PreprintPDF Available

Understudied, underrepresented, and unknown: methodological biases that limit detection of early diverging fungi from environmental samples

April 2021

April 2021

DOI:10.22541/au.161839272.27561225/v1

Authors:

Michelle A. Jusino

US Forest Service

Preprints and early-stage research may not have been peer reviewed yet.

Metabarcoding is an important tool for understanding fungal communities. The internal transcribed spacer (ITS) rDNA is the accepted fungal barcode but has known problems. The large subunit (LSU) rDNA has also been used to investigate fungal communities but available LSU metabarcoding primers were mostly designed to target Dikarya (Ascomycota + Basidiomycota) with little attention to early diverging fungi (EDF). However, evidence from multiple studies suggests that EDF comprise a large portion of unknown diversity in community sampling. Here we investigate how DNA marker choice and methodological biases impact recovery of EDF from environmental samples. We focused on one EDF lineage, Zoopagomycota, as an example. We evaluated three primer sets (ITS1F/ITS2, LROR/LR3, and LR3 paired with new primer LR22F) to amplify and sequence a Zoopagomycota mock community and a set of 146 environmental samples with Illumina MiSeq. We compared two taxonomy assignment methods and created an LSU reference database compatible with AMPtk software. The two taxonomy assignment methods recovered strikingly different communities of fungi and EDF. Target fragment length variation exacerbated PCR amplification biases and influenced downstream taxonomic assignments, but this effect was greater for EDF than Dikarya. To improve identification of LSU amplicons we performed phylogenetic reconstruction and illustrate the advantages of this critical tool for investigating identified and unidentified sequences. Our results suggest much of the EDF community may be missed or misidentified with “standard” metabarcoding approaches and modified techniques are needed to understand the role of these taxa in a broader ecological context.

Content uploaded by Michelle A. Jusino

Content may be subject to copyright.

Posted on Authorea 14 Apr 2021 | The copyright holder is the author/funder. All rights reserved. No reuse without permission. | https://doi.org/10.22541/au.161839272.27561225/v1 | This a preprint and has not been peer reviewed. Data may be preliminary.

Understudied, underrepresented, and unknown: methodological

biases that limit detection of early diverging fungi from

environmental samples

Nicole Reynolds1, Michelle Jusino2, Jason Stajich3, and Matthew Smith1

1University of Florida

2Williams and Mary College

3University of California Riverside

April 14, 2021

Abstract

Metabarcoding is an important tool for understanding fungal communities. The internal transcribed spacer (ITS) rDNA is the

accepted fungal barcode but has known problems. The large subunit (LSU) rDNA has also been used to investigate fungal

communities but available LSU metabarcoding primers were mostly designed to target Dikarya (Ascomycota + Basidiomycota)

with little attention to early diverging fungi (EDF). However, evidence from multiple studies suggests that EDF comprise a

large portion of unknown diversity in community sampling. Here we investigate how DNA marker choice and methodological

biases impact recovery of EDF from environmental samples. We focused on one EDF lineage, Zoopagomycota, as an example.

We evaluated three primer sets (ITS1F/ITS2, LROR/LR3, and LR3 paired with new primer LR22F) to amplify and sequence

a Zoopagomycota mock community and a set of 146 environmental samples with Illumina MiSeq. We compared two taxonomy

assignment methods and created an LSU reference database compatible with AMPtk software. The two taxonomy assignment

methods recovered strikingly diﬀerent communities of fungi and EDF. Target fragment length variation exacerbated PCR

ampliﬁcation biases and inﬂuenced downstream taxonomic assignments, but this eﬀect was greater for EDF than Dikarya. To

improve identiﬁcation of LSU amplicons we performed phylogenetic reconstruction and illustrate the advantages of this critical

tool for investigating identiﬁed and unidentiﬁed sequences. Our results suggest much of the EDF community may be missed

or misidentiﬁed with “standard” metabarcoding approaches and modiﬁed techniques are needed to understand the role of these

taxa in a broader ecological context.

Title: Understudied, underrepresented, and unknown: methodological biases that limit detec-

tion of early diverging fungi from environmental samples

Short title: Metabarcoding biases limit detection of EDF

Nicole K. Reynolds1*, Michelle A. Jusino2, Jason E. Stajich3, Matthew E. Smith1

1University of Florida, Department of Plant Pathology, Gainesville, Florida 32611

2Williams and Mary College, Department of Biology, Williamsburg, Virginia 23185

3Department of Plant Pathology & Microbiology and Institute for Integrative Genome Biology, University

of California–Riverside, Riverside, California 92521

*Corresponding author: nicolereynolds1@uﬂ.edu

Abstract

Metabarcoding is an important tool for understanding fungal communities. The internal transcribed spacer

(ITS) rDNA is the accepted fungal barcode but has known problems. The large subunit (LSU) rDNA has

also been used to investigate fungal communities but available LSU metabarcoding primers were mostly

designed to target Dikarya (Ascomycota + Basidiomycota) with little attention to early diverging fungi

(EDF). However, evidence from multiple studies suggests that EDF comprise a large portion of unknown

diversity in community sampling. Here we investigate how DNA marker choice and methodological biases

impact recovery of EDF from environmental samples. We focused on one EDF lineage, Zoopagomycota, as

an example. We evaluated three primer sets (ITS1F/ITS2, LROR/LR3, and LR3 paired with new primer

LR22F) to amplify and sequence a Zoopagomycota mock community and a set of 146 environmental samples

with Illumina MiSeq. We compared two taxonomy assignment methods and created an LSU reference

database compatible with AMPtk software. The two taxonomy assignment methods recovered strikingly

diﬀerent communities of fungi and EDF. Target fragment length variation exacerbated PCR ampliﬁcation

biases and inﬂuenced downstream taxonomic assignments, but this eﬀect was greater for EDF than Dikarya.

To improve identiﬁcation of LSU amplicons we performed phylogenetic reconstruction and illustrate the

advantages of this critical tool for investigating identiﬁed and unidentiﬁed sequences. Our results suggest

much of the EDF community may be missed or misidentiﬁed with “standard” metabarcoding approaches and

modiﬁed techniques are needed to understand the role of these taxa in a broader ecological context.

Key Words

Illumina MiSeq, ITS, LSU, Metabarcoding, Mock community, Zoopagomycota

Introduction

Metabarcoding of fungal communities using high-throughput technologies is a powerful tool for investigating

fungal ecology. The internal transcribed spacer (ITS) region of the rDNA operon has been used extensively

as the DNA barcode for fungi, particularly in environmental sequencing studies (Gardes & Bruns, 1996;

Schoch et al., 2012). Because the ITS region has been used as a barcode for approximately two decades, the

bioinformatic tools for processing amplicon data are well developed and the ITS reference databases (UNITE

and INSD) are consistently curated and updated (Abarenkov et al., 2020; Nilsson et al., 2019). However,

even with the widespread usage of these databases there are problems, including poor representation of

some taxonomic groups, low quality sequences, and incorrect taxonomic annotation (Abarenkov et al., 2018;

Hofstetter et al., 2019; Nilsson et al., 2012). These problems mean that results from taxonomic assignments

of operational taxonomic units (OTUs) must be interpreted with caution (Nilsson et al., 2006; Yahr et al.,

2016). Despite its widespread use, the ITS is not a suitable barcode for all taxonomic groups. There are

known issues with sequencing ITS in some fungal lineages, including high variability in ITS length between

groups (resulting in favored PCR and sequencing for shorter fragments) (Casta˜no et al., 2020; Engelbrektson

et al., 2010; Manter & Vivanco, 2007), lack of interspeciﬁc discrimination (and therefore inability to use the

marker for species-level determinations) (e.g. Gazis et al., 2011), interspeciﬁc rDNA copy number variation

(Lindner & Banik, 2011), and primer biases which exclude some groups (Bellemain et al., 2010; Li et al.,

2020; Tedersoo & Lindahl, 2016). For example, among arbuscular mycorrhizal fungi (AMF) the ITS is

hypervariable and has high intraspeciﬁc and intra-spore variation compared to the small (SSU) and large

(LSU) rDNA subunits (Egan et al., 2018; Thi´ery et al., 2012). One study found up to 6% divergence among

sequences from a single spore (Lloyd-MacGilp et al., 1996). A recent study also found wide rDNA copy

number variation across kingdom Fungi that was uncorrelated with trophic mode (Lofgren et al., 2019),

making such variation unpredictable in environmental samples. Interspeciﬁc rDNA variation can lead to

the formation of multiple OTUs derived from a single individual and individuals with more rDNA copies

could potentially dominate during PCR ampliﬁcation and sequencing from mixed templates. Among early

diverging fungal (EDF) lineages, direct comparisons of markers for metabarcoding have not been performed

for many groups. An exception are the AMF for which the SSU, LSU, and ITS have been evaluated and

some combination of two markers is commonly used (Hart et al., 2015; ¨

Opik et al., 2014). Additionally, the

LSU was suggested to perform better than ITS as a barcode for EDF due to greater PCR success and a

larger barcode gap (i.e. diﬀerence between inter- and intraspeciﬁc variation) than SSU (Schoch et al., 2012).

Another critical advantage of the LSU and SSU is the ability to perform phylogenetic reconstruction with

the OTUs to provide preliminary placement of unidentiﬁed sequences.

The majority of fungal metabarcoding studies have focused on the Dikarya. In contrast, EDF such as chytrids

and zygomycetes are often overlooked or ignored in environmental sequencing studies. EDF are also generally

missing or underrepresented during attempts to develop “universal” fungal primers for metabarcoding. For

example, mock communities used to validate primer performance often contain none or a few EDF, despite

the fact that these fungi often have highly divergent target sequences relative to Dikarya (e.g. Ihrmark et

al., 2012; P´erez-Izquierdo et al., 2017; Stielow et al., 2015; Tedersoo et al., 2015). Early diverging lineages

are among the least studied fungi and are generally challenging to collect and manipulate in the lab. Many

EDF are not culturable using standard techniques, are obligate symbionts, and have limited or no sequence

data available (e.g. Benny et al. 2016; Corsaro et al. 2014, 2017; Lazarus & James, 2015; Letcher & Powell,

2019; Malar et al. 2021). Available data suggest that a large portion of undescribed taxa belong to EDF

lineages (Tedersoo et al. 2017, 2020; Torres-Cruz et al., 2017; Walsh et al., 2020). Metabarcoding methods can

provide an essential tool for learning more about these “dark matter fungi” (Grossart et al., 2016), especially

if taxonomy can be reliably assigned to the OTUs. Among the most understudied groups of EDF is the

Zoopagomycota. The placement of Zoopagomycota is still unresolved but phylogenomic studies indicate this

lineage is either sister to all other terrestrial fungi (Dikarya + Mucoromycota – Spatafora et al., 2016) or

sister to the Mucoromycota (Li et al., 2021). This phylum is ecologically diverse and includes fungal parasites

(mycoparasites) as well as parasites of small animals. Specialized enrichment methods indicate that some

taxa are diverse and widespread in soils, leaf litter, and dung (e.g. Benjamin, 1958; Benny et al., 2016;

Drechsler, 1938; Duddington, 1955). Despite these ﬁndings, Zoopagomycota species are absent or found in

low abundance in most metabarcoding studies (Lazarus et al., 2017; Reynolds et al., 2019).

We focused on some species of Zoopagomycota as a test case for evaluating methodological biases because:

1) these fungi are routinely found in soil during culture-based studies (e.g. Benjamin 1958; Benny et al.,

2016) and yet they are not readily detected with metabarcoding, 2) there are limited reference sequence

data available, 3) among species for which sequence data are available there is large ITS sequence length

variation (Lazarus et al., 2017; Reynolds et al., 2019), and 4) they are diﬃcult to isolate, culture, and work

with in the lab, making environmental sampling a particularly important investigative tool. Out of the three

subphyla (Entomophthoromycotina, Kickxellomycotina, and Zoopagomycotina) we focused on taxa that ha-

ve been isolated from soil samples. This includes members of the Zoopagomycotina and Kickxellomycotina

that have been regularly isolated from our study sites and for which we have a large culture collection.

Zoopagomycotina species are mycoparasites that primarily attack host fungi in the Mucoromycota (and

sometimes Ascomycota) or are parasites of microinvertebrates (e.g. amoebae, nematodes, rotifers) (Zoopa-

gales). The Kickxellomycotina is the most diverse subphylum and includes mycoparasites (Dimargaritales)

that are parasitic on Mucoromycota and Ascomycota, commensalistic arthropod gut-dwelling species (Asel-

lariales, Harpellales, Orphellales), and putative saprotrophs (Kickxellales). We collected soil (Dimargaritales,

Kickxellales, Zoopagales), freshwater sediments and water (Harpellales, Orphellales), and microinvertebrate

(Zoopagales) samples from multiple sites in California (CA) and Florida (FL), two geographically distant

states with divergent climate and soils. Both locations have been heavily sampled for Zoopagomycota fun-

gi using selective culturing methods (Benjamin 1958, 1959, 1961; Benny et al., 2016; Lazarus et al., 2017;

Reynolds et al., 2019). We did not include Entomophthoromycotina because they are obligate arthropod

pathogens and likely to be rare or absent from many environmental samples.

In order to detect fungal communities, we compared the ITS1 marker region to two diﬀerent regions of the

LSU rDNA, using two primer pairs for LSU (LROR/LR3 and LR3 paired with the newly designed primer

LR22F). Because relatively few studies have used LSU as a sole marker for proﬁling fungal communities, we

also compared the eﬃcacy of two methods for LSU OTU taxonomy assignment: the RDP Na¨ıve Bayesian

Classiﬁer (hereafter RDP classiﬁer) (Wang et al., 2007) and UTAX (Edgar, 2010) to search a manually

curated reference database. We created the new LSU database by combining the RDP (Cole et al., 2014)

and SILVA (Quast et al., 2013) reference databases along with additional sequences from GenBank and

our lab. Finally, we used a subset of LSU OTUs generated from each primer pair that were identiﬁed as

Posted on Authorea 14 Apr 2021 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.161839272.27561225/v1 — This a preprint and has not been peer reviewed. Data may be preliminary.

Zoopagomycota or only to kingdom Fungi and included them in phylogenetic reconstructions. Analyzing am-

plicons in this way allowed us to validate the taxonomic assignments made by our pipeline as well as identify

putative errors, a step that is not possible with ITS data. We created a Zoopagomycota mock community

that was evaluated alongside the environmental samples to investigate how the metabarcoding pipeline af-

fected that group speciﬁcally. We aimed to address the following questions: 1) Are Zoopagomycota truly rare

in the environment?, 2) Are methodological biases (such as marker choice and fragment length) inhibiting

the detection of Zoopagomycota?, and 3) Are both factors contributing to the absence of Zoopagomyco-

ta in metabarcoding studies? We hypothesized that: 1) fragment length ampliﬁcation and sequencing bias

would decrease the detection of Zoopagomycota species, 2) the LR22F/LR3 primer pair would outperform

the LROR/LR3 primer pair in terms of OTU clustering and phylogenetic reconstruction, and 3) taxonomy

assignment methods would strongly diﬀer in the identiﬁcation of Zoopagomycota and other EDF OTUs.

Materials and Methods

Zoopagomycota fungi in the mock community –A mock community comprised of Zoopagomycota fungi was

created by combining equilibrated aliquots of genomic DNA from species of Dimargaritales, Kickxellales, and

Zoopagales. DNA of mock community members was obtained from cultures grown from the University of

Florida Gerald Benny culture collection or received from collaborators (TABLE 1). Species of Dimargaritales,

Piptocephalis, and Syncephalis are haustorial mycoparasites and were grown in dual cultures with their

host fungi (Benny et al., 2016b). Accordingly, the DNA extracts from these cultures also contained an

unknown quantity of host DNA. Similarly, genomic DNAs from the Davis et al. (2019) study were single

cell genomes ampliﬁed by multiple displacement methods and also contained DNA from both the fungi

and their host organisms. Species of Kickxellales are saprotrophic and were grown axenically (Benjamin,

1958). A preliminary test of ITS1F/ITS2 primers on DNA from Piptocephalis andSyncephalis species mixed

with soil DNA indicated that increased ITS1 length in these taxa resulted in reduced ampliﬁcation and

sequencing (SUPP FIG 1). We included those taxa and additional taxa with known length variation in our

mock community to further examine these potential biases across diﬀerent markers. The ﬁnal community

contained 30 isolates of Zoopagomycota fungi (TABLE 1). We also generated reference Sanger sequences from

individual mock community members using primer pair LROR/LR5 (Hopple and Vilgalys 1994; Vilgalys

and Hester 1990) for LSU and primer pair ITS1F/ITS4 (Gardes & Bruns, 1993; White et al., 1990) for

ITS. Reference sequences were veriﬁed by BLAST analysis against NCBI GenBank and with phylogenetic

reconstruction (data not shown). A non-biological, equimolar DNA mock community which consisted of a

mixture of 12 synthetic single-copy sequences (SYNMO) was included alongside the ITS1 samples to help

detect index bleed between samples and evaluate bioinformatic parameters (Palmer et al., 2018). All OTUs

recovered from the mock community samples were submitted for BLAST searches to assess the taxonomic

identity of the OTUs for comparison against the bioinformatic output.

Environmental samples –We collected environmental samples from ﬁve sites in CA and two in FL (SUPP

TABLE 1). At each site ﬁve substrates were collected: 1) bulk water from a freshwater stream or pond

(water), 2) saturated sediment from the edge of the water body (mud), 3) the upper soil layers and leaf

litter, consisting of the visible organic layer (topsoil), 4) the mineral soil layers below the topsoil (deep soil),

and 5) microinvertebrates collected from the soil samples using Baermann funnels (invertebrates). At each

site ﬁve replicates of each sample type were collected. The topsoil and deep soil replicates were collected at

increasing distances from the water source along a 25 m transect (i.e. sample 1 was closest to the water and

soil sample 5 was furthest from the water). For each sample approximately 15-25 mL of soil was collected

into sterile 50 mL tubes and ﬁlled with sterile 2x Cetyl Trimethyl Ammonium Bromide lysis buﬀer (CTAB)

to 30-35 mL. Water samples were collected in 950 mL sterilized Mason jars by dipping the jar into the water

from the embankment. Water samples included some sediment and debris present on the bottom or ﬂoating

on the top of the water. Large debris such as sticks, rocks, and clumps of leaves were removed. Vacuum

ﬁltration and a sterile B¨uchner funnel were used to ﬁlter the water through ﬁlter paper (6 μm pore size).

