Supplementary material

Peeking into the black box – integrated taxonomy of Archaeorhizomycetes

Faheema Kalsoom Khan1,2 **, Kerri Kluting 1 **, Jeanette Tångrot 3, Hector Urbina 1,3, Tea

1 1 1 2 1* Ammunet , Shadi Eshghi Sahraei , Martin Rydén , Martin Ryberg , Anna Rosling

Supplementary tables

Table S1. Sample names, soil horizon origin, primer number and barcode Table S2. Number of long reads throughout the quality control and filtering Table S3. Overview of 41 Archaeorhizomycetes ASVs in “phylogenetic” dataset Table S4. Overview of Archaeorhizomycetes reads in the “ecological” dataset Table S5. Summary statistics for 2-way ANOVA for “ecological” dataset Table S6. Summary statistics for MANOVA on PSHs Table S7. Niche specific distribution of PHSs, Kruskal-Wallis test Table S8. Frequency and abundance of Archaeorhizomycetes PSHs Table S9. Niche specific distribution across two orthogonal contrasts, MANOVA test

Supplementary figures

Figure S1. Overview of sampled plots at Ivantjärnsheden field station Figure S2. All fungal ASV tree identifying ASVs in Archaeorhizomycetes Figure S3. NMDS ordination of total fungal ASV community in ugit102 Figure S4. Ranked relative abundance and cumulative abundance of PSHs in Fig. 1 Figure S5. Proportion (%) of Archaeorhizomycetes reads across soil horizons Figure S6. Number of itASVs and PSHs across soil horizon Figure S7. Tree corresponding to Fig 1 including node labeles Figure S8. nMDS of relative abundance of 68 taxa across soil horizons Figure S9. Relative abundance of PSH_7 vs. PSH_8

1 Supplementary tables

Table S1. Metabarcoding of the fungal ITS2 region codes for Sample names, soil horizon (Horizon), primer number and barcode used for Ion Torrent sequencing of “ecological“ sequence dataset. Right column indicates samples included in PacBio libraries (SWO, SWE and SWB) for the “phylogenetic” sequence dataset. Plot nr. /cont. Horizon Primer nr. Barcode sequence In PB library 4 O 39 TAACAATCGGCGAT 4 B 8 TTCCGATAACGAT SWB 4 E 84 CTTCCATAACGAT SWE 17 O 15 TCTAGAGGTCGAT SWO 17 B 90 CTAACCACGGCGAT 17 E 29 TCGACCACTCGAT 1F O 26 TTACAACCTCGAT SWO 1F E 18 AGGCAATTGCGAT 1F B 4 TACCAAGATCGAT 2O O 63 CCTTAGAGTTCGAT 2O E 68 TCAAGAAGTTCGAT SWE 2O B 70 CCTACTGGTCGAT 3I O 34 TCGCATCGTTCGAT SWO 3I E 58 TCCTAGAACACGAT SWE 3I B 12 TAGGTGGTTCGAT 11 O 67 TTCCTACCAGTCGAT SWO 11 E 27 AACCATCCGCGAT SWE 11 B 79 CCTGGTTGTCGAT 12 O 54 CCGGAGAATCGCGAT 12 E 44 TTGGAGGCCAGCGAT 12 B 20 CAGATCCATCGAT 13 O 81 CCTGCCATTCGCGAT SWO 13 E 19 TTAGTCGGACGAT SWE 13 B 77 CGAAGCGATTCGAT SWB 16 O 51 TTGAGCCTATTCGAT 16 E 55 TCCACCTCCTCGAT SWE 16 B 32 TCTTACACACGAT 21 O 37 CTTGAGAATGTCGAT 21 E 89 TCCTGAATCTCGAT 21 B 35 TAAGCCATTGTCGAT 23 O 69 TTCAATTGGCGAT 23 E 48 TTCTAAGAGACGAT SWE 23 B 43 CTTGACACCGCGAT SWB 24 O 92 CTAGGAACCGCGAT 24 E 86 CTTGGTTATTCGAT SWE 24 B 62 TTCCTGCTTCACGAT ugit.13 Neg. cont. 13 TCTAACGGACGAT ugit.14 Pos. cont. 14 TTGGAGTGTCGAT

2 Table S2. Number of long reads and amplicon sequence variants (ASVs) from initial PacBio CCS reads throughout the quality control and filtering steps used to generate a sequence dataset representing the Archaeorhizomycetes (Arch.) diversity at the study site. Numbers are presented for the three sequenced libraries (SWO, SWE, SWB) separate and combined (Across libraries). ASVs taxonomically assigned to Archaeorhizomycetes were aligned and manually checked for chimers, this identified one chimeric ASV (ASV_135) that was removed. In the end, the Archaeorhizomycetes “phylogenetic” sequence dataset (Included Arch. ASVs) represented 52,274 reads across three libraries.

Steps SWO SWE SWB Across libraries PacBio CCS 67,159 69,083 61,914 198,156 ASVs with ITS2 region 208 239 178 276 ASVs incl. All fungal ASVs 207 236 178 273 ASVs assigned to Arch. 33 35 31 42 Chimeric ASV_135 1 1 1 1 Included Arch. ASVs 32 34 30 41 Reads represented by 21,324 13,302 17,648 52,274 Arch. ASVs Rel. ab. Archaeo. 32% 19% 29% 26%

3 Table S3. Summary of 41 Archaeorhizomycetes ASVs in phylogenetic dataset Total SWO SWE SWB Fungi ASVs with ITS 112,485 35,526 38,767 38,192 Tot Archy seq 52,274 21,324 13,302 17,648 Rel abund Archy 46% 60% 34% 46% nr of Archy ASVs 42 33 35 31 ASV ID PSH ASV_1 14,049 262 395 13,392 PSH_9 A.fin ASV_3 9,472 5,630 3,468 374 PSH 8 ASV_4 7,837 5,862 1,089 886 PSH 7 ASV_5 4,916 1,503 2,348 1,065 PSH 1 ASV_7 2,935 2,093 832 10 PSH 2 ASV_8 1,971 1,922 49 PSH 2 ASV_12 1,797 1,326 353 118 PSH 8 ASV_19 1,111 1,078 30 3 PSH 1 ASV_23 811 6 805 PSH 1 ASV_24 785 20 765 PSH 2 ASV_29 681 80 601 PSH 2 ASV_31 604 183 29 392 PSH 3 ASV_35 533 1 532 PSH 2 ASV_37 507 500 4 3 PSH 8 ASV_38 505 9 495 1 PSH 2 ASV_48 371 4 159 208 PSH 2 ASV_59 278 1 277 PSH_9 A.fin ASV_63 266 3 227 36 PSH 1 ASV_66 261 241 11 9 PSH 8 ASV_68 251 2 231 18 PSH 1 ASV_73 241 4 234 3 PSH 5 ASV_82 219 2 106 111 PSH 2 ASV_84 213 138 32 43 PSH 8 ASV_86 202 20 179 3 PSH 5 ASV_88 193 140 30 23 PSH 8 ASV_90 184 5 32 147 PSH 6 ASV_104 154 79 75 PSH 3 ASV_106 149 149 PSH 3 ASV_120 121 121 PSH 5 ASV_122 118 118 PSH_9 A.fin ASV_134 100 100 PSH_9 A.fin ASV_136 98 98 PSH 3 ASV_155 76 76 PSH 3 ASV_176 57 9 48 PSH 5 ASV_181 53 2 51 PSH 8 ASV_210 36 28 4 4 PSH 8 ASV_220 32 32 PSH 6 ASV_230 25 1 24 PSH 3 ASV_238 21 2 3 16 PSH 5 ASV_239 21 21 PSH 4 ASV_240 20 5 15 PSH 3 Removed sequence ASV_135 Chimer 100 94 2 4

4 Table S4. Overview of Archaeorhizomycetes reads in the “ecological” dataset Rel. ab. Archaeo. PSHs without Fungal Archaeo. Archaeo. Archaeo. Archaeo. long reads Plot Horizon reads (%) reads itASVs PSHs Rel. ab. (%) Nr. of PSHs 1 B 22967 29 6613 30 12 0,3 5 1 E 28885 34 9680 20 7 0,0 0 1 * O 19882 32 6283 23 9 0,4 4 2 B 1175 10 113 19 11 0,5 3 2 * E 22255 63 13978 24 9 0,0 2 2 O 22950 46 10567 25 10 1,2 2 3 B 20046 30 5984 21 10 0,0 4 3 * E 21285 45 9574 21 6 0,0 0 3 * O 20297 39 7941 19 10 0,1 2 4 * B 24477 41 10132 38 19 0,4 11 4 * E 22234 31 6797 27 15 0,2 9 4 O 17508 34 6040 16 8 0,1 3 11 B 5673 21 1202 17 12 0,1 5 11 * E 17747 31 5465 12 4 0,0 0 11 * O 9569 28 2643 17 7 0,0 0 12 B 1141 14 165 18 14 1,2 9 12 E 15263 0 3 3 3 0,0 0 12 O 16290 4 684 20 16 0,2 13 13 * B 23884 59 14128 26 10 0,1 2 13 * E 22283 72 15976 17 8 0,0 0 13 * O 23818 44 10508 19 10 0,6 2 16 B 12695 27 3384 36 24 2,7 16 16 * E 20181 43 8672 13 7 0,0 1 16 O 9845 29 2903 16 8 0,1 2 17 B 24957 55 13760 57 31 31,3 22 17 E 20622 37 7722 38 21 0,4 12 17 * O 20329 38 7708 21 10 0,4 3 21 B 828 6 53 16 10 1,0 6 21 E 11915 30 3545 17 8 0,0 2 21 O 8682 52 4531 20 11 0,1 5 23 * B 9143 10 874 21 13 0,2 8 23 * E 22783 43 9888 22 9 0,2 3 23 O 18786 42 7804 13 5 4,2 1 24 B 22155 53 11791 49 32 42,0 26 24 * E 15570 16 2563 13 7 0,0 3 24 O 21056 18 3693 13 6 0,0 3 Average 17,199 33 6,483 22 12 2 5 Plot – Plot number, * indicates that the sample was included in “phylogenetic” dataset Horizon – organic (O), mineral elluvial (E) and mineral illuvial (B) Fungal reads – number of fungal reads Rel. ab. Archaeo (%) – relative abundance of Archaeorhiozmycetes Archaeo. reads – number of Archaeorhiozmycetes reads Archaeo. itASVs – number of itASVs assigned to Archaeorhiozmycetes Archaeo. PSHs – number of itASVs assigned to Archaeorhiozmycetes Archaeo. PSHs without long reads –Archaeorhiozmycetes PSHs not represented in “phylogenetic” dataset presented in two where Rel. Ab. (%) is the relative abundance and Nr. of PSHs is the number of PSH detected only in the “ecological” dataset.

5

Table S5. Summary statistics for 2-way ANOVAs on number of fungal reads, number of Archaeorhizomycetes itASVs and PSHs as well as relative abundance of Archaeorhizomycetes reads across Treatment and Horizon (sequentially added). Significant effects are indicated by *. Variable Df Sum Sq Mean Sq F value Pr(>F) Nr. of Treatment 3 1.681e+08 56021477 1.149 0.346 fungal Horizon 2 2.162e+08 108075315 2.216 0.127 reads Residual 30 1.463e+09 48776781 Nr. of Treatment 3 138.1 46.0 0.485 0.695 Arch. Horizon 2 848.4 424.2 4.472 0.020 * itASVs Residual 30 2845.8 94.9 Nr. of Treatment 3 93.8 31.26 0.916 0.44487 Archy Horizon 2 514.7 257.33 7.542 0.00222 ** PSHs Residual 30 1023.6 34.12 % reads in Treatment 3 1.747e-06 5.823e-07 0.849 0.478 Archy Horizon 2 3.217e-06 1.609e-06 2.345 0.113 Residual 30 2.058e-05 6.860e-07

Table S6. PERMANOVA testing for variables that significantly affect the relative abundance of nine PSHs. Model selection started with a full model with all variables (Horizon, Treatment, Plot) and their interactions. Variables were reduced by omitting the highest order non-significant term at each step. Fd ChiSquare F Pr(>F) Horizon 2 0.43471 4.6070 <0.001 ***

Plot 1 0.1896 4.0190 <0.001 ***

Treatment 3 0.5680 4.0130 <0.001 *** Plot*Treatment 3 0.2880 2.0344 <0.005 ** Residual 26 1.22665

6 Table S7. Kruskal Wallis tests of niche-specific distribution for nine PSHs (supported by long read data) based on the relative read abundance in the “ecological” dataset as the response variable and horizon as the explanatory variable. The p-values were corrected for multiple comparisons using the Benjamini & Hochberg method. Corrected p-values below 0.1 denoted by •.

Kruskal-Wallis test Chi-squared Degrees of Corrected P-value freedom (df) PSH_1 2.1613106 2 0.6109 PSH_2 1.6028322 2 0.6730 PSH_3 3.9627832 2 0.4137 PSH_4 0.8558524 2 0.7334 PSH_5 1.1127558 2 0.7334 PSH_6 3.0828607 2 0.4817 PSH_7 0.4166110 2 0.8120 PSH_8 5.2672673 2 0.3231 PSH_9 A. finlayi 9.4570478 2 0.0792•

Table S8. Frequency of observation, mean and maximum relative abundance in “ecological” dataset of Archaeorhizomycetes PSHs supported by long read “phylogenetic” dataset. PSH Nr of sample Ave rