and Fusarium. Our study shows that that potential plant species effects on the abundance and diversity of the rhizosphere mycobiomes are governed by long-term fertilization. Fertilization management could therefore be used to manipulate rhizosphere fungal communities and soilborne pathogen suppressiveness.
Article
Full-text available
Fungi typically live in highly diverse communities composed of multiple ecological guilds. Although high-throughput sequencing has greatly increased the ability to quantify the diversity of fungi in environmental samples, researchers currently lack a simple and consistent way to sort large sequence pools into ecologically meaningful categories. We address this issue by introducing FUNGuild, a tool that can be used to taxonomically parse fungal OTUs by ecological guild independent of sequencing platform or analysis pipeline. Using a database and an accompanying bioinformatics script, we demonstrate the application of FUNGuild to three high-throughput sequencing datasets from different habitats: forest soils, grassland soils, and decomposing wood. We found that guilds characteristic of each habitat (i.e., saprotrophic and ectomycorrhizal fungi in forest soils, saprotrophic and arbuscular mycorrhizal fungi in grassland soils, saprotrophic, wood decomposer, and plant pathogenic fungi in decomposing wood) were each well represented. The example datasets demonstrate that while we could quickly and efficiently assign a large portion of the data to guilds, another large portion could not be assigned, reflecting the need to expand and improve the database as well as to gain a better understanding of natural history for many described and undescribed fungal species. As a community resource, FUNGuild is dependent on third-party annotation, so we invite researchers to populate it with new categories and records as well as refine those already in existence.
Chapter
Full-text available
Community phylogenetics combines ideas from community ecology and evolutionary biology, using species phylogeny to explore the processes underlying ecological community assembly. Here, we describe the development of the field’s comparative methods and their roots in conservation biology, biodiversity quantification, and macroevolution. Next, we review the multitude of community phylogenetic structure metrics and place each into one of four classes: shape, evenness, dispersion, and dissimilarity. Shape metrics examine the structure of an assemblage phylogeny, while evenness metrics incorporate species abundances. Dispersion metrics examine assemblages given a phylogeny of species that could occupy those assemblages (the source pool), while dissimilarity metrics compare phylogenetic structure between assemblages. We then examine how metrics perform in simulated communities that vary in their phylogenetic structure. We provide an example of model-based approaches and argue that they are a promising area of future research in community phylogenetics. Code to reproduce all these analyses is available in the Online Practical Material (http:// www. mpcm-evolution. org). We conclude by discussing future research directions for the field as a whole.
Article
Full-text available
Next generation fungal amplicon sequencing is being used with increasing frequency to study fungal diversity in various ecosystems; however, the influence of sample preparation on the characterization of fungal community is poorly understood. We investigated the effects of four procedural modifications to library preparation for high-throughput sequencing (HTS). The following treatments were considered: 1) the amount of soil used in DNA extraction, 2) the inclusion of additional steps (freeze/thaw cycles, sonication, or hot water bath incubation) in the extraction procedure, 3) the amount of DNA template used in PCR, and 4) the effect of sample pooling, either physically or computationally. Soils from two different ecosystems in Minnesota, USA, one prairie and one forest site, were used to assess the generality of our results. The first three treatments did not significantly influence observed fungal OTU richness or community structure at either site. Physical pooling captured more OTU richness compared to individual samples, but total OTU richness at each site was highest when individual samples were computationally combined. We conclude that standard extraction kit protocols are well optimized for fungal HTS surveys, but because sample pooling can significantly influence OTU richness estimates, it is important to carefully consider the study aims when planning sampling procedures.
Article
Net pairwise plant-soil feedbacks (PSF) may be an important factor structuring plant communities, yet the influence of abiotic context on PSF is not yet understood. Abiotic factors such as light availability can alter plant-soil interactions, potentially resulting in strong context dependence of PSF. Here, we present an experiment in which we measured whole-soil net pairwise feedbacks amongst six common forest understory species across a gradient of light availability. Light treatments were imposed throughout both phases (the 'conditioning phase' and the 'response phase') of the feedback experiment. Across the plant community, PSF shifted from negative