After ﬁltration, the ﬁlter papers were immediately placed in sterile 50 mL tubes ﬁlled with CTAB. No water

samples were collected from the “Sweeney wash” site in California because there was no standing water

although mud samples were collected from a wet depression between rocks. Microinvertebrates (protists,

nematodes, tardigrades, etc.) were collected using Baermann funnels. A 50/50 mixture of topsoil and deep

soil was added to a funnel, ﬁlled with sterilized water, and covered with Paraﬁlm. One mL of water was

collected in a sterile tube after 24 hours incubation and stored in a -20 C freezer. A second mL of water

was collected during the second 24-hour period. The two samples were centrifuged at 15000 x g for 15

minutes, the excess water was drained, and then the samples were combined into one tube with CTAB for

DNA extraction. For the three “Sweeney” desert sites in California, three rather than ﬁve Baermann funnel

replicates were obtained due to limited funnel availability. Five microinvertebrate replicates were performed

for all other sites. Baermann funnel samples are enriched for potential hosts of Zoopagales parasites and

were used to target these taxa.

DNA Extraction –DNA extraction followed a modiﬁed CTAB protocol (Gardes & Bruns, 1993). Topsoil,

deep soil, and ﬁlter papers from the water samples were subjected to several cycles of freezing and thawing

in 50 mL tubes with CTAB prior to DNA extraction. Several sterilized glass beads were then added, and

samples were shaken for 1 minute at 1500 RPM in a 1600 MiniG tissue homogenizer (SPEX, Metuchen, NJ,

USA). Two mL of CTAB was collected from each 50 mL tube and placed in sterile microcentrifuge tubes.

Samples in CTAB were incubated with a 1:1 mixture of phenol:chloroform overnight and then washed with

an additional chloroform step. The remainder of the CTAB protocol was performed without modiﬁcation.

Extractions from cultures followed the methods of Reynolds et al. (2019) and also used the CTAB protocol.

Following extraction, DNA yield was estimated by a Nanodrop 2000 spectrophotometer (ThermoFisher

Scientiﬁc, Waltham, MA, USA). Samples with genomic DNA concentrations >200 ng/ul were further cleaned

with the DNeasy PowerClean Pro cleanup kit (Qiagen, Germantown, MD, USA); samples with low DNA

yield (<200 ng/uL) were not cleaned due to loss of DNA during the clean-up process.

Primer Selection –A multiple sequence alignment containing Dikarya, Mucoromycota, and Zoopagomycota

sequences from GenBank, as well as the newly sequenced Zoopagomycota species, was created in Mesquite

(Maddison & Maddison, 2019) and aligned with MUSCLE (Edgar, 2004). This reference alignment was

used to evaluate the number of mismatches between published primer sequences and Zoopagomycota fungi.

After testing several modiﬁed primer combinations on four samples, we chose the primers ITS1F and ITS2

(White et al., 1990) for the ITS1 region for further comparisons. For LSU, we chose the LROR/LR3 (Hopple

& Vilgalys, 1994) primer set, which has been successfully used in metagenomics studies by processing only

the forward reads (Benucci et al., 2019; Bonito et al., 2014; Johansen et al., 2016). We also included a

modiﬁed forward primer LR22F (5´-GAGACCGATAGHRHACAAG-3´) used in combination with reverse

primer LR3. LR22F is the reverse compliment of primer LR22 to which we added three degenerate positions

to maximize compatibility with EDF. This primer is similar to LR22R (Mueller et al., 2015, 2016) but shifted

upstream 8 bp because target Zoopagomycota taxa in our alignment had mismatches to LR22R (SUPP FIG

2). Hereafter we refer to these primer sets by the forward primer: ITS1F, LROR or LR22F.

Library Preparation –We prepared the ITS1F Illumina library using the thermocycling protocols of Truong

et al. (2019). Brieﬂy, we used Phusion high ﬁdelity polymerase (ThermoFisher Scientiﬁc, Waltham, MA,

USA) and a dual-indexing approach with the following modiﬁcations. Ampliﬁcation and sequencing with

LROR were the same as for ITS1F except that the thermocycling program for the initial ampliﬁcation step

was 95°C for 1 minute, step down of -0.1°C per cycle from 55°C, and 72°C for 1:30 with a total of 30 cycles

and a ﬁnal elongation step of 72°C for 10 minutes. Temperature gradient tests were performed for LR22F

to determine the best annealing temperature. The thermocycling program for LR22F was the same as for

LROR except that the annealing temperature was a step down of -0.2°C per cycle starting from 65°C. All

samples were ampliﬁed in three separate replicates and the replicates were combined prior to the indexing

reaction. Negative PCR controls were included in all reactions. Amplicons from both the initial ampliﬁcation

and index attachment reaction were veriﬁed on 1.5% agarose gels. Indices were added using the Nextera XT

index kit v2 (Illumina, San Diego, CA, USA) in a separate PCR step with the following protocol: GoTaq

Green master mix (Promega, Madison, WI, USA) was used for the reaction and the thermocycling program

included nine cycles at 94°C for 1 minute, 55°C for 1 minute, 72°C for 1:30, and a ﬁnal elongation step of 72°C

for 10 minutes. Indexed PCR products were pooled into groups of three based on similar band intensity and

then cleaned with the Select-a-Size DNA Clean and Concentrator kit (Zymo, Irvine, CA, USA). All samples

were then quantiﬁed using a Qubit 4.0 Fluorometer (Invitrogen, Waltham, MA, USA), equilibrated, and

combined for the ﬁnal library. Libraries were further puriﬁed with the Agencourt AMPure XP kit (Beckman

Coulter, Brea, CA, USA) to remove primer dimers and then sequenced with the Illumina MiSeq 300 bp

PE protocol using V3 chemistry (Illumina, San Diego, CA, USA) at either the UF Interdisciplinary Center

for Biotechnology Research or the UC Riverside Institute for Integrative Genome Biology. Raw data are

available at NCBI’s Sequence Read Archive (BioProject PRJNA660245).

Bioinformatics – All sequencing data were processed with the AMPtk pipeline v 1.4.1 (Palmer et al., 2018).

Up to two nucleotide mismatches were allowed in each primer, a maximum of two expected errors were

allowed during demultiplexing and quality ﬁltering, the maximum length was set at 600 bp, a standard

index bleed of 0.5% was set, and reference-based chimera ﬁltering was used during clustering. The 600 bp

maximum length refers to the truncation of reads for downstream processing in AMPtk. Reads are only

truncated if they are above the length cutoﬀ. We tested several diﬀerent length cutoﬀs and found that 600

bp returned the best results for consistent contig formation. For the ITS1F data, a minimum length of 150

bp was used, the clustering threshold was 97% using the UNOISE3 (Edgar 2010) algorithm, and the built-in

UNITE ITS database in AMPtk was used for taxonomic identiﬁcation via the hybrid method. The ITS

OTUs were also assigned taxonomy with the RDP classiﬁer for comparisons across primer sets. The LSU

data from both primer sets were processed with a minimum length of 225 bp and a clustering threshold

of 98% using the UNOISE3 method. Because some fungal species have ITS1 sequences under 200 bp, a

shorter minimum length was used for ITS1F than LSU. A custom LSU database was installed in AMPtk for

taxonomy assignment (see below) because the built-in LSU database was outdated (RDP training set 8) and

no other comprehensive LSU database compatible with AMPtk was available. The clustering thresholds for

the LROR and LR22F datasets were determined by evaluating which mock community output best matched

the true community composition after processing the data at diﬀerent cutoﬀs (97, 98, and 99% for both

primer sets) and comparing UNOISE3 against DADA2 (Callahan et al., 2016). For both LSU data sets,

UNOISE3 was used instead of DADA2 because DADA2 doubled the number of OTUs in the output and

did not signiﬁcantly improve recovery of the mock community. Additionally, using UNOISE3 enabled direct

comparisons between datasets generated with the three primer sets. Any remaining OTUs detected in the

negative controls after ﬁltering were deleted from the OTU tables.

To create the LSU database, the RDP (training set 11) (Liu et al., 2012; Wang et al., 2007) and SILVA

(LSU Ref 132) (Quast et al., 2013) fungal LSU reference FASTA ﬁles were downloaded, concatenated, and

the taxonomy strings reformatted for use in AMPtk. Reference sequences from Zoopagomycota in the mock

community were added to the database along with >200 sequences of animals, fungi, plants, and protists

from GenBank. We selected these additional sequences based on BLAST results of OTUs in the dataset that

were initially unidentiﬁed or misidentiﬁed by the databases. Due to limitations in formatting, the database

is only compatible with the UTAX or DADA2 taxonomy assignment methods in AMPtk. The ﬁnal updated

database consisted of 115,382 dereplicated sequences and is freely available on OSF (https://osf.io/cz3mh/).

We refer to this modiﬁed database as RDP+SILVA, and the program used for taxonomy assignment was

UTAX. All three OTU datasets were also assigned taxonomy with the RDP classiﬁer v 2.12 (Wang et al.,

2007) executed in AMPtk using the ITS UNITE and LSU training sets, a commonly used methodology for

metabarcoding studies.

Statistical analyses – Once OTU tables were obtained, further analyses were conducted in R (R Core Team,

2019) with scripts available at OSF (https://osf.io/cz3mh/). To check sampling coverage, rarefaction plots

were made using the package iNEXT (Chao et al., 2014; Hsieh et al., 2020). Community comparisons were

performed on the subset of OTUs identiﬁed as EDF and compared to all fungi. To visualize community

diﬀerences among primer sets, the OTU table was converted into presence/absence format and Bsim dis-

similarity matrices were calculated with the betadiver function in the vegan package (Oksanen et al., 2019)

using the “w” method (Koleﬀ et al., 2003). Non-metric multidimensional scaling (NMDS) ordinations were

performed with the metaMDS function in vegan and plotted using ggplot2 (Wickham, 2016). These ordina-

tions were visualized for the entire fungal community and on subsets of the EDF OTUs. NMDS ordinations

were performed for each primer set separately and the scores were extracted as dataframe objects which

were then overlayed onto a single plot without modiﬁcation of the coordinates. To compare the eﬀect of the

sampling location and sample type on fungal and EDF community composition, the dissimilarity matrices

were analyzed with permutational multivariate analysis of variance (PERMANOVA - Anderson, 2001) of the

Bsim distance matrices (Anderson et al., 2006) using the adonis2 function in the vegan package. Because

PERMANOVA tests are sensitive to dispersion of the samples within groups, an ANOVA analysis of the

Bsim distance matrices was conducted to test for signiﬁcant diﬀerences between groups of interest. Signi-

ﬁcant ANOVA results of Bsim output indicate that dispersion between groups may confound results from

PERMANOVA analyses. Centroid diﬀerences were used for Bsim (type = “centroid”), and both Bsim

and PERMANOVA used 9,999 permutations. To further examine similarities between primer sets, Mantel

tests of correlation between the dissimilarity matrices of each primer set were conducted using Spearman’s

rank correlation and 999 permutations. Finally, taxonomy bar plots and alpha diversity metrics for all fungi

and the subset of EDF were conducted in phyloseq (McMurdie & Holmes, 2013) or the R base functions.

Alpha diversity measures were evaluated for fungal community diﬀerences based on site and sample type

with ANOVA and Tukey’s Honest Signiﬁcant Diﬀerences using the agricolae package (de Mendiburu, 2020).

Mantel tests of fungal OTUs utilized a subset of the data that contained only the 127 samples (118 for EDF)

that worked for all three primer sets.

We selected 50 OTUs from the LROR dataset and 50 OTUs from the LR22F dataset to study using phy-

logenetic analyses. We selected those that had the greatest number of reads, were detected in more than

one sample, and were identiﬁed only as kingdom Fungi with our pipeline (i.e. the UTAX algorithm and

RDP+SILVA database). The OTUs were also subjected to BLAST searches and OTUs with high matches to

non-fungal sequences were not included. These 100 unknown OTUs were added to a sequence alignment of

Zoopagomycota, other EDF taxa, and additional Dikarya. As a check on the taxonomic identiﬁcation by our

pipeline, OTUs classiﬁed within the Zoopagomycota (Kickxellomycotina or Zoopagomycotina) were included

in a smaller alignment and processed the same way as the larger alignment. The sequences were aligned with

MUSCLE, ambiguously aligned regions were excluded, and Maximum Likelihood analyses were performed in

RAxML v 8 (Stamatakis, 2014) using the GTRGAMMA model and 1,000 bootstraps. The unknown OTUs

were also examined using BLAST searches against GenBank using both default parameters and excluding

uncultured/environmental sample sequences and the results were compared to the placement of the OTUs

in the phylogeny. Resulting ﬁgures were modiﬁed in FigTree v 1.4.3 (Rambaut, 2012) and InkScape v 0.92.2

(https://inkscape.org/en/).

RESULTS

Comparisons of sequencing and bioinformatic processing–The ITS1F dataset included 146 samples (105

CA, 42 FL), had the most reads (14.2 million), and resulted in 7,609 OTUs. The LR22F dataset included

142 samples (106 CA, 36 FL), had the fewest reads (7.4 million), and resulted in 10,028 OTUs. The LROR

dataset included 144 samples (107 CA, 37 FL), had an intermediate number of reads (9.1 million) and

resulted in 5,786 OTUs (FIG 1). The ITS1F dataset had the highest percentage of reads discarded due to

primer incompatibility (6.37%) while the LR22F set had the lowest percentage of primer incompatibility

(0.38%), but the highest percentage of reads discarded due to short length (5.76%). Data from the LR22F

primer set were mostly assembled into contigs that used both forward and reverse reads (FIG 1). In contrast,

pairing between forward and reverse reads for ITS1F data was widely variable, whereas reads from the LROR

primer set generally could not be compiled into contigs. This resulted in the majority of LROR OTUs being

comprised of only forward reads (283 bp). Rarefaction plots indicate suﬃcient sampling from each state for

each primer set (SUPP FIG 3).

The proportion of fungal OTUs was diﬀerent for each of the three primer pairs and the two taxonomy

assignment methods (FIG 2, SUPP FIG 4). The ITS1F dataset recovered the most fungal OTUs followed

by LROR and LR22F according to the hybrid (for ITS1F) and UTAX (for LSU) methods of taxonomy

assignment. The RDP classiﬁer assigned 100% of ITS1F OTUs to Fungi but assigned a greater proportion

of LROR and LR22F OTUs to non-fungal groups than the UTAX method (47.2-54% RDP vs 38.8-33.2%

UTAX) (FIG 2 C). The majority of fungal reads from all three datasets were assigned to Ascomycota followed

by Basidiomycota but the three datasets diﬀered in the proportion of sequences that were assigned to EDF

groups (FIGS 2, 4). The maximum OTU length was similar for all primer sets (ITS1F 534 bp, LR22F 548

bp, LROR 546 bp). Correlation plots of OTU length versus read number indicate that these variables are

signiﬁcantly negatively related for ITS1F, but not for the LSU datasets (SUPP FIG 5). Finally, the ITS1F

dataset had the lowest percentage of unidentiﬁed OTUs with 1.9%, followed by LROR with 23.8%, and

LR22F with 35.5% using the hybrid and UTAX methods, respectively. The RDP classiﬁer recovered fewer

unidentiﬁed OTUs with 0% for ITS1F, 18.3% for LR22F, and 24.1% for LROR.

Within Fungi, several orders of Dikarya were dominant across all primer sets and samples (SUPP FIG 6;

SUPP TABLE 3). The greatest diﬀerences between primer sets were among less OTU-rich groups that

were recovered by only one or two of the three primer sets (FIG 3; SUPP FIG 6). The LROR and LR22F

datasets recovered more EDF OTUs from Blastocladiales, Calcarisporiellales, Chytridiales, Cladochytriales,

Endogonales, Entomophthorales, Gromochytriales, Monoblepharidales, Neocallimastigales, Zoopagales, and

Microsporidia than ITS1F (FIG 3). The LROR and LR22F datasets also recovered more OTUs from fungal-

like organisms in class Oomycetes (kingdom SAR) and slime molds in Physariida (kingdom Amoebozoa).

In other cases, the ITS1F dataset had several times the number of OTUs than either of the LSU datasets.

For example, Rozellomycota was the dominant taxon among EDF for ITS1F, but Chytridiomycota was

dominant for both LSU markers (FIG 3). Likewise, the ITS1F dataset had 36 OTUs assigned to Kickxellales

compared to ﬁve or fewer for each of the LSU primer sets. This mirrors the inﬂation of Kickxellales in the

mock community in the ITS1F dataset (see below).

Comparison of mock communities –Inspection of the ITS1F mock community shows some likely errors in

the OTU taxonomic assignment (TABLE 1). The AMPtk hybrid taxonomy assignment method identiﬁed

two ITS1F OTUs as Stramenopiles (SAR), but each had BLAST matches to Acaulopage (Zoopagales), a

mock community member (73% coverage, 91.5% identity and 72% coverage, 78.08% identity). The OTU

identiﬁed as Dimargaris cristilligena had a BLAST match of 100% coverage and 97.4% identity to Cokero-

myces recurvatus,the host of this mycoparasite. Five other OTUs found in the ITS1F Zoopagomycota mock

community were identiﬁed as Basidiomycota, Metazoa, or Haptista. These Dikarya and non-fungal OTUs

were not found in the negative controls and they were not returned in the mock communities sequenced

with the LSU primers. These OTUs could have been ampliﬁed from the mixed genomic DNA present in

the non-axenic mock isolates. The ITS1F hybrid dataset also returned 12 OTUs assigned to Kickxellales in

the mock community whereas only nine taxa were actually included. Similarly, the ITS1F dataset had only

three Zoopagales OTUs even though 21 isolates were originally added. The RDP classiﬁer method identi-

ﬁed 5 putative Mucoromycota host fungi OTUs and one Kickxellales OTU from the mock community. The

remaining 21 OTUs were classiﬁed only to kingdom Fungi by RDP.

Mock community recovery was more accurate with the LSU datasets using the UTAX taxonomy assignment

and RDP+SILVA database than the ITS1F dataset (TABLE 1). Both LSU primer sets recovered almost all

the members of the mock community except that the LR22F dataset identiﬁed one lessCoemansia OTU. Both

LSU primer sets also recovered OTUs assigned to putative Mucoromycota host fungi of the mycoparasites

included in the mock. Comparison of UTAX against the RDP+SILVA LSU database to the RDP classiﬁer

shows that few of the mock community members were identiﬁed by RDP (TABLE 1). Many mock isolates

were classiﬁed as Metazoa or remained unclassiﬁed with RDP but were accurately identiﬁed as fungi by

UTAX and the RDP+SILVA LSU database. Across all primer sets, the Zoopagales mock members were un-

derrepresented with several isolates remaining undetected by all three primer sets. There are multiple possible

reasons these taxa did not amplify or sequence well: 1) we found that Acaulopage dichotoma ,A. tetraceros

, andStylopage species lack part of the ITS2 priming site (see alignment ﬁles at https://osf.io/cz3mh/), 2)

ampliﬁcation competition from shorter host DNA fragments present in the community, and/or 3) these fungi

may have uncharacterized sequence features that reduce ampliﬁcation, such as high G:C content (Dutton et

al., 1993).

Ordination plots and statistical analyses – Community analyses are based on the subset of OTUs identiﬁed

as Fungi or the subset of OTUs identiﬁed as EDF. The ITS1F primer set recovered 6,140 fungal OTUs, 1,757

of which were EDF, LROR recovered 2,095 fungi (369 EDF), and LR22F recovered 3,126 fungi (447 EDF).

There were no observable diﬀerences in the fungal community recovered between primer sets based on the

grouping of samples in the NMDS plots (FIG 4) (individual plots for each primer set separately are given

in SUPP FIG 7). However, fungal communities from CA and FL were distinct with additional partitioning

based on sampling site. The Evey Canyon and San Jacinto sites clustered separately from the Mojave Desert

Sweeney sites in CA, whereas the two FL sites overlapped (FIG 4B, D). For the EDF, CA site clusters

were separated in the ITS1F dataset, but those communities overlapped in the LSU datasets (FIG 4 A).

PERMANOVA analyses found signiﬁcant eﬀects of site and sample type on the fungal communities among

each of the primer sets (TABLE 2), but the eﬀect size (R2) was small. Generally, sample type had a greater

eﬀect on EDF and fungal communities from FL, whereas site had a stronger impact on communities in

CA. The ANOVA results of theBsim matrices for CA were all signiﬁcant except for ITS1F site (fungi) and

sample (EDF). Mantel test comparisons between primer sets were all signiﬁcant for both fungi and EDF,

indicating that the recovered communities were signiﬁcantly correlated between the three datasets (TABLE

3). The correlation (R) was less than 0.50 for fungal ITS1F vs. LROR and LROR vs. LR22F and less than

0.60 for all EDF comparisons. The highest Mantel R statistic was for fungal ITS1F vs. LR22F, with a value

of 0.71.

Plots of fungal richness between each primer set, site, and sample type showed that overall fungal richness

was similar across all three primer sets, but soil samples generally had the highest diversity (SUPP FIG 8).

Evey Canyon and San Jacinto had among the highest OTU richness for all sample types across all three

primers. Fungal richness was lower for invertebrate samples across all primer sets but was higher for ITS1F

than either LSU primer set. However, EDF diversity patterns diﬀered. For ITS1F, many water, mud, and

invertebrate samples had greater EDF diversity than soil (FIG 5). Conversely, soil and mud from Sweeney

sites in the Mojave Desert had among the highest EDF diversity for both LSU markers (FIG 5). Tukey’s test

results varied for EDF and all fungi and by primer (SUPP TABLE 4) and separation of EDF communities

by sample type and site are also observable in the NMDS plots (SUPP FIG 7).

Phylogenetic reconstruction of LSU OTUs –The 50 LROR and 50 LR22F OTUs identiﬁed only as “Fungi”

were added to the sequence alignment that included 436 taxa. After exclusion of ambiguous sites, the LSU

alignment contained 1,104 characters and OTUs had a ﬁnal length of 216-239 bp for LR22F and 186-230 bp

for LROR. Figure 7 shows the phylogeny and SUPP TABLE 5 lists the LSU OTUs used in the phylogeny and

BLAST results for each OTU. Backbone nodes of the phylogeny were mostly unsupported whereas nodes near

the tips had higher bootstrap support (>70). Both primer sets recovered monophyletic clades of OTUs that

had BLAST matches to protist sequences (SUPP TABLE 5), but the LR22F dataset had fewer than LROR

(FIG 6). The majority of these LROR OTUs had higher BLAST identity scores to the protist sequences

than the LR22F OTUs, a maximum of 81% coverage and 100% identity for LROR, but only 20% coverage

and 98.86% identity for LR22F. The LROR OTUs that had BLAST matches to protists were resolved in two

diﬀerent clades, one with three OTUs that had matches to Stramenopiles, and a larger clade with matches

to Rhizaria. The LR22F protist clade was nested within the Chytridiomycetes, but all the matches (except

the one mentioned above) had identity scores in the 75-78% range. Most fungal OTUs from both primer

sets were resolved in the Chytridiomycetes and Orbiliomycetes. The OTUs placed in the Orbiliomycetes

had close BLAST matches to Orbiliomycetes. In contrast, many OTUs placed in the Chytridiomycetes had

matches to other fungal orders. Only LROR OTUs were placed within Aphelidiomycetes, and clades in the

Basidiomycota, Eurotiomyctes, Umbelopsidomycetes, and Zoopagomycetes. The Archaeorhizomycetes, En-

dogonomycetes 1, and Sordariomycetes contained LR22F OTUs but none from LROR. The Glomeromycetes

contained four LROR OTUs and one LR22F OTU. In the Zoopagomycota-only phylogeny, the two subphyla

are recovered as polyphyletic, contrary to other studies (Davis et al., 2019), with the Dimargaritales and

Ramicandelaber nested within the Zoopagomycotina (FIG 7). Additionally, the Harpellales are sister to the

two subphyla rather than nested within Kickxellomycotina as found by other studies (Wang et al., 2019).

Only OTU 6654 (LR22F) did not place in a clade matching its taxonomic classiﬁcation from the taxonomy

pipeline.

DISCUSSION

Taxonomic assignment of OTUs – Recovery of EDF communities is impacted more by methodological choi-

ces during metabarcoding sampling than Dikarya communities. In particular, target fragment length and

taxonomy assignment method can each have a profound impact on EDF detection. We found considera-

ble diﬀerences in taxonomic identiﬁcations between taxonomy methods for all primer sets. This was even

true at the Kingdom level where LSU OTUs were more likely to be identiﬁed as non-fungal by the RDP

classiﬁer (47-54%) compared to our RDP+SILVA database (33-38%). The ITS1F dataset had the greatest

number of OTUs identiﬁed as Fungi. This result was expected due to diﬀerences in primer speciﬁcity (i.e.

the ITS1F/ITS2 primer combination is more fungal-speciﬁc but has known biases for Dikarya – Bellemain

et al., 2010; Tedersoo et al., 2015) and the completeness of the ITS versus LSU databases. Because the ITS

databases have >1,000,000 reference sequences compared to <200,000 for LSU databases, we expected more

accurate taxonomic assignment for ITS1F OTUs.

In the mock community analyses, both LSU datasets more closely recapitulated the community than ITS1F

(TABLE 1). Errors in the taxonomy assignment of ITS1F OTUs from the mock community indicate mi-

sidentiﬁcation of reference sequences (e.g. Dimargaris ) and also a lack of reference sequences for some taxa

(e.g. Zoopagales) (TABLE 1). Correlation plots of fungal OTU length and taxonomic identity score had

signiﬁcant negative relationships for the LSU primer sets (SUPP FIG 9), but a positive relationship for

ITS1F. The negative relationship for LSU likely reﬂects the greater representation of the shorter length refe-

rence sequences from Dikarya species in the databases than longer non-Dikarya. However, while there was a

negative relationship between OTU length and read number for ITS1F, there was no signiﬁcant correlation

for the LSU datasets (SUPP FIG 5). This implies that taxonomic representation in the reference databases

rather than OTU length has a greater impact on taxonomy scores of LSU OTUs. However, our RDP+SILVA

LSU database substantially improved identiﬁcation of Zoopagomycota isolates in the mock community com-

pared to the RDP classiﬁer (TABLE 1) with ≤10% of mock members identiﬁed with RDP versus 63% with

RDP+SILVA. In contrast, ITS1F recovered 27% of the mock members using the hybrid taxonomy method.

It is important to note, however, that populating our RDP+SILVA database with additional non-fungal se-

quences from GenBank improved the accuracy of determining fungal versus non-fungal sequences compared

to RDP. The supported placement of Zoopagomycota OTUs in our reduced phylogeny also reinforces the

accuracy of the taxonomic classiﬁcations made by our pipeline (FIG 7). Therefore, a robust reference data-

base should include a diversity of eukaryotic sequences, especially for LSU because primers for this region

are often less fungal-speciﬁc.

These results demonstrate the profound eﬀect that reference databases have on the classiﬁcation of OTUs.

Other taxonomic assignment approaches, like BLAST followed by MEGAN (Huson et al., 2011) or the

Statistical Assignment Package (Munch et al., 2008), have the potential to improve LSU OTU identiﬁcation

(Porter & Golding, 2012). However, many of our unidentiﬁed OTUs had best matches to other unidentiﬁed

sequences in GenBank (e.g. SUPP TABLE 5) or best matches that did not reﬂect the phylogenetic placement

of the OTU, indicating that BLAST and reference-based methods cannot completely alleviate identiﬁcation

problems (L¨

ucking et al., 2020). The disparity between named fungal species and unnamed environmental

sequences is substantial and available data suggest that many unidentiﬁed sequences represent EDF, including

Zoopagomycota (Lazarus & James, 2015; Tedersoo et al., 2017, 2020). Even within a relatively small order,

Zoopagales, 13 out of 22 genera lack any sequence data (Davis et al., 2019b). Although much has been done

to improve fungal LSU databases (e.g. Vu et al., 2019; Hanafy et al., 2020), further additions and curation

can bring LSU on par with ITS as an rDNA metabarcoding marker. Until intensive eﬀorts are made to

curate and ﬁll the taxonomic gaps within databases, it is clear that taxonomic assignment issues will be

problematic irrespective of improvements in sequencing and bioinformatics. For example, the long reads and

simultaneous sequencing of multiple rDNA markers at once oﬀered by PacBio technology was not able to

entirely overcome the pitfalls of reference-based taxonomic assignment (Furneaux et al., 2021; Heeger et al.,

2018).

Primer comparison – Comparisons between metabarcoding datasets from diﬀerent studies are challenging

due to variable methods of sample collection, PCR ampliﬁcation, sequencing, and bioinformatics. However,

our results using data from >127 environmental samples support the broad pattern of fungal community

congruence between ITS and LSU markers as evidenced by signiﬁcant Mantel tests (TABLE 3) and found

by others (Amend et al., 2010; Benucci et al., 2019; Bonito et al., 2014; Brown et al., 2014; Johansen et al.,

2016; Mota-Gutierrez et al., 2019; Nelson & Shaw, 2019; Skelton et al., 2019; Xue et al., 2019). Our NMDS

plots (FIG 4; SUPP FIG 7) demonstrate that the ITS1F, LROR, and LR22F datasets recovered similar

fungal communities and detected mostly the same OTU-rich lineages (SUPP FIG 6). However, there were

discrepancies among rarer groups, like some chytrid orders for which more LSU OTUs were detected than

ITS1F (FIG 3). Similarly, Nelson and Shaw (2019) and Benucci et al., (2018) reported greater numbers of

Chytridiomycota OTUs with LROR than ITS. Conversely, the ITS1F marker inﬂated the number of OTUs

assigned to the Kickxellales in the mock community; we originally added nine taxa but recovered 12 OTUs.

This could be a result of biological (e.g. intraspeciﬁc rDNA copy number variation) and/or methodological

(e.g. inappropriate OTU clustering identity threshold) factors. Artiﬁcial inﬂation of the number of ITS

OTUs has been shown for various taxa in mock communities in metabarcoding studies (Casta˜no et al.,

2020; De Filippis et al., 2017; Jusino et al., 2019; Nguyen et al., 2015; Vˇetrovsk´y et al., 2016). These

results underscore the importance of mock communities for detecting methodological errors and reﬁning

bioinformatic parameters such as clustering thresholds (Caporaso et al., 2011; Palmer et al., 2018; Taylor et

al., 2016). Furthermore, taxonomic assignments of ITS OTUs cannot be tested with phylogenetic analyses.

Thus, groups with dramatic diﬀerences in representation between markers (e.g. Rozellomycota in our dataset,

FIG 3) cannot easily be evaluated for accuracy.

We further aimed to test LSU primer pairs for their ability to amplify fungi broadly, and Zoopagomycota

fungi speciﬁcally, as well as compare their performance in the bioinformatics pipeline. The LR22F dataset

had more fungal and total OTUs than LROR and fungal diversity analyses found signiﬁcant diﬀerences

between groups for LR22F not found with LROR (SUPP TABLE 4; SUPP FIG 8). Similar results were

found by Mueller et al. (2016) for the related LR22R primer, which recovered richness estimates closer to

ITS than LROR. We found that the forward and reverse reads from the LR22F dataset were consistently

paired into contigs, contrary to both the ITS1F and LROR primer sets (FIG 1). The LROR target fragments

for the mock community members were commonly >700 bp (TABLE 1), well beyond the sequenceable length

of Illumina MiSeq chemistry. The resulting data are therefore almost entirely restricted to the forward reads,

resulting in signiﬁcant data loss. However, paired reads have several advantages over unpaired reads: more

data are utilized, overlapping sequences reduce sequence errors, and longer sequences reduce problems during

OTU clustering and taxonomy assignment (Bartram et al., 2011; Truong et al., 2019). Furthermore, longer

sequences are more accurately identiﬁed using bioinformatic methods (Liu et al., 2012; Porras-Alfaro et al.,

2014; Porter & Golding, 2012). As a result, the taxonomic identiﬁcation of LROR OTUs may be less reliable

than the longer LR22F OTUs. We also found that sequence length inﬂuenced our phylogenetic analysis where

the longer LR22F OTUs were generally placed with higher resolution than shorter LROR OTUs (FIG 6).

Twenty-two of the 50 LR22F OTUs were placed in a clade that matched the fungal class of their BLAST

match, compared to only 15 of the 50 LROR OTUs. Conversely, both LROR and LR22F OTUs resolved well

in the smaller Zoopagomycota tree (FIG 7), and the majority were placed in clades that correspond to the

taxonomy assignment output from the UTAX/RDP+SILVA pipeline. These results illustrate the utility of

phylogenetic reconstruction of LSU OTUs for identifying potential divergent EDF fungal sequences as well

as sequence artifacts or taxonomic errors that need further investigation (Glass et al., 2013). For example,

the OTUs that were placed within the Orbiliomycetes (and had high BLAST matches to Orbiliomycetes)

indicate that the corresponding reference sequences in the database could be identiﬁed beyond kingdom to

increase the accuracy of the taxonomy.

Methodological biases impacting detection of EDF: the example of Zoopagomycota –We found additional

evidence that target region length (for both ITS1 and LSU fragments) strongly aﬀects the metabarcoding

process at diﬀerent steps. For ITS1F, the strongest bias putatively occurs during PCR when shorter fragments

are preferentially ampliﬁed (Casta˜no et al., 2020; Jusino et al., 2019; Palmer et al., 2018). In our initial

experiments with the ITS1F primer set, Zoopagomycota species with the longest ITS1 ([?]400 bp) were

not recovered from mixed samples. This was true despite adding DNA from those species at twice the

concentration of the “background” DNA (SUPP FIG 1). This pattern was reiterated in mock community

results (TABLE 1) where target species with the longest ITS1 ([?]400 bp) were not detected. This bias

towards short fragments is also taxonomically biased. Many species of Ascomycota, Basidiomycota, and

Mucoromycota have short ITS1 regions (<200-300 bp) (Bellemain et al., 2010; Bokulich & Mills, 2013),

with a few notable exceptions (e.g. Cantharellales may have ITS1 >1,000 bp – Feibelman et al., 1994). On

the other hand, most Zoopagomycota species we examined have an ITS1 >300 bp with signiﬁcant variation

between taxa. For example, Piptocephalis and Syncephalis species have ITS1 that ranges 300 – 800 bp

(Lazarus et al., 2017; Reynolds et al., 2019) and we have also found variation among Coemansia species

(TABLE 1). However, the Coemansia isolates had shorter ITS1 than many other mock members, which

likely contributed to their overrepresentation in the ITS1F results. Species of Harpellales have extreme

variation with an ITS1 range of 250–1,000 bp (Gottlieb & Lichtwardt, 2001). Similar length variation occurs

in the ITS2 region and in some cases ITS2 is longer than ITS1, such as in some Harpellales that have 1,100

bp for ITS2 but only 500 bp for ITS1 (Gottlieb & Lichtwardt, 2001). Although the fragment length for

LR22F is variable among Zoopagomycota, the variation is lower than for ITS1F. In our mock community,

the diﬀerence in fragment length between the longest and the shortestSyncephalis species was only 45 bp

for LR22F versus 508 bp for ITS1F. Likewise, the range among Piptocephalis species was 85 bp for LR22F

versus 247 bp for ITS1F.

Beyond sequencing and bioinformatics, the biology of Zoopagomycota must also be considered. The symbiotic

nature of Zoopagomycota fungi means that their abundance in any given sample is linked with the abundance

of their host organisms and dependent on host/parasite interactions. As a result, the distribution of these

fungi is likely patchy and highly variable through time, lowering the probability of their detection from

any single sample. Little is known about the host/parasite dynamics among Zoopagomycota parasites,

making the choice of locations and sample types for metabarcoding less straightforward. For example,

although the mycoparasitic species can be isolated from soil, they also are frequently isolated from dung.

Are these species mainly coprophilic, or do they actively attack hosts in the soil environment as well?

How long do their spores persist in the soil? Similarly, Zoopagales species that attack microinvertebrates

have been isolated from wet substrates like moss and decaying plant material (Drechsler 1938; Duddington

1955). We attempted to concentrate potential hosts of these fungi and thereby increase our chance of

detecting them by using Baermann funnels. However, these invertebrate samples were generally dominated

by Chytridiomycota OTUs followed by Blastocladiomycota and Mucoromycota (FIG 3). A small number

of Zoopagales OTUs were detected from other sample types by each primer set, and phylogenetic analyses

supported their identiﬁcation (FIG 7). Nonetheless, the number of OTUs detected is still less than the

number of isolates recovered from these sites using specialized culturing techniques. For example, Benny

et al., (2016) found that 46% of 520 soil samples (mostly from Florida) contained at least one Syncephalis

species, but we only recovered one Syncephalis OTU from two samples using the LROR primer set. Likewise,

the spores of Harpellales fungi are thought to pass between hosts through transmission in the water column

(Lichtwardt 1986). However, we were unable to detect Harpellales OTUs from water or mud samples,

which were mostly dominated by chytrid OTUs. Species of Kickxellales grow axenically and can be isolated

from soil or dung, but their trophic modes remain unclear. Although Kickxellales fungi are assumed to be

saprotrophic, there are reports of some species growing on other fungi or in association with arthropods

and many species exhibit fastidious growth in culture (Jackson & Deardon, 1948; Linder 1943; Kurihara et

al., 2001, 2008). Furthermore, most species are assumed to be rare, but the dearth of reports could be an

artifact of undersampling. Our results demonstrate that Kickxellales can be detected from soil (FIG 3) and

phylogenetic analyses indicate the clade may be more diverse and more widely distributed than currently

recognized (FIG 7).

The combined eﬀects of methodological biases and environmental sample heterogeneity (with symbiotic

Zoopagomycota likely having lower abundance) may have a synergistic impact on metabarcoding outcomes,

leading to artiﬁcially inﬂated representation of some groups and absence of others. These biases are rooted

in methodology and can aﬀect any group of organisms with highly variable target fragment lengths and/or

poor reference sequence representation. Casta˜no et al. (2020) found such a pattern in mock communities

where longer fragments added in unequal proportions to shorter fragments were severely underrepresented

in Illumina MiSeq results while shorter fragments could be highly overrepresented (up to 57% greater output

than input). Although PacBio RS II sequencing produced less severe discrepancies, a length bias was still

detected (Casta˜no et al., 2020), an eﬀect that can also be inﬂuenced by sample loading method (Tedersoo

et al., 2017b). Our results demonstrate that both PCR length bias and lack of reference sequences severely

impact the detection of Zoopagomycota from mixed samples. Other lineages of EDF are similarly involved

in symbiotic associations and lack reference sequence data (e.g. Blastocladiomycota, Rozellomycota), indi-

cating they are likewise often missed or misidentiﬁed in metabarcoding studies. Many unanswered questions

about Zoopagomycota remain, such as their roles in microbial food webs and their eﬀect on host popula-

tions. Metabarcoding has the potential to help unravel some of these mysteries but novel approaches are

needed to overcome methodological biases, such as PCR-free on-array hybrid capture (e.g. Mamanova et al.,

2010). Combined with targeted culturing approaches, improved environmental sampling methods can help

illuminate the diversity and ecological roles of “dark matter fungi” (Grossart et al., 2016).

Acknowledgements

This work was supported by the National Science Foundation (DEB 1441677 to MES; DEB 1441715 to

JES). Additional funding was provided by the Department of Plant Pathology, the USDA-NIFA under award

number FLA-PLP-005289 and the Institute for Food and Agricultural Sciences at the University of Florida.

We thank Tim James (University of Michigan) and William J. Davis (USDA Mycology and Nematology

Genetic Diversity and Biology Lab) for sharing Zoopagales DNA samples used to generate reference rDNA

sequences and in the mock community. We also appreciate the guidance and input from Gerald Benny

(University of Florida) who provided access to the zygomycete culture collection and helped with the culture

work. The environmental sampling was performed at the University of California Natural Reserve System

James San Jacinto Mountains Reserve (doi: 10.21973/N3KQ0T) and Sweeney Granite Mountains Desert

Research Center (doi: 10.21973/N3S942) and at the University of Florida Ordway-Swisher Biological Station

(https://ordway-swisher.uﬂ.edu/) in Putnam County, Florida.

Author Contributions

NKR and MES planned the research. MES and JES provided funding for the research. NKR collected samples

from California and Florida and JES helped arrange access to reserves and collect samples from California.

NKR performed lab work with the assistance of MAJ. NKR and MAJ conceived the bioinformatics and

analyses and NKR analyzed the data with troubleshooting help from MAJ. NKR wrote the manuscript

which was thoroughly edited by MES and included input from all co-authors.

Data Availability Statement

Jupyter notebooks used for raw data processing in AMPtk, R scripts used for data analyses, and the

RDP+SILVA LSU database FASTA ﬁle are available for download from OSF (https://osf.io/cz3mh/). The

GenBank accession numbers for the Sanger sequences reported in this paper are available in Table 1. Illumina

raw sequences are available on NCBI’s sequence read archive (SRA) at BioProject PRJNA660245 for the

main dataset and PRJNA428770 for the supplemental data.

References

Abarenkov, K, Zirk, A., Piirmann, T., P¨

oh¨

onen, R., Ivanov, F., Nilsson, R., & K˜oljalg, U. (2020). Full UNI-

TE+INSD dataset for Fungi . UNITE Community. https://doi.org/https://doi.org/10.15156/BIO/786372

Abarenkov, Kessy, Somervuo, P., Nilsson, R. H., Kirk, P. M., Huotari, T., Abrego, N., & Ovaskainen, O.

(2018). Protax-fungi: a web-based tool for probabilistic taxonomic placement of fungal internal transcribed

spacer sequences. New Phytologist ,220 (2), 517–525. https://doi.org/10.1111/nph.15301

Amend, A. S., Seifert, K. A., Samson, R., & Bruns, T. D. (2010). Indoor fungal composition is geographically

patterned and more diverse in temperate zones than in the tropics. Proceedings of the National Academy of

Sciences of the United States of America ,107 (31), 13748–13753. https://doi.org/10.1073/pnas.1000454107

Anderson, M. J., Ellingsen, K. E., & McArdle, B. H. (2006). Multivariate dispersion as a measure of beta

diversity. Ecology Letters ,9(6), 683–693. https://doi.org/10.1111/j.1461-0248.2006.00926.x

Anslan, S., Nilsson, R. H., Wurzbacher, C., Baldrian, P., Tedersoo, L., & Bahram, M. (2018). Great dif-

ferences in performance and outcome of high-throughput sequencing data analysis platforms for fungal

metabarcoding. MycoKeys ,39 , 29–40. https://doi.org/10.3897/mycokeys.39.28109

Asemaninejad, A., Weerasuriya, N., Gloor, G. B., Lindo, Z., & Thorn, G. (2016). New primers for

discovering fungal diversity using nuclear large ribosomal DNA. PLoS ONE ,11 (7), 1–15. htt-

ps://doi.org/10.1371/journal.pone.0159043

Badotti, F., De Oliveira, F. S., Garcia, C. F., Vaz, A. B. M., Fonseca, P. L. C., Nahum, L. A., Oliveira, G.,

& G´oes-Neto, A. (2017). Eﬀectiveness of ITS and sub-regions as DNA barcode markers for the identiﬁcation

of Basidiomycota (Fungi). BMC Microbiology ,17 (1), 1–12. https://doi.org/10.1186/s12866-017-0958-x

Bartram, A. K., Lynch, M. D. J., Stearns, J. C., Moreno-Hagelsieb, G., & Neufeld, J. D. (2011).

Generation of multimillion-sequence 16S rRNA gene libraries from complex microbial communities by

assembling paired-end Illumina reads. Applied and Environmental Microbiology ,77 (11), 3846–3852.

https://doi.org/10.1128/AEM.02772-10

Bellemain, E., Carlsen, T., Brochmann, C., Coissac, E., Taberlet, P., & Kauserud, H. (2010). ITS as an

environmental DNA barcode for fungi: an in silico approach reveals potential PCR biases. BMC Microbiology

,10 (189). https://doi.org/10.1016/j.enzmictec.2010.03.010

Benjamin, R. K. (1958). Sexuality in the Kickxellaceae. Aliso ,4(1), 149–169. htt-

ps://doi.org/10.5642/aliso.19580401.14

Benjamin, R. K. (1959). The Merosporangiferous Mucorales. Aliso, 4 (2), 321–433.

Benjamin, R. K. (1961). Addenda to ”The Merosporangiferous Mucorales”.Aliso, 5 (1), 11-19.

Benny, G. L., Ho, H. M., Lazarus, K. L., & Smith, M. E. (2016). Five new species of the obliga-

te mycoparasite Syncephalis (Zoopagales, Zoopagomycotina) from soil. Mycologia ,108 (6), 1114–1129.

https://doi.org/10.3852/15-073

Benny, G. L., Smith, M. E., Kirk, P. M., Tretter, E. D., & White, M. M. (2016). Challenges and Fu-

ture Perspectives in the Systematics of Kickxellomycotina, Mortierellomycotina, Mucoromycotina, and

Zoopagomycotina. In D. Li (Ed.), Biology of Microfungi (pp. 65–126). Springer International Publishing.

https://doi.org/10.1007/978-3-319-29137-6

Benucci, G. M. N., Bonito, V., & Bonito, G. (2019). Fungal, Bacterial, and Archaeal Diversity in

Soils Beneath Native and Introduced Plants in Fiji, South Paciﬁc. Microbial Ecology ,78 (1), 136–146.

https://doi.org/10.1007/s00248-018-1266-1

Bokulich, N. A., & Mills, D. A. (2013). Improved selection of internal transcribed spacer-speciﬁc primers

enables quantitative, ultra-high-throughput proﬁling of fungal communities. Applied and Environmental Mi-

crobiology ,79 (8), 2519–2526. https://doi.org/10.1128/AEM.03870-12

Bonito, G., Reynolds, H., Robeson, M. S., Nelson, J., Hodkinson, B. P., Tuskan, G., Schadt, C. W., &

Vilgalys, R. (2014). Plant host and soil origin inﬂuence fungal and bacterial assemblages in the roots of

woody plants. Molecular Ecology ,23 (13), 3356–3370. https://doi.org/10.1111/mec.12821

Brown, S. P., Rigdon-Huss, A. R., & Jumpponen, A. (2014). Analyses of ITS and LSU gene regi-

ons provide congruent results on fungal community responses. Fungal Ecology ,9(1), 65–68. htt-

ps://doi.org/10.1016/j.funeco.2014.02.002

Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., & Holmes, S. P. (2016).

DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods ,13 (7), 581–583.

https://doi.org/10.1038/nmeth.3869

Caporaso, J. G., Lauber, C. L., Walters, W. A., Berg-Lyons, D., Lozupone, C. A., Turnbaugh, P. J., Fierer,

N., & Knight, R. (2011). Global patterns of 16S rRNA diversity at a depth of millions of sequences per

sample. Proceedings of the National Academy of Sciences of the United States of America ,108 (SUPPL.

1), 4516–4522. https://doi.org/10.1073/pnas.1000080107

Casta˜no, C., Berlin, A., Brandstr¨

om Durling, M., Ihrmark, K., Lindahl, B. D., Stenlid, J., Clemmensen, K.

E., & Olson, ˚

A. (2020). Optimized metabarcoding with Paciﬁc biosciences enables semi-quantitative analysis

of fungal communities. New Phytologist . https://doi.org/10.1111/nph.16731

Chao A, Gotelli NJ, Hsieh TC, Sande EL, Ma KH, Colwell RK, Ellison AM (2014). “Rarefaction and

extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies.”

Ecological Monographs, 84, 45–67.

Cole, J. R., Wang, Q., Fish, J. A., Chai, B., McGarrell, D. M., Sun, Y., Brown, C. T., Porras-Alfaro, A.,

Kuske, C. R., & Tiedje, J. M. (2014). Ribosomal Database Project: Data and tools for high throughput

rRNA analysis. Nucleic Acids Research ,42 (D1), 633–642. https://doi.org/10.1093/nar/gkt1244

Corsaro, D., K¨

ohsler, M., Wylezich, C., Venditti, D., Walochnik, J., Michel, R. (2017). New insights from

molecular phylogenetics of amoebophagous fungi (Zoopagomycota, Zoopagales). Parasitology Research, 117,

157-167. https://doi.org/10.1007/s00436-017-5685-6

Corsaro, D., Walochnik, J., Venditti, D., Steinmann, J., M¨

uller, K. D., & Michel, R. (2014). Microsporidia-

like parasites of amoebae belong to the early fungal lineage Rozellomycota. Parasitology Research ,113 (5),

1909–1918. https://doi.org/10.1007/s00436-014-3838-4

Davis, W. J., Amses, K. R., Benny, G. L., Carter-House, D., Chang, Y., Grigoriev, I., Smith, M. E.,

Spatafora, J. W., Stajich, J. E., & James, T. Y. (2019). Genome-scale phylogenetics reveals a monophy-

letic Zoopagales (Zoopagomycota, Fungi). Molecular Phylogenetics and Evolution ,133 , 152–163. htt-

ps://doi.org/10.1016/j.ympev.2019.01.006

Davis, W. J., Amses, K. R., James, E. S., & James, T. Y. (2019b). A new 18S rRNA phylogeny of uncultured

predacious fungi (Zoopagales).Mycologia ,111 (2), 291–298. https://doi.org/10.1080/00275514.2018.1546066

De Filippis, F., Laiola, M., Blaiotta, G., & Ercolini, D. (2017). Diﬀerent amplicon targets for

sequencing-based studies of fungal diversity. Applied and Environmental Microbiology ,83 (17), 1–9.

https://doi.org/10.1128/AEM.00905-17

de Mendiburu, F. (2020). Statistical Procedures for Agricultural Research Package “agricolae” (1.3-3). The

R Project for Statistical Computing. https://cran.r-project.org/package=agricolae

Drechsler, C. (1938). New Zoopagaceae Capturing and Consuming Soil Amoebae. Mycologia ,30 (2), 137–

157.

Duddington, C. L. (1955). Fungi That Attack Microscopic Animals.Botanical Review ,21 (7), 377–439.

Dutton, C. M., Christine, P., & Sommer, S. S. (1993). General method for amplifying regions of very high

G + C content. Nucleic Acids Research, 21(12), 2953–2954. https://doi.org/10.1093/nar/21.12.2953

Edgar, R. C. (2004). MUSCLE: Multiple sequence alignment with high accuracy and high throughput.

Nucleic Acids Research ,32 (5), 1792–1797. https://doi.org/10.1093/nar/gkh340

Edgar, R. C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics ,26

(19), 2460–2461. https://doi.org/10.1093/bioinformatics/btq461

Egan, C. P., Rummel, A., Kokkoris, V., Klironomos, J., Lekberg, Y., & Hart, M. (2018). Using mock com-

munities of arbuscular mycorrhizal fungi to evaluate ﬁdelity associated with Illumina sequencing. Fungal

Ecology ,33 , 52–64. https://doi.org/10.1016/j.funeco.2018.01.004

Engelbrektson, A., Kunin, V., Wrighton, K. C., Zvenigorodsky, N., Chen, F., Ochman, H., & Hugenholtz, P.

(2010). Experimental factors aﬀecting PCR-based estimates of microbial species richness and evenness.ISME

Journal ,4(5), 642–647. https://doi.org/10.1038/ismej.2009.153

Feibelman, T., Bayman, P., & Cibula, W. G. (1994). Length variation in the internal transcribed spacer

of ribosomal DNA in chanterelles.Mycological Research ,98 (6), 614–618. https://doi.org/10.1016/S0953-

7562(09)80407-3

Furneaux, B., Bahram, M., Rosling, A., Yorou, N.S., Ryberg, M. (2021). Long- and short-read meta-

barcoding technologies reveal similar spatio-temporal structures in fungal communities. Authorea , doi:

10.22541/au.160340221.18389016/v1

Gardes, M., & Bruns, T. D. (1993). ITS primers with enhanced speciﬁcity for basidiomycetes -

application to the identiﬁcation of mycorrhizae and rusts. Molecular Ecology ,2(2), 113–118.

https://doi.org/10.1111/j.1365-294X.1993.tb00005.x

Gardes, M., & Bruns, T. D. (1996). Community strucutre of ectomycorrhizal fungi in a Pinus muricata

forest: above- and below-ground views. Canadian Journal of Botany ,74 , 1572–1583.

Gazis, R., Rehner, S., & Chaverri, P. (2011). Species delimitation in fungal endophyte diversity studies

and its implications in ecological and biogeographic inferences. Molecular Ecology ,20 (14), 3001–3013.

https://doi.org/10.1111/j.1365-294X.2011.05110.x

Glass, D. J., Takebayashi, N., Olson, L. E., & Taylor, D. L. (2013). Evaluation of the authenticity of a highly

novel environmental sequence from boreal forest soil using ribosomal RNA secondary structure modeling.

Molecular Phylogenetics and Evolution ,67 (1), 234–245. https://doi.org/10.1016/j.ympev.2013.01.018

Gottlieb, A. M., & Lichtwardt, R. W. (2001). Molecular variation within and among species of Harpellales.

Mycologia ,93 (1), 66–81. https://doi.org/10.1080/00275514.2001.12061280

Grossart, H. P., Wurzbacher, C., James, T. Y., & Kagami, M. (2016). Discovery of dark matter fungi in

aquatic ecosystems demands a reappraisal of the phylogeny and ecology of zoosporic fungi.Fungal Ecology ,

19 , 28–38. https://doi.org/10.1016/j.funeco.2015.06.004

Hart, M. M., Aleklett, K., Chagnon, P. L., Egan, C., Ghignone, S., Helgason, T., Lekberg, Y., Opik, M.,

Pickles, B. J., & Waller, L. (2015). Navigating the labyrinth: A guide to sequence-based, community ecology

of arbuscular mycorrhizal fungi. New Phytologist ,207 (1), 235–247. https://doi.org/10.1111/nph.13340

Hofstetter, V., Buyck, B., Eyssartier, G., Schnee, S., & Gindro, K. (2019). The un-

bearable lightness of sequenced-based identiﬁcation.Fungal Diversity ,96 , 243–284.

https://doi.org/https://doi.org/10.1007/s13225-019-00428-3 The

Holm, S. (1979). A Simple Sequentially Rejective Multiple Test Procedure. Scandinavian Journal of Statis-

tics ,6(2), 65–70.

Hopple, J. S., & Vilgalys, R. (1994). Phylogenetic Relationships among Coprinoid Taxa and Allies Based

on Data from Restriction Site Mapping of Nuclear rDNA. Mycologia ,86 (1), 96–107.

Hsieh TC, Ma KH, Chao A (2020). iNEXT: Interpolation and Extrapolation for Species Diversity. R package

version 2.0.20, http://chao.stat.nthu.edu.tw/wordpress/software download/.

Huson, D., Mitra, S., & Ruscheweyh, H. (2011). MEGAN4 paper.Genome Research ,21 (9), 1552–1560.

https://doi.org/10.1101/gr.120618.111.Freely

Ihrmark, K., Bodeker, I. T. M., Cruz-Martinez, K., Friberg, H., Kubartova, A., Schenck, J., Strid, Y., Stenlid,

J., Brandstrom-Durling, M., Clemmensen, K. E., & Lindahl, B. D. (2012). New primers to amplify the fungal

ITS2 region - evaluation by 454-sequencing of artiﬁcial and natural communities. FEMS Microbiology Ecology

,82 (3), 666–677. https://doi.org/10.1111/j.1574-6941.2012.01437.x

Jackson, H.S., & Dearden, E.R. (1948). Martensella corticiiThaxter and its distribution. Mycologia, 40 (2),

168-176.

Johansen, R. B., Johnston, P., Mieczkowski, P., Perry, G. L. W., Robeson, M. S., Burns, B. R., & Vilgalys,

R. (2016). A native and an invasive dune grass share similar, patchily distributed, root-associated fungal

communities. Fungal Ecology ,23 , 141–155. https://doi.org/10.1016/j.funeco.2016.08.003

Jusino, M. A., Banik, M. T., Palmer, J. M., Wray, A. K., Xiao, L., Pelton, E., Barber, J. R., Kawahara, A.

Y., Gratton, C., Peery, M. Z., & Lindner, D. L. (2019). An improved method for utilizing high-throughput

amplicon sequencing to determine the diets of insectivorous animals. Molecular Ecology Resources ,19 (1),

176–190. https://doi.org/10.1111/1755-0998.12951

Koleﬀ, P., Gaston, K. J., & Lennon, J. J. (2003). Measuring Beta Diversity for Presence-Absence Data.

Journal of Animal Ecology ,72 (3), 367–382.

Kurihara, Y., Degawa, Y., Tokumasu, S. (2001). A new genusMyconymphaea (Kickxellales) with peculiar

septal plugs.Mycological Research, 105 (11), 1397-1402.

Kurihara, Y., Sukarno, N., Ilyas, M., Yuniarti, E., Mangunwardoyo, W., Park, J-Y., Saraswati, R., Widyas-

tuti, Y., Ando, K. (2008). Indonesian Kickxellales: two species of Coemansia and Linderina .Mycoscience,

49 , 250-257.

Lazarus, K. L., Benny, G. L., Ho, H. M., & Smith, M. E. (2017). Phylogenetic systematics of Syn-

cephalis (Zoopagales, Zoopagomycotina), a genus of ubiquitous mycoparasites. Mycologia ,109 (2), 333–349.

https://doi.org/10.1080/00275514.2017.1307005

Lazarus, K. L., & James, T. Y. (2015). Surveying the biodiversity of the Cryptomycota using a targeted

PCR approach. Fungal Ecology ,14 , 62–70. https://doi.org/10.1016/j.funeco.2014.11.004

Letcher, P.M., Powell, M.J. (2019). A taxonomic summary of Aphelidiaceae. IMA Fungus , 10(4), 1-11.

https://doi.org/10.1186/s43008-019-0005-7

Li, S., Deng, Y., Wang, Z., Zhang, Z., Kong, X., Zhou, W., Yi, Y., & Qu, Y. (2020). Exploring the accuracy

of amplicon-based internal transcribed spacer markers for a fungal community. Molecular Ecology Resources

,20 (1), 170–184. https://doi.org/10.1111/1755-0998.13097

Li, Y., Steenwyk, J.L., Chang, Y., Wang, Y., James, T.Y., Sta jich, J.E., Spatafora, J.W., Groenewald, M.,

Dunn, C.W., Hittinger, C.T., Shen, X-X., Rokas, A. (2021). A genome-scale phylogeny of Fungi; insights

into early evolution, radiations, and the relationship between taxonomy and phylogeny, bioRxiv preprint

https://doi.org/10.1101/2020.08.23.262857

Lichtwardt, R. W. 1986. The Trichomycetes: Fungal Associates of Arthropods . Springer-Verlag, New York.

343 pp.

Linder, D. H. (1943). The genera Kickxella, Martensella, and Coemansia .Farlowia ,1(1), 49–77.

https://doi.org/10.2307/2420811

Lindner, D. L., & Banik, M. T. (2011). Intragenomic variation in the ITS rDNA region obscures phylogenetic

relationships and inﬂates estimates of operational taxonomic units in genus Laetiporus .Mycologia ,103 (4),

731–740. https://doi.org/10.3852/10-331

Liu, K. L., Porras-Alfaro, A., Kuske, C. R., Eichorst, S. A., & Xie, G. (2012). Accurate, rapid taxonomic

classiﬁcation of fungal large-subunit rRNA Genes. Applied and Environmental Microbiology ,78 (5), 1523–

1533. https://doi.org/10.1128/AEM.06826-11

Lloyd-MacGilp, S. A., Chambers, S. M., Dodd, J. C., Fitter, A. H., Walker, C., & Young, J. P. W. (1996). Di-

versity of the ribosomal internal transcribed spacers within and among isolates of Glomus mosseae and related

mycorrhizal fungi. New Phytologist ,133 (1), 103–111. https://doi.org/10.1111/j.1469-8137.1996.tb04346.x

Lofgren, L. A., Uehling, J. K., Branco, S., Bruns, T. D., Martin, F., & Kennedy, P. G. (2019). Genome-based

estimates of fungal rDNA copy number variation across phylogenetic scales and ecological lifestyles.Molecular

Ecology ,28 (4), 721–730. https://doi.org/10.1111/mec.14995

Lucking, R., Aime, M. C., Robbertse, B., Miller, A., Ariyawansa, H., Aoki, T., Cardinali, G., Crous, P.,

Druzhinina, I., Geiser, D., Hawksworth, D., Hyde, K., Irinyi, L., Jeewon, R., Johnston, P., Kirk, P., Malosso,

E., May, T., Meyer, W., . . . Schoch, C. (2020). Unambiguous identiﬁcation of fungi: where do we stand and

how accurate and precise is fungal DNA barcoding? IMA Fungus ,11 (1). https://doi.org/10.1186/s43008-

020-00033-z

Maddison, W. P., & Maddison, D. R. (2019). Mesquite Version 3.61 . http://www.mesquiteproject.org

Malar, C.M., Kruger, M., Kruger, C., Wang, Y., Stajich, J.E., Keller, J., Chen, E.C.H., Yildirir, G.,

Villeneuve-Laroche, M., Roux, C., Delaux, P.-M., Corradi, N. (2021). The genome of Geosiphon pyri-

formis reveals ancestral traits linked to the emergence of the arbuscular mycorrhizal symbiosis. Current

Biology , 31, 1-8. https://doi.org/10.1016/j.cub.2021.01.058

Mamanova, L., Coﬀey, A. J., Scott, C. E., Kozarewa, I., Turner, E. H., Kumar, A., Howard, E., Shendure,

J., & Turner, D. J. (2010). Target-enrichment strategies for next-generation sequencing.Nature Methods ,7

(2), 111–118. https://doi.org/10.1038/nmeth.1419

Manter, D. K., & Vivanco, J. M. (2007). Use of the ITS primers, ITS1F and ITS4, to characterize fungal

abundance and diversity in mixed-template samples by qPCR and length heterogeneity analysis.Journal of

Microbiological Methods ,71 (1), 7–14. https://doi.org/10.1016/j.mimet.2007.06.016

McMurdie, P.J., & Holmes, S. (2013) phyloseq: An R Package for Reproducible Interactive Analysis and

Graphics of Microbiome Census Data. PLoS ONE, 8 (4), e61217

Mota-Gutierrez, J., Ferrocino, I., Rantsiou, K., & Cocolin, L. (2019). Metataxonomic comparison between

internal transcribed spacer and 26S ribosomal large subunit (LSU) rDNA gene. International Journal of

Food Microbiology ,290 (October 2018), 132–140. https://doi.org/10.1016/j.ijfoodmicro.2018.10.010

Mueller, R. C., Belnap, J., & Kuske, C. R. (2015). Soil bacterial and fungal community responses to

nitrogen addition across soil depth and microhabitat in an arid shrubland. Frontiers in Microbiology ,6

(SEP). https://doi.org/10.3389/fmicb.2015.00891

Mueller, R. C., Gallegos-Graves, L. V., & Kuske, C. R. (2016). A new fungal large subunit riboso-

mal RNA primer for high-throughput sequencing surveys. FEMS Microbiology Ecology ,92 (2), 1–11.

https://doi.org/10.1093/femsec/ﬁv153

Munch, K., Boomsma, W., Huelsenbeck, J. P., Willerslev, E., & Nielsen, R. (2008). Statistical as-

signment of DNA sequences using Bayesian phylogenetics. Systematic Biology ,57 (5), 750–757.

https://doi.org/10.1080/10635150802422316

Nelson, J., & Shaw, A. J. (2019). Exploring the natural microbiome of the model liverwort: fungal endophyte

diversity in Marchantia polymorpha L. Symbiosis ,78 (1), 45–59. https://doi.org/10.1007/s13199-019-00597-

Nguyen, N. H., Smith, D., Peay, K., & Kennedy, P. (2015). Parsing ecological signal from noise in next

generation amplicon sequencing.New Phytologist ,205 (4), 1389–1393. https://doi.org/10.1111/nph.12923

Nilsson, R. Henrik, Ryberg, M., Kristiansson, E., Abarenkov, K., Larsson, K. H., & Koljalg, U. (2006).

Taxonomic reliability of DNA sequences in public sequences databases: A fungal perspective.PLoS ONE ,1

(1). https://doi.org/10.1371/journal.pone.0000059

Nilsson, R. Henrik, Tedersoo, L., Abarenkov, K., Ryberg, M., Kristiansson, E., Hartmann, M., Schoch,

C. L., Nylander, J. A. A., Bergsten, J., Porter, T. M., Jumpponen, A., Vaishampayan, P., Ovaskainen, O.,

Hallenberg, N., Bengtsson-Palme, J., Eriksson, K. M., Larsson, K.-H., Larsson, E., & Koljalg, U. (2012). Five

simple guidelines for establishing basic authenticity and reliability of newly generated fungal ITS sequences.

MycoKeys ,4, 37–63. https://doi.org/10.3897/mycokeys.4.3606

Nilsson, R. Henrik, Wurzbacher, C., Bahram, M., Coimbra, V. R. M., Larsson, E., Tedersoo, L., Eriksson,

J., Ritter, C. D., Svantesson, S., Sanchez-Garcia, M., Ryberg, M., Kristiansson, E., & Abarenkov, K. (2016).

Top 50 most wanted fungi. MycoKeys ,12 , 29–40. https://doi.org/10.3897/mycokeys.12.7553

Nilsson, Rolf Henrik, Larsson, K. H., Taylor, A. F. S., Bengtsson-Palme, J., Jeppesen, T. S., Schigel, D.,

Kennedy, P., Picard, K., Glockner, F. O., Tedersoo, L., Saar, I., Koljalg, U., & Abarenkov, K. (2019).

The UNITE database for molecular identiﬁcation of fungi: Handling dark taxa and parallel taxonomic

classiﬁcations. Nucleic Acids Research ,47 (D1), D259–D264. https://doi.org/10.1093/nar/gky1022

Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P. R., O’Hara,

R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H., Szoecs, E., & Wagner, H. (2019). Community Ecology

Package ‘ vegan ’ (2.5-6). The R Project for Statistical Computing. https://cran.r-project.org

Opik, M., Davison, J., Moora, M., & Zobel, M. (2014). DNA-based detection and identiﬁcation

of Glomeromycota: The virtual taxonomy of environmental sequences. Botany ,92 (2), 135–147.

https://doi.org/10.1139/cjb-2013-0110

Palmer, J. M., Jusino, M. A., Banik, M. T., & Lindner, D. L. (2018). Non-biological synthetic

spike-in controls and the AMPtk software pipeline improve mycobiome data. PeerJ ,2018 (5).

https://doi.org/10.7717/peerj.4925

Perez-Izquierdo, L., Morin, E., Maurice, J. P., Martin, F., Rincon, A., & Buee, M. (2017). A new

promising phylogenetic marker to study the diversity of fungal communities: The Glycoside Hydrolase 63

gene.Molecular Ecology Resources ,17 (6), e1–e11. https://doi.org/10.1111/1755-0998.12678

Picard, K. T. (2017). Coastal marine habitats harbor novel early-diverging fungal diversity. Fungal Ecology

,25 , 1–13. https://doi.org/10.1016/j.funeco.2016.10.006

Porras-Alfaro, A., Liu, K. L., Kuske, C. R., & Xiec, G. (2014). From genus to phylum: Large-

subunit and internal transcribed spacer rRNA operon regions show similar classiﬁcation accuracies

inﬂuenced by database composition. Applied and Environmental Microbiology ,80 (3), 829–840.

https://doi.org/10.1128/AEM.02894-13

Porter, T. M., & Golding, G. B. (2012). Factors that aﬀect large subunit ribosomal DNA amplicon se-

quencing studies of fungal communities: Classiﬁcation method, primer choice, and error. PLoS ONE ,7(4).

https://doi.org/10.1371/journal.pone.0035749

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., Peplies, J., & Glockner, F. O. (2013).

The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools.Nucleic

Acids Research ,41 (D1), 590–596. https://doi.org/10.1093/nar/gks1219

R Core Team. (2019). R: A Language and Environment for Statistical Computing (3.6.2). https://www.r-

project.org/

Rambaut, A. (2012). FigTree v1.4.3 . Institute of Evolutionary Biology, University of Edinburgh.

http://tree.bio.ed.ac.uk/software/ﬁgtree

Reynolds, N. K., Benny, G. L., Ho, H.-M., Hou, Y.-H., Crous, P. W., & Smith, M. E. (2019). Phy-

logenetic and morphological analyses of the mycoparasitic genus Piptocephalis .Mycologia ,111 (1).

https://doi.org/10.1080/00275514.2018.1538439

Ross, M. G., Russ, C., Costello, M., Hollinger, A., Lennon, N. J., Hegarty, R., Nusbaum, C., &

Jaﬀe, D. B. (2013). Characterizing and measuring bias in sequence data. Genome Biology ,14 (5).

https://doi.org/10.1186/gb-2013-14-5-r51

Schoch, C. L., Seifert, K. A., Huhndorf, S., Robert, V., Spouge, J. L., Levesque, C. A., Chen, W.,

Bolchacova, E., Voigt, K., Crous, P. W., Miller, A. N., Wingﬁeld, M. J., Aime, M. C., An, K. D., Bai,

F. Y., Barreto, R. W., Begerow, D., Bergeron, M. J., Blackwell, M., .. . Schindel, D. (2012). Nu-

clear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi.

Proceedings of the National Academy of Sciences of the United States of America ,109 (16), 6241–6246.

https://doi.org/10.1073/pnas.1117018109

Skelton, J., Jusino, M. A., Carlson, P. S., Smith, K., Banik, M. T., Lindner, D. L., Palmer, J. M., & Hulcr, J.

(2019). Relationships among wood-boring beetles, fungi, and the decomposition of forest biomass.Molecular

Ecology ,28 (22), 4971–4986. https://doi.org/10.1111/mec.15263

Spatafora, J. W., Chang, Y., Benny, G. L., Lazarus, K., Smith, M. E., Berbee, M. L., Bonito, G., Corradi,

N., Grigoriev, I., Gryganskyi, A., James, T. Y., O’Donnell, K., Roberson, R. W., Taylor, T. N., Uehling, J.,

Vilgalys, R., White, M. M., & Stajich, J. E. (2016). A phylum-level phylogenetic classiﬁcation of zygomycete

fungi based on genome-scale data. Mycologia ,108 (5), 1028–1046. https://doi.org/10.3852/16-042

Stamatakis, A. (2014). RAxML version 8: A tool for phylogenetic analysis and post-analysis of large

phylogenies. Bioinformatics ,30 (9), 1312–1313. https://doi.org/10.1093/bioinformatics/btu033

Stielow, J. B., Levesque, C. A., Seifert, K. A., Meyer, W., Irinyi, L., Smits, D., Renfurm, R., Verkley, G. J.

M., Groenewald, M., Chaduli, D., Lomascolo, A., Welti, S., Lesage-Meessen, L., Favel, A., Al-Hatmi, A. M.

S., Damm, U., Yilmaz, N., Houbraken, J., Lombard, L., .. . Robert, V. (2015). One fungus, which genes?

Development and assessment of universal primers for potential secondary fungal DNA barcodes.Persoonia:

Molecular Phylogeny and Evolution of Fungi ,35 (1), 242–263. https://doi.org/10.3767/003158515X689135

Taylor, D. L., Walters, W. A., Lennon, N. J., Bochicchio, J., Krohn, A., Caporaso, J. G., & Pennanen,

T. (2016). Accurate Estimation of Fungal Diversity and Abundance through Improved Lineage-Speciﬁc

Primers Optimized for Illumina Amplicon Sequencing. Applied and Environmental Microbiology ,82 (24),

7217–7226. https://doi.org/http://dx.doi.org/10.1128 /AEM.02576-16

Tedersoo, L., Anslan, S., Bahram, M., Koljalg, U., & Abarenkov, K. (2020). Identifying the ‘unidentiﬁed’

fungi: a global-scale long-read third-generation sequencing approach. Fungal Diversity ,103 (1), 273–293.

https://doi.org/10.1007/s13225-020-00456-4

Tedersoo, L., Anslan, S., Bahram, M., Polme, S., Riit, T., Liiv, I., Koljalg, U., Kisand, V., Nilsson, R.

H., Hildebrand, F., Bork, P., & Abarenkov, K. (2015). Shotgun metagenomes and multiple primer pair-

barcode combinations of amplicons reveal biases in metabarcoding analyses of fungi. MycoKeys ,10 , 1–43.

https://doi.org/10.3897/mycokeys.10.4852

Tedersoo, L., Bahram, M., Puusepp, R., Nilsson, R. H., & James, T. Y. (2017). Novel soil-inhabiting clades

ﬁll gaps in the fungal tree of life. Microbiome ,5(1), 1–10. https://doi.org/10.1186/s40168-017-0259-5

Tedersoo, L., & Lindahl, B. (2016). Fungal identiﬁcation biases in microbiome projects. Environmental

Microbiology Reports ,8(5), 774–779. https://doi.org/10.1111/1758-2229.12438

Tedersoo, L., Tooming-Klunderud, A., Anslan, S. (2017b). PacBio metabarcoding of Fungi and other eu-

karyotes: errors, biases and perspectives. New Phytologist, 217 , 1370-1385. doi: 10.1111/nph.14776

Thiery, O., Moora, M., Vasar, M., Zobel, M., & Opik, M. (2012). Inter- and intrasporal nuclear ribosomal

gene sequence variation within one isolate of arbuscular mycorrhizal fungus, Diversispora sp.Symbiosis ,58

(1–3), 135–147. https://doi.org/10.1007/s13199-012-0212-0

Toju, H., Tanabe, A. S., Yamamoto, S., & Sato, H. (2012). High-coverage ITS primers for the DNA-

based identiﬁcation of ascomycetes and basidiomycetes in environmental samples. PLoS ONE ,7(7).

https://doi.org/10.1371/journal.pone.0040863

Torres-Cruz, T. J., Tobias, T. L. B., Almatruk, M., Hesse, C. N., Kuske, C. R., Desiro, A., Benucci, G. M.

N., Bonito, G., Stajich, J. E., Dunlap, C., Arnold, A. E., & Porras-Alfaro, A. (2017).Biﬁguratus adelaidae

, gen. et sp. nov., a new member of Mucoromycotina in endophytic and soil-dwelling habitats.Mycologia ,

109 (3), 363–378. https://doi.org/10.1080/00275514.2017.1364958

Truong, C., Gabbarini, L. A., Corrales, A., Mujic, A. B., Escobar, J. M., Moretto, A., & Smith, M. E. (2019).

Ectomycorrhizal fungi and soil enzymes exhibit contrasting patterns along elevation gradients in southern

Patagonia. New Phytologist ,222 (4), 1936–1950. https://doi.org/10.1111/nph.15714

Vandepol, N., Liber, J., Desiro, A., Na, H., Kennedy, M., Barry, K., Grigoriev, I. V., Miller, A. N., O’Donnell,

K., Stajich, J. E., & Bonito, G. (2020). Resolving the Mortierellaceae phylogeny through synthesis of multi-

gene phylogenetics and phylogenomics. Fungal Diversity ,In Press .

Vˇetrovsk´y, T., Kolaˇr´ık, M., ˇ

Zifˇc´akov´a, L., Zelenka, T., & Baldrian, P. (2016). The rpb2 gene represents a

viable alternative molecular marker for the analysis of environmental fungal communities. Molecular Ecology

Resources ,16 (2), 388–401. https://doi.org/10.1111/1755-0998.12456

Vilgalys, R., & Hester, M. (1990). Rapid genetic identiﬁcation and mapping of enzymatically ampli-

ﬁed ribosomal DNA from severalCryptococcus species. Journal of Bacteriology ,172 (8), 4238–4246.

https://doi.org/10.1128/jb.172.8.4238-4246.1990

Vu, D., Groenewald, M., Vries, M. De, Gehrmann, T., Stielow, B., Eberhardt, U., Groenewald, J. Z.,

Cardinali, G., Houbraken, J., Boekhout, T., Crous, P. W., & Robert, V. (2019). Large-scale genera-

tion and analysis of ﬁlamentous fungal DNA barcodes boosts coverage for kingdom fungi and reveals

thresholds for fungal species and higher taxon delimitation. Studies in Mycology ,154 , 135–154. htt-

ps://doi.org/10.1016/j.simyco.2018.05.001

Walsh, E., Luo, J., Khiste, S., Scalera, A., Sajjad, S., & Zhang, N. (2020). Pygmaeomycetaceae, a new

root associated family in Mucoromycotina from the pygmy pine plains. BioRxiv , 2020.07.03.187096. htt-

ps://doi.org/10.1101/2020.07.03.187096

Wang, Q., Garrity, G. M., Tiedje, J. M., & Cole, J. R. (2007). Na¨ıve Bayesian classiﬁer for rapid assignment

of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology ,73 (16),

5261–5267. https://doi.org/10.1128/AEM.00062-07

Wang, Y., White, M.M., Moncalvo, J-M. (2019). Diversiﬁcation of the gut fungi Smittium and allies (Harpel-

lales) co-occurred with the origin of complete metamorphosis of their symbiotic insect hosts (lower Diptera).

Molecular Phylogenetics and Evolution, 139, 1-13. https://doi.org/10.1016/j.ympev.2019.106550

White, T., Bruns, T. D., Lee, S., & Taylor, J. (1990). Ampliﬁcation and direct sequencing of fungal ribosomal

RNA genes for phylogenetics. In M. A. Innis, D. H. Gelfand, J. J. Sninsky, & T. J. White (Eds.),PCR

protocols: a guide to methods and applications (pp. 315–322). Academic Press.

Wickham, H. (2016). ggplot2: Elegant Graphics for Data Analysis . Springer-Verlag. htt-

ps://ggplot2.tidyverse.org

Xue, C., Hao, Y., Pu, X., Ryan Penton, C., Wang, Q., Zhao, M., Zhang, B., Ran, W., Huang, Q., Shen, Q.,

& Tiedje, J. M. (2019). Eﬀect of LSU and ITS genetic markers and reference databases on analyses of fungal

communities. Biology and Fertility of Soils ,55 (1), 79–88. https://doi.org/10.1007/s00374-018-1331-4

Yahr, R., Schoch, C. L., & Dentinger, B. T. M. (2016). Scaling up discovery of hidden diversity in fungi:

Impacts of barcoding approaches.Philosophical Transactions of the Royal Society B: Biological Sciences ,

371 (1702). https://doi.org/10.1098/rstb.2015.0336

[dataset] Reynolds, N.K., Jusino, M.A., Stajich, J.E., Smith, M.E.; 2020; OSF (https://osf.io/cz3mh/); doi:

Figure 1. Violin plots of the average percentage of forward and reverse reads that were merged to form contigs

from each primer set (ITS1F/ITS2, LROR/LR3 and LR22F/LR3). The number of total reads returned for

each dataset is listed above the boxes and the number of total (i.e. fungal and non-fungal) OTUs found after

ﬁltering and quality control is given below.

Figure 2. Primer set variation (ITS1F/ITS2, LROR/LR3, LR22F/LR3) in OTU length and read number

assigned to each OTU according to taxonomic assignment by Kingdom and fungal phylum using the hybrid

(for ITS1F) or UTAX (for LSU) method (A, B) and the RDP classiﬁer (C, D). Box plot height and whiskers

represent OTU length range, whereas the box plot width represents the proportion of reads assigned to each

group. The LROR primer set was almost entirely comprised of forward reads with a length of 283 bp, resulting

in a ﬂat line. SUPP FIG 4 contains additional graphs of the LROR data separately. Percentages indicate the

proportion of non-fungal (A, C) and fungal (B, D) OTUs in each dataset. Note that Glomeromycotina is a

subphylum within the Mucoromycota but is categorized separately for comparison with the ITS taxonomy.

Figure 3. Relative abundance of reads assigned to each phylum of early diverging fungi by sample type for

each primer set.

Figure 4. Non-metric multi-dimensional scaling (NMDS) ordination plots for all fungi and early diverging

fungal communities recovered by the primer sets (ITS1F/ITS2, LROR/LR3, LR22F/LR3) for all California

(CA) and Florida (FL) environmental sampling sites. Point colors represent diﬀerent sampling locations and

point shapes indicate primer set. Stress values are listed for each dataset.

Figure 5. Comparison of alpha diversity measures for early diverging fungi for each primer set by site (colors)

and sample type (shapes).

Figure 6. Maximum likelihood phylogenetic reconstruction of fungal LSU sequences including references from

GenBank and newly sequenced isolates from this study. 50 OTUs identiﬁed only as “Fungi” from each of

the LROR and LR22F datasets were included and the numbers are bolded. Analyses were performed in

RAxML v 8 using the GTR + GAMMA model and 1,000 bootstrap replicates. Classes of fungi are colored if

they include unknown LSU OTUs or shaded grey if they do not. The dark grey “BLAST match to protists”

shading indicates that these clades are comprised of OTUs that had GenBank matches to protist sequences

with the OTU IDs in red. Asterisks indicate early diverging clades. SUPP TABLE 5 has a list of all OTUs

included in the phylogeny along with their BLAST matches.

Figure 7. Maximum likelihood phylogenetic reconstruction of Zoopagomycota LSU sequences including re-

ferences from GenBank and newly sequenced isolates from this study. OTUs identiﬁed as Zoopagomycota

species from each of the LROR and LR22F datasets were included and the numbers are bolded and include

the order to which each OTU was classiﬁed. Analyses were performed in RAxML v 8 using the GTR +

GAMMA model and 1,000 bootstrap replicates. Branch supports [?]70 are shown.

Table 1. Zoopagomycota mock community members and results of mock community OTU taxonomy com-

parisons across the ITS1F/ITS2, LROR/LR3, and LR22F/LR3 primer sets, including the target fragment

length in base pairs (bp), GenBank accession numbers (ITS, LSU), and the primer columns list the number

of OTUs of each mock community member detected by the RDP classiﬁer taxonomy (outside parentheses)

versus the RDP+SILVA LSU database (for LSU) or the AMPtk hybrid method (for ITS1F) (in parentheses).

Isolate +ITS1F/ITS2 length bp ++§LR22F/LR3 length bp LROR/LR3 length bp GenBank Accession #¶ITS1F OTUs LROR OTUs LROR OTUs LR22F OTUs LR22F OTUs

Coemansia aciculifera NRRL 2694 412 365 689 1 (1) 0 (1) 0 (1) 0 (1) 0 (1)

Coemansia thaxteri IMI 214463 412 363ˆ700ˆ0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Piptocephalis cruciata CBS 826.97 527 446 773 MG764652, MG764611 0 (0) 1 (1) 1 (1) 0 (1) 0 (1)

Piptocephalis microcephala CBS 418.77 457 437 764 MG764650, MG764609 0 (0) 1 (1) 1 (1) 0 (1) 0 (1)

Syncephalis californica A23985 433+503+845+KY001705, KY001776 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Syncephalis pseudoplumigaleata S71 778ˆ503 817 KY001697, KY001764 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Coemansia erecta IMI 312319 287+361+687+0 (0) 0 (2) 0 (2) 0 (1) 0 (1)

Coemansia interrupta BCRC 34489 216ˆ241 568 JN942674, JN982932 0 (4) 0 (2) 0 (2) 0 (2) 0 (2)

Dimargaris xerosporica NRRL 3178 728+261+592+AY997043, DQ273791 0 (1) 0 (0) 0 (0) 0 (0) 0 (0)

Piptocephalis cylindrospora RSA 2659 427 455 783 MG764680, MG764623 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Piptocephalis moniliformis NRRL 13723 335ˆ468 780 MG764647, MG775651 0 (0) 3 (1) 3 (1) 2 (1) 2 (1)

Syncephalis cornu NRRL A-5447 (61) 407+501 813 KT601335, KY001803 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Isolate +ITS1F/ITS2 length bp ++§LR22F/LR3 length bp LROR/LR3 length bp GenBank Accession #¶ITS1F OTUs LROR OTUs LROR OTUs LR22F OTUs LR22F OTUs

Syncephalis depressa S116 (4) 271ˆ491 833 KY001683, KY001766 0 (1) 0 (1) 0 (1) 0 (1) 0 (1)

Tieghemiomyces parasiticus RSA 861 Unknown 227 544ˆNA, KF848916 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Acaulopage tetraceros T2 281 416 733 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Coemansia sp. RSA 1933 219 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Coemansia sp. RSA 2604 186 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Piptocephalis graefenhanii S99022101 280+522+848+JX312513, JX128019 0 (1) 0 (1) 0 (1) 0 (1) 0 (1)

Syncephalis digitata S521 (K12) 270 458 808 KY001695, KY001806 0 (1) 0 (1) 0 (1) 0 (1) 0 (1)

Syncephalis obconica S227 (K10) 321 458 808 KU317676, KY001788 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Mycoemilia scoparia NBRC 100468 Unknown 354 683ˆNA, AB287999 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Myconymphaea yatsukahoi NBRC 100467 Unknown 370 698ˆNA, AB287998 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Pinnaticoemansia coronantispora CBS 131509 Unknown 370ˆ698ˆNA, AB288000 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Stylopage hadra B35 483+421 745 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Zoophagus pectospora B39 Unknown 354 677 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Acaulopage acanthospora Ac1 Unknown 416 746 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Zoopage sp. C3 326 461 794 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Acaulopage sp. Ap 246ˆ416 789 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Cochlonema odontospora E1 Unknown 389 721 0 (0) 0 (1) 0 (1) 0 (1) 0 (1)

Zoopage sp. Zo2 Unknown 461 811 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Hosts (Cokeromyces, Cunninghamel la, Rhizopus, Umbelopsis)Typically short (˜300) 5 (4) 5 (5) 5 (5) 5 (5) 5 (5)

Unidentiﬁed Kickxellales Typically short (˜300) 0 (7) 1 (0) 1 (0) 0 (0) 0 (0)

Non-zygomycetes (animals, protists, Dikarya fungi) 0 (7) 4 (0) 4 (0) 8 (0) 8 (0)

“Unclassiﬁed” or “Fungi” 21 (0) 21 (0) 11 (0) 11 (0) 9 (1)

+Isolate names in bold indicate species for which genomic DNAs were combined to make the mock community.

The remaining “isolates” refer to OTU identiﬁcations that were recovered after ampliﬁcation, sequencing,

and bioinformatic processing.++ Lengths with ˆindicate that the number is an estimate due to the reference

sequence missing one or more priming sites. Lengths with +indicate that the length is an estimate based

on a diﬀerent isolate of the same species, with the exception of Dimargaris xerosporica for which estimates

are based on the only species with available rDNA sequences, D. bacillispora .¶Bolded GenBank accession

numbers indicate isolates newly sequenced for this study.

Table 2. Results of PERMANOVA and whether ANOVA analyses of theBsim output were signiﬁcant

for all fungi and early diverging fungal (EDF) communities recovered by each primer set (ITS1F/ITS2,

LROR/LR3, LR22F/LR3) by sites within states (California, CA and Florida, FL) and by sample type

(including invertebrates, mud, soil, and water), as well as the interaction between site and sample type, with

asterisks indicating signiﬁcant p-values (<0.05).

PERMANOVA

Degrees of

freedom Sum squares R2F value P value

ANOVA Bsim

signiﬁcant?

ITS1F CA

fungi Site

4 9.845 0.21154 8.8969 0.0001* no

ITS1F CA

fungi

Sample

3 3.755 0.08068 4.5242 0.0001* yes

ITS1F CA

fungi Site x

Sample

11 9.149 0.19658 3.0065 0.0001*

ITS1F FL

fungi Site

1 1.3797 0.0855 4.5509 0.0001* yes

PERMANOVA

Degrees of

freedom Sum squares R2F value P value

ANOVA Bsim

signiﬁcant?

ITS1F FL

fungi

Sample

3 3.4 0.2107 3.7382 0.0001* yes

ITS1F FL

fungi Site x

Sample

3 2.2622 0.14018 2.4872 0.0001*

ITS1F CA

EDF Site

4 9.678 0.19172 7.4519 0.0001* yes

ITS1F CA

EDF Sample

3 3.784 0.07495 3.8844 0.0001* no

ITS1F CA

EDF Site x

Sample

11 9.746 0.19306 2.7287 0.0001*

ITS1F FL

EDF Site

1 1.4805 0.093 4.7734 0.0001* no

ITS1F FL

EDF Sample

3 3.8367 0.241 4.1235 0.0001* yes

ITS1F FL

EDF Site x

Sample

3 1.9184 0.1205 2.0617 0.0007*

LROR CA

fungi Site

4 9.328 0.18756 7.6923 0.0001* yes

LROR CA

fungi

Sample

3 5.096 0.10247 5.6033 0.0001* yes

LROR CA

fungi Site x

Sample

11 9.542 0.19185 2.8612 0.0001*

LROR FL

fungi Site

1 1.2176 0.08762 4.4197 0.0001* no

LROR FL

fungi

Sample

3 3.6358 0.26163 4.399 0.0001* no

LROR FL

fungi Site x

Sample

3 2.1557 0.15512 2.6082 0.0001*

LROR CA

EDF Site

4 8.618 0.1528 5.4378 0.0001* yes

LROR CA

EDF Sample

3 5.205 0.09228 4.3787 0.0001* yes

LROR CA

EDF Site x

Sample

11 8.899 0.15779 2.0418 0.0001*

LROR FL

EDF Site

1 1.2958 0.12837 6.4159 0.0001* no

LROR FL

EDF Sample

3 3.4713 0.34388 5.7291 0.0001* yes

PERMANOVA

Degrees of

freedom Sum squares R2F value P value

ANOVA Bsim

signiﬁcant?

LROR FL

EDF Site x

Sample

3 1.692 0.16761 2.7925 0.0001*

LR22F CA

fungi Site

4 9.531 0.2103 9.1561 0.0001* yes

LR22F CA

fungi

Sample

3 4.397 0.09702 5.6324 0.0001* yes

LR22F CA

fungi Site x

Sample

11 9.274 0.20461 3.2395 0.0001*

LR22F FL

fungi Site

1 1.32 0.0919 5.1877 0.0001* no

LR22F FL

fungi

Sample

3 3.7993 0.26452 4.9772 0.0001* yes

LR22F FL

fungi Site x

Sample

3 2.1189 0.14753 2.7758 0.0001*

LR22F CA

EDF Site

4 9.335 0.16055 5.7059 0.0001* yes

LR22F CA

EDF Sample

3 4.364 0.07506 3.5568 0.0001* yes

LR22F CA

EDF Site x

Sample

11 9.68 0.16648 2.1516 0.0001*

LR22F FL

EDF Site

1 1.3695 0.10074 6.0836 0.0001* no

LR22F FL

EDF Sample

3 4.1174 0.30287 6.0965 0.0001* no

LR22F FL

EDF Site x

Sample

3 2.2544 0.16583 3.3381 0.0001*

Table 3. Mantel test results for the correlation analyses between the distance matrices of each primer set

(ITS1F/ITS2, LROR/LR3, LR22F/LR3) for fungal and early diverging fungal (EDF) OTUs using Spear-

man’s rank correlation, asterisks indicate signiﬁcant p-values (<0.05).

Comparison R statistic Signiﬁcance+90% quantile 95% quantile 97.5% quantile 99% quantile

ITS1F vs LROR EDF 0.5397 0.006* 0.0345 0.0436 0.0539 0.0641

ITS1F vs LR22 EDF 0.5603 0.006* 0.0318 0.0408 0.0487 0.0543

LROR vs LR22F EDF 0.5849 0.006* 0.0356 0.0483 0.0598 0.0698

ITS1F vs LROR fungi 0.4716 0.006* 0.0389 0.0532 0.0637 0.0740

ITS1F vs LR22F fungi 0.7101 0.006* 0.0373 0.0481 0.0545 0.0667

LROR vs LR22F fungi 0.4483 0.006* 0.0424 0.0518 0.0601 0.0771

+P-values adjusted for multiple comparisons using the Holm method (Holm, 1979).

10,028 OTUs

ITS1F LR22F LROR

7,609 OTUs

Average % merged reads

5,786 OTUs

Total reads

9,135,843

Total reads

7,403,607

Total reads

14,233,141

OTU Length bp

ITS1F

LR22F

LROR

ITS1F

LR22F

LROR

OTU Length bp

Aphelidiomycota

Ascomycota

Basidiomycota

Blastocladiomycota

Chytridiomycota

Entorrhizomycota

Glomeromycotina

Microsporidia

Mucoromycota

Rozellomycota

Zoopagomycota

Not assigned

Ascomycota

Basidiomycota

Blastocladiomycota

Chytridiomycota

Glomeromycotina

Mucoromycota

Zoopagomycota

200

300

400

500

200

300

400

500

UTAX/hybrid

Fungal Phylum

200

300

400

500

200

300

400

500

RDP

Fungal Phylum

17.3% 33.2% 38.8%

0% 54% 47.2%

80.7% 36.2%31.2%

Amoebozoa

Apusozoa

Choanozoa

Eukaryota

Excavata

Fungi

Hacrobia

Haptista

Metazoa

Rhodophyta

SAR

Viridiplantae

Filasterea

UTAX/hybrid

Kingdom

RDP

Kingdom

Amoebozoa

Eukaryota

Fungi

Metazoa

Rhodophyta

SAR

Viridiplantae

Filasterea

100% 27.7% 28.5%

Phylum

Relative Abundance

ITS1F

500000

1000000

1500000

2000000

Invertebrate Mud Soil

1000000

2000000

3000000

Invertebrate Mud Soil Water

LROR

LR22F

Water

Aphelidiomycota

Blastocladiomycota

Chytridiomycota

Glomeromycotina

Microsporidia

Mucoromycota

Rozellomycota

Zoopagomycota

1000000

2000000

2500000

NMDS1

NMDS2

Primer

ITS1F

LR22F

LROR

Location

Evey Canyon

San Jacinto

Sweeney Wash

Sweeney 1

Sweeney 2

California early diverging fungi

−0.4

−0.2

0.0

0.2

0.4

−0.4 −0.2 0.0 0.2 0.4 0.6

Florida early diverging fungi Primer

ITS1F

LR22F

LROR

Location

Blue Pond

Goose Lake

NMDS2

NMDS1

-0.25

0.00

0.25

0.50

-0.25 0.00 0.25

-0.25

0.00

0.25

0.50

-0.50 -0.25 0.00 0.25

California all fungi

NMDS1

NMDS2

-0.50

-0.25

0.00

0.25

-0.50 -0.25 0.00 0.25

Florida all fungi

NMDS2

NMDS1

EDF stress: 0.1367924

Fungi stress: 0.1168332

EDF stress: 0.1556327

Fungi stress: 0.1429781

EDF stress: 0.1502559

Fungi stress: 0.1281375

EDF stress: 0.07588488

Fungi stress: 0.06929218

EDF stress: 0.07262848

Fungi stress: 0.09405482

EDF stress: 0.05850112

Fungi stress: 0.08640903

A B

Observed

Chao1

Shannon

Simpson

0.00

0.25

0.50

0.75

1.00

100

150

100

150

Alpha Diversity Measure

ITS1F

Observed

Chao1

Shannon

Simpson

0.00

0.25

0.50

0.75

1.00

Samples

Location

Blue Pond, FL

Evey Canyon, CA

Goose Lake, FL

San Jacinto, CA

Sweeney wash, CA

Sweeney 1, CA

Sweeney 2, CA

Substrate

Invertebrate

Mud

Soil

Water

LR22F

Observed

Chao1

Shannon

Simpson

0.00

0.25

0.50

0.75

1.00

LROR

Cystoﬁlobasidium inﬁrmominiatum CBS323

Anthracocystis andrewmitchellii JQ9953701

Symmetrospora pseudomarina KJ7012141

Paratritirachium cylindroconium CBS838.71

Olpidium brassicae S000622265

Mortierella macrocystopsis MH8737831

Basidiobolus haptosporus var. minor ATCC16579

Piptocephalis lepidula NRRLA5450

Piptocephalis tieghemiana RSA1564

Piptocephalis freseniana D98122701

Piptocephalis pseudocephala IMI210884

Piptocephalis formosana BCRC34206

Piptocephalis indica BCRC34516

Piptocephalis tieghemiana Horse Idaho

Piptocephalis unispora RSA1396

Piptocephalis trachypus RSA737

Piptocephalis brijmohanii NRRLA21622

Piptocephalis xenophila S98092309

Uncultured soil fungus clone FunCON4 02H

Piptocephalis curvata CBS100869

Piptocephalis graefenhanii CBS100814

Piptocephalis graefenhanii BCRC34436

Zoophagus pectospora B38

Acaulopage acanthospora Ac1

Zoophagus insidians Zi2

Cochlonema odontospora E1

Syncephalis cf.plumigaleata NRRLA13833

Syncephalis depressa STDTFEBa

Dimargaris bacillispora AFTOLID 136

Coemansia linderi BCRC34191

NBRC100470 Pinnaticoemansia coronantispora

NBRC100467 Myconymphaea yatsukahoi

AF031066.1 Martensiomyces pterosporus

BCRC 31802 Linderina macrospora

OR-14-W25A Orphella pseudohiemalis

NRRL3067 Spiromyces minutus

NF-15-5A Harpella melusinae

S003865602 Genistelloides hibernus

NG0600761 Ramicandelaber taiwanensis

EF3924001 Zoophthora phalloides

GQ2858741 Entomophthora planchoniana

NG0672751 Spizellomyces acuminatus

DQ27378513155Dendrochytrium crassum

NG0602511 Kappamyces laurelensis

FJ8041531 Mesochytrium penetrans

KJ6681211 Monoblepharis hypogyna

DQ2738221 Neocallimastix sp.

CBS112.53 Ascochyta medicaginicola var. macrospora

FJ1768561 Mycosphaerella pneumatophorae

JN7125511 Fusicladium proteae

NG0662301 Chlamydotubeuﬁa cylindrica

EF1756491 Aliquandostipite khaoyaiensis

FJ1768871 Symbiotaphrina buchneri

NG0661601 Xylona heveae

KT2322171 Elaphomyces granulatus

DQ9123361 Hypotrachyna caraccensis

KF1579841 Chaenotheca phaeocephala

MG5733361 Apiosphaeria guaranitica

AY3462591 Apiospora setosa

NG0673701 Botryotinia ﬁcariarum

HAILUO215 Tetracladium globosum

S003854464 Vermispora spermatophaga

S004606307 Microdochiella fusarioidea

HQ1107021 Brachyphoris brevistipitata

S003827891 Dactylellina robusta

S003827888 Monacrosporium shuzhengense

FJ1768641 Arthrobotrys elegans

NG0597301 Tuber verrucosivolvum

JF8360221 Archaeorhizomyces ﬁnlayi

FG15P2b Archaeorhizomyces borealis

NG0661861 Schizosaccharomyces octosporus

AM9464721 Tricholoma frondosae

Psathyrella candolleana AFTOL ID1507

Ganoderma enigmaticum NG0581561

Marchandiomyces marsonii NG0594341

Tomentella pulvinulata BAFC52370

Sistotrema albopallescens KHL11070

Geastrum smithii KF9885751

Tremella globispora CBS6972

Cerinomyces crustulinus S000451434

Rhodotorula chungnamensis AY9539491

Phragmidium tuberculatum KJ8419191

Meredithblackwellia eburnea NG0583331

Macalpinomyces muelleri KX6869491

Geminibasidium hirsutum S003870098

Cunninghamella bertholletiae JA6068584484035

Fennellomyces heterothallicus JN2065391

Protomycocladus faisalabadensis JN2065581

Umbelopsis autotrophica CBS310.93

Jimgerdemannia ﬂammicorona KPMNC0024738

0.3

Choanephora infundibulifera JN2065141

OTU3201 LR22F

KJ4623661 Xylographa opegraphella

Piptocephalis debaryana CBS794.97

Piptocephalis moniliformis NRRL13723

Mortierella jenkinii S003872849

Piptocephalis microcephala CBS418.77

ARSEF6175 Ramicandelaber longisporus

EF6437491 Clavascidium umbrinum

DQ27380413972 Oedogoniomyces sp.

Sphaerocreas pubescens LC4311101

Piptocephalis cruciata CBS826.97

Archaeospora trappei S003841656

Mucor racemosus JA6068602263946

OTU3001 LR22F

CCF4001 Aspergillus citocrescens

Tieghemiomyces parasiticus NRRL2924

LROR OTU2235

Umbelopsis gibberispora CBS109328

TB8870 Entoloma vinaceum

Diversispora celata AM7134192

OTU4779 LROR

Piptocephalis cylindrospora RSA786

RSA2049 Coemansia sp.

LR22F OTU2179

Dentiscutata heterogama INVAM FL225

Trechisporasp. MPM 2013

Paraglomus occultum S003824525

Crucibulum laeve TUB011564

AY2611231 Orbilia auricolor

KP6987271 Curvularia eragrostidis

Tomentella botryoides

Acaulopage sp. Ap

Boletellus deceptivus KP3276191

OTU3880 LR22F

Basidiobolus ranarum EF3924191

Acaulopage acanthospora Ac3

Dispira cornuta NRRLA-16106

OTU2952 LR22F

LROR OTU2016

LROR OTU2626

JN9391821Erynia radicans

OTU373 LR22F

LROR OTU3050

LROR OTU3054

AM9464141 Calocybe gambosa

Pilobolus crystallinus JX6445171

Piptocephalis sp. HTS P14

JX9466931 Conidiobolus undulatus

Piptocephalis xenophila NRRLA10082

LROR OTU3048

Mortierella parvispora MH8685881

Piptocephalis cylindrospora RSA2659

DQ2737841 Entophlyctishelioformis

Piptocephalis indica RSA1275

Marasmiellus synodicus AY6394351

OTU2653 LR22F

Acaulopage tetraceros T2

KT5915611 Hebeloma sacchariolens

LR22F OTU2454

JN9409991 Allochytridium luteum

Thamnostylum lucknowense JN2065461

Apophysomyces variabilis JN9806991

KY3505381 Laboulbenia ﬂagellata

NRRL1564 Coemansia reversa

KX4642811 Diplodia seriata

Piptocephalis cylindrospora RSA525

Piptocephalis sp. Sky0202

AM9464151 Camarophyllopsis schulzeri

OTU1935 LROR

Sporodiniella umbellata FLAS-F-62758

Piptocephalis ﬁmbriata CBS950.95

OTU3126 LROR

DQ7676541 Tubeuﬁa helicomyces

LROR OTU3006

NRRL3115 Coemansia spiralis

KF8489091 Smittium culicis

OTU3095 LR22F

Geosiphon pyriformis AM1839202

Mortierella kuhlmanii CBS157.71

KY2496411 Aphelidium desmodesmi

LR22F OTU8379

RSA720 Coemansia sp.

OTU2799 LR22F

Mycotypha indica NG0641351

Umbelopsis ramanniana S000622244

Coemansia aciculifera AB287993.1

OTU2070 LR22F

CBS122642 Cladophialophora subtilis

LROR OTU2661

AFTOLID 271 Cochliobolus sativus

OTU3056 LR22F

JX0000851 Chaenotheca trichialis

LROR OTU1183

OTU2728 LROR

Piptocephalis indica CBS927.95

JQ0047921 Conidiobolus coronatus

Coemansia sp. S171

Piptocephalis curvata Symg0108

NG0599191 Tuber spinoreticulatum

EF3923951 Entomophaga maimaiga

OTU1667 LR22F

LC0943861 Saccharomyces cerevisiae

Gigaspora margarita S004127012

Umbelopsis ovata CBS499.82

OTU1150 LROR

Basidiobolus microsporus CBS130.62

EF3923771 Massospora cicadina

OTU1711 LR22F

Syncephalis pyriformis BCRC34472

JF9220301 Eurotium amstelodami

OTU2208 LROR

LROR OTU876

Syncephalis sphaerica SCANSAa

Syncephalis cornu NRRL6268

Rhizophagus intraradices FJ23557113086

Basidiobolus meristosporus EF3924221

NG0602571 Lecanactis submollis

Stylopage hadra B35

Diversispora sp EE1

LROR OTU2835

NG0426821 Trinosporium guianense

OTU2253 LR22F

NG0578101 Gymnascella littoralis

FJ5156331 Didymella vitalbina

EF3924011 Batkoa major

KX2449761 Taeniolella pyrenulae

LROR OTU3765

IB2005309 Hebeloma mesophaeum

EF3923891 Furia americana

Cyathus striatus AF3362471

Glomus sp MUCL 43206

KF8162291 Abrothallus parmeliarum

S004058249 Fimicolochytrium jonesii

JN0120181 Wolﬁna aurantiopsis

Absidia panacisoliNG0639481

OTU3604 LR22F

Piptocephalis sp. S114

OTU4046 LR22F

OTU5666 LROR

LR22F OTU1998

LROR OTU5816

Entorrhiza casparyana AF0098521

Piptocephalis sp. FD1

AY5713811 Arthonia dispersa

OTU1509 LROR

OTU4294 LR22F

KY7422721Epicoccum viticis

Maravaliacryptostegiae KT1994011

Syncephalis depressa S114

OR-11-W8 Pteromaktron sp.

1-1-5 Cenococcum geophilum

LROR OTU875

OTU2307 LR22F

Sistotrema conﬂuens AY5867121

AB8075101 Aquastroma magniostiolata

SMH2961 Cercophora atropurpurea

Coemansia furcata BCRC34190

MN614386 Amanita sp. texasorora

LROR OTU2199

MFLUCC17 0783 Alternaria hampshirensis

OTU2802 LR22F

Rhopalomyces elegans AFTOLID 142

Piptocephalis brijmohanii NRRL66062

Xerocomus badius Xb2

Piptocephalis indica NRRLA12035

DQ4709731 Taphrina deformans

JX2425911 Batkoa gigantea

Trechispora nivea AY5867201

OTU2374 LR22F

OTU3554 LR22F

Mortierella turﬁcola CBS432.76

NG0426371 Scheffersomyces stipitis

Glomus intraradices MUCL 43194

NG0275661 Boothiomyces macroporosum

JX6698671 Urnula hiemalis

Neolentinus cyathiformis KM4114771

JX50729824085774 Amoeboaphelidium protococcarum

Zoopage sp. C3

Absidia californica EU7363011

OTU1714 LR22F

Diversispora celata pHS006 01

LR22F OTU336

RSA1358 Coemansia sp.

LROR OTU5731

Rhizophagus irregularisDAOM22 9456

Schizangiella serpentis EF3924281

Jimgerdemannia lactiﬂua LC4311011

DQ4709851 Neolecta vitellina

NG0276191 Batrachochytrium dendrobatidis

OTU4179 LROR

OTU1177 LR22F

Naganishia vishniacii KY1086251

OTU8142 LR22F

Lactiﬂuus medusae KR3641981

Circinella angarensis JN2065511

Endogone incrassata LC1073621

OTU3248 LR22F

Mortierella verticillata DQ2737941

BP B Hebeloma sacchariolens

LROR OTU1539

Syzygites megalocarpus JQ0432301

OTU1843 LR22F

NRRL3781 Linderina pennispora

GJS9526 Clonostachys pityrodes

OTU1878 LR22F

Mycocladus corymbiferus FJ3453501

OTU3214 LR22F

KF7974501 Bifusella camelliae

OTU2434 LR22F

KY1088121 Pichia fermentans

OTU3084 LR22F

AFTOLID 283 Pyrenophora phaeocomes

Benjaminiella youngii MH8738481

KC1210581 Conidiobolus paulus

IMI279145 Coemansia erecta

NRRL2925 Dipsacomyces acuminosporus

OTU4841 LROR

OTU1188 LROR

AY2073041 Sphagnurus paluster

EL37 99Tricholoma_apium

LROR OTU853

OTU3368 LR22F

LROR OTU3431

MUT4379 Pleospora typhicola

CBS885.95 Knuﬁa perforans

OTU3817 LR22F

LROR OTU3137

Piptocephalis debaryana CBS260.77

OTU1822 LR22F

Tulasnella eremophila KJ7011891

NBRC100468 Mycoemilia scoparia

Syncephalis aurantia NRRL6269

Dentiscutata heterogama DQ27379213188

OTU5379 LR22F

OTU1317 LROR

Piptocephalis trachypus RSA1397

OTU773 LR22F

OTU4118 LROR

MF13025 Trichoderma harzianum

AJ8644751 Orpinomyces sp.

OTU3586 LR22F

CBS127115 Trichoderma peltatum

Syncephalis parvula S439-M11

CBS26759 Pyrenochaeta lycopersici

OTU413 LROR

Piptocephalis sp. NRRLA10083

Acaulospora colliculosa GU3263391

CA-18-W17 Bojamyces sp.

Piptocephalis lepidula RSA2468

Scutellospora heterogama AFTOL ID138

Zoophagus pectospora B39

KT1552281 Nannizzia corniculata

LT6272411 Acidiella americana

OTU1859 LROR

LROR OTU2720

Rhizopus lyococcus KJ4085621

Piptocephalis lepidula RSA2551

OTU1790 LROR

BRACR16695 Clavaria ﬂavostellifera

LROR OTU4134

Paraglomus occultum AFTOL ID844

Stylopage hadra B36

Craterocolla cerasi AY5055421

Glomus etunicatum S003857707

Russula solaris 559

Piptocephalis sp. CBS945.95

Piptocephalis ﬁmbriata BCRC34715

FJ34535213830 Fusarium solani

Spiromyces aspiralis

Rhizophagus fasciculatus BEG53

LROR OTU733

Sebacina incrustans FJ6445132

AF031068.1 Spirodactylon aureum

Rhizopus delemar MH8666671

Zoopage sp. Zo2

OTU3519 LR22F

KJ6680981 Monoblepharis polymorpha

LROR OTU1205

Entorrhiza tenuis KP4130811

FJ8900381 Amanita muscaria

Marasmius anomalus EF1600861

EU8285091 Nowakowskiella sp.

OTU4126 LR22F

OTU3437 LR22F

OTU2685 LR22F

OTU1976 LROR

Syncephalis californica NRRLA-23985

Abortiporus biennis FD319

Pilaira moreaui JX6445141

RSA1939 Coemansia sp.

LROR OTU1847

Cokeromyces recurvatus NG0588131

JQ8626081 Xylaria grammica

Coemansia furcata NRRL5531

EF6437651 Placidiumarboreum

OTU2378 LR22F

Piptocephalis sp. BennyS2-5

OTU1260 LROR

Endogone pisiformis KPMNC0024229

DQ27382613589 Triparticalcar arcticum

OTU761 LR22F

LROR OTU507

MH9339701 Rhopalophlyctis sarcoptoides

Ellisomyces anomalus NG0673651

OTU889 LR22F

Backusella variabilis KC0126581

Mortierella alpina KC0183281

AY3008681 Toninia sedifolia

LROR OTU1531

MH6264961 Zygorhizidium afﬂuens

Piptocephalis sp. NRRL26521

CBS115979 Neocucurbitaria cava

OTU1708 LR22F

Syncephalis intermedia NRRL6286

JX9672741 Amoeboaphelidium sp.

LROR OTU1679

NG0602501 Lobulomyces angularis

OTU3750 LROR

Piptocephalis platyclados Brazil2

OTU2011 LROR

OTU3241 LR22F

Coemansia sp. RSA1694

CBS126.86 Cladophialophora boppii

OTU3390 LR22F

AF3362581 Hymenogaster vulgaris

Thelephora sp. IR 2013

WS36 Wilcoxina mikolae

Gigaspora margarita BEG152

OTU3547 LROR

Umbelopsis longicollis S003857285

KY3505001 Coreomyces sp.

Saksenaea trapezispora LT6074071

NG0424061 Protomyces inouyei

EF3923841 Pandora delphacis

Rhodotorula mucilaginosa KY1091371

FJ8900411 Amanita ﬂavoconia

Gongronella butleri JN2066071

Coemansia sp. RSA2424

BCRC34628 Coemansia asiatica

OTU1412 LROR

JX9683691 Bolbitius titubans

Abundisporus mollissimus NG0570151

Syncephalis intermedia DTSTM

RSA522 Coemansia sp.

NRRL2693 Kickxella alabastrina

Glomeromycetes

Endogonomycetes 1

Endogonomycetes 2

Umbelopsidomycetes

Mucoromycetes

Geminibasidiomycetes

Pucciniomycotina

Tremellomycetes

BLAST match to protists

Agaricomycetes

Archaeorhizomycetes

Orbiliomycetes

Sordariomycetes

Eurotiomycetes

Aphelidiomycetes

BLAST match to protists

Chytridiomycetes

Entomophthoromycetes

Harpellomycetes

Kickxellomycetes

Zoopagomycetes

Basidiobolomycetes

Mortierellomycetes

BLAST match to protists

ResearchGate has not been able to resolve any citations for this publication.

Long- and short-read metabarcoding technologies reveal similar spatio-temporal structures in fungal communities

Preprint

Full-text available

Oct 2020

A genome-scale phylogeny of Fungi; insights into early evolution, radiations, and the relationship between taxonomy and phylogeny

Preprint

Full-text available

Aug 2020

Phylogenomic studies based on genome-scale amounts of data have greatly improved understanding of the tree of life. Despite their diversity, ecological significance, and biomedical and industrial importance, large-scale phylogenomic studies of Fungi are lacking. Furthermore, several evolutionary relationships among major fungal lineages remain controversial, especially those at the base of the fungal phylogeny. To begin filling these gaps and assess progress toward a genome-scale phylogeny of the entire fungal kingdom, we compiled a phylogenomic data matrix of 290 genes from the genomes of 1,644 fungal species that includes representatives from most major fungal lineages; we also compiled 11 additional data matrices by subsampling genes or taxa based on filtering criteria previously shown to improve phylogenomic inference. Analyses of these 12 data matrices using concatenation- and coalescent-based approaches yielded a robust phylogeny of the kingdom in which ~85% of internal branches were congruent across data matrices and approaches used. We found support for several relationships that have been historically contentious (e.g., for the placement of Wallemiomycotina (Basidiomycota), as sister to Agaricomycotina), as well as evidence for polytomies likely stemming from episodes of ancient diversification (e.g., at the base of Basidiomycota). By examining the relative evolutionary divergence of taxonomic groups of equivalent rank, we found that fungal taxonomy is broadly aligned with genome sequence divergence, but also identified lineages, such as the subphylum Saccharomycotina, where current taxonomic circumscription does not fully account for their high levels of evolutionary divergence. Our results provide a robust phylogenomic framework to explore the tempo and mode of fungal evolution and directions for future fungal phylogenetic and taxonomic studies.

Optimized metabarcoding with Pacific biosciences enables semi‐quantitative analysis of fungal communities

Article

Full-text available

Jul 2020
NEW PHYTOL

Recent studies have questioned the use of high‐throughput sequencing of the nuclear ribosomal internal transcribed spacer (ITS) region to derive a semi‐quantitative representation of fungal community composition. However, comprehensive studies that quantify biases occurring during PCR and sequencing of ITS amplicons are still lacking. We used artificially assembled communities consisting of 10 ITS‐like fragments of varying lengths and guanine‐cytosine (GC) contents to evaluate and quantify biases during PCR and sequencing with Illumina MiSeq, PacBio RS II and PacBio Sequel I technologies. Fragment length variation was the main source of bias in observed community composition relative to the template, with longer fragments generally being under‐represented for all sequencing platforms. This bias was three times higher for Illumina MiSeq than for PacBio RS II and Sequel I. All 10 fragments in the artificial community were recovered when sequenced with PacBio technologies, whereas the three longest fragments (> 447 bases) were lost when sequenced with Illumina MiSeq. Fragment length bias also increased linearly with increasing number of PCR cycles but could be mitigated by optimization of the PCR setup. No significant biases related to GC content were observed. Despite lower sequencing output, PacBio sequencing was better able to reflect the community composition of the template than Illumina MiSeq sequencing.

Relationships among wood‐boring beetles, fungi, and the decomposition of forest biomass

Article

Full-text available

Oct 2019

A prevailing paradigm in forest ecology is that wood‐boring beetles facilitate wood decay and carbon cycling, but empirical tests have yielded mixed results. We experimentally determined the effects of wood borers on fungal community assembly and wood decay within pine trunks in the southeastern United States. Pine trunks were made either beetle‐accessible or inaccessible. Fungal communities were compared using culturing and high‐throughput meta‐barcode sequencing of DNA and RNA. Prior to beetle infestation, living pines had diverse fungal endophyte communities. Endophytes were displaced by beetle‐associated fungi in beetle‐accessible trees, whereas some endophytes persisted as saprotrophs in beetle‐excluded trees. Beetles increased fungal diversity several fold. Over forty taxa of Ascomycota were significantly associated with beetles, but beetles were not consistently associated with any known wood‐decaying fungi. Instead, increasing ambrosia beetle infestations caused reduced decay, consistent with previous in vitro experiments that showed beetle‐associated fungi reduce decay rates by competing with decay fungi. No effect of bark‐inhabiting beetles on decay was detected. Platypodines carried significantly more fungal taxa than scolytines. Molecular results were validated by synthetic and biological mock communities and were consistent across methodologies. RNA sequencing confirmed that beetle‐associated fungi were biologically active in the wood. Meta‐barcode sequencing of the LSU/28S marker recovered important fungal symbionts that were missed by ITS2, though community‐level effects were similar between markers. In contrast to the current paradigm, our results indicate ambrosia beetles introduce diverse fungal communities that do not extensively decay wood, but instead reduce decay rates by competing with wood decay fungi.

A taxonomic summary of Aphelidiaceae

Article

Full-text available

Jun 2019

Aphelids are parasitoids of various algae and diatoms, and in a recent classification are contained in family Aphelidiaceae, phylum Aphelidiomycota, kingdom Fungi. Family Aphelidiaceae (the only family in the phylum) is composed of four genera: Aphelidium, Paraphelidium, Amoeboaphelidium, and Pseudaphelidium. All species are known morphologically, and most have been illustrated. Few have been examined ultrastructurally, and even fewer have been sequenced for molecular comparisons. Recent studies in molecular phylogenetics have revealed an abundance of related environmental sequences that indicate unrealized biodiversity within the group. Herein, we briefly summarize the history of aphelids and acknowledge the controversy of placement of the group with related organisms. With light microscopic images and transmission electron micrographs, we illustrate typical life cycle stages for aphelids, provide updated descriptions and taxonomy for all described species, and provide a key to the species.

The unbearable lightness of sequence-based identification

Article

Full-text available

May 2019

Using the basic GenBank local alignment search tool program (BLAST) to identify fungi collected in a recently protected beech forest at Montricher (Switzerland), the number of ITS sequences associated to the wrong taxon name appears to be around 30%, even higher than previously estimated. Such results rely on the in-depth re-examination of BLAST results for the most interesting species that were collected, viz. first records for Switzerland, rare or patrimonial species and problematic species (when BLAST top scores were equally high for different species), all belonging to Agaricomycotina. This paper dissects for the first time a number of sequence-based identifications, thereby showing in every detail-particularly to the user community of taxonomic information-why sequence-based identification in the context of a fungal inventory can easily go wrong. Our first conclusion is that in-depth examination of BLAST results is too time consuming to be considered as a routine approach for future inventories: we spent two months on verification of approx. 20 identifications. Apart from the fact that poor taxon coverage in public depositories remains the principal impediment for successful species identification, it can be deplored that even very recent fungal sequence deposits in GenBank involve an uncomfortably high number of misidentifications or errors with associated metadata. While checking the original publications associated with top score sequences for the few examples that were here reexamined , a positive consequence is that we uncovered over 80 type sequences that were not annotated as types in GenBank. Advantages and pitfalls of sequence-based identification are discussed, particularly in the light of undertaking fungal inventories. Recommendations are made to avoid or reduce some of the major problems with sequence-based identification. Nevertheless, the prospects for a more reliable sequence-based identification of fungi remain quite dim, unless authors are ready to check and update the metadata associated with previously deposited sequences in their publications.

New Zoopagaceae Capturing and Consuming Soil Amoebae

Article

Sep 2018

Charles Drechsler

Exploring the accuracy of amplicon‐based internal transcribed spacer (ITS) markers for fungal community

Article

Oct 2019

With the continual improvement of high‐throughput sequencing technology and constant updates to fungal reference databases, the use of amplicon‐based DNA markers as a tool to reveal fungal diversity and composition in various ecosystems has become feasible. However, both primer selection and experimental procedure still require meticulous verification. Here, we computationally and experimentally evaluated the accuracy and specificity of three widely used or newly designed internal transcribed spacer (ITS) primer sets (ITS1F/ITS2, gITS7/ITS4 and 5.8S‐Fun/ITS4‐Fun). In silico evaluation revealed that the primer coverage varied at different taxonomy levels due to the differences in degeneracy and location of primer sets. Using even and staggered mock community standards, we identified different proportions of chimeric and mismatch reads generated by different primer sets, as well as great variation in species abundances, suggesting that primer selection would affect the results of amplicon‐based metabarcoding studies. Choosing a proofreading and high‐fidelity polymerase (KAPA HiFi) could significantly reduce the percentage of chimeric and mismatch sequences, further reducing OTU inflation. Moreover, for two types of environmental fungal communities, plant endophytic and soil fungi, it was demonstrated that the three primer sets could not reach a consensus on fungal community compositions or diversities, and primer selection, not the experimental treatment, determines observed soil fungal community diversity and composition. Future DNA marker surveys should pay greater attention to potential primer effects and improve the experimental scheme to increase credibility and accuracy.

Molecular variation within and among species of Harpellales

Article

Jan 2001

Intra- and interspecific variation of the nuclear ribosomal internal transcribed spacers (ITS) of 77 Smittium isolates and related genera was studied using RFLP. Due to the sequence and length variation encountered, the ITS is unsuitable for comparison at the species level within Smittium, as presently delimited. Intra-and interspecific variation of 18S rDNA was also analyzed. Cladistic analysis of 32 new 18S rDNA sequences resolved at least five lineages and suggests that Smittium is not monophyletic. In addition, 17 18S rDNA sequences from 5 orders of Zygomycota were included to assess phylogenetic relationships with the Trichomycetes. The kickxellid origin of Trichomycetes was not strongly supported.

A new 18S rRNA phylogeny of uncultured predacious fungi (Zoopagales)

Article

Mar 2019

Previous molecular phylogenetic studies have shown that families in Zoopagales are not monophyletic. To test the monophyly of genera and species in the order, we used a single-cell approach to generate nuclear 18S rRNA (18S) sequences for 10 isolates representing nine taxa. We provide the first sequences for the genus Zoopage and additional sequences for taxa in Cocholonema, Acaulopage, and Zoophagus. Our results reveal that Zoophagus, Zoopage, and Acaulopage tetraceros are not monophyletic. We conclude that morphology alone is not sufficient to delineate genera and species in the order and encourage studies that increase genetic sampling of taxa including type species.

Understudied, underrepresented, and unknown: methodological biases that limit detection of early diverging fungi from environmental samples

Abstract

Recommended publications

Employing 454 amplicon pyrosequencing to reveal intragenomic divergence in the internal transcribed...

The Unusual Dominance of the Yeast Genus Glaciozyma in the Deeper Layer in an Antarctic Permafrost C...

Pine wilt disease alters soil properties and root-associated fungal communities in Pinus tabulaeform...

Understudied, underrepresented, and unknown: Methodological biases that limit detection of early div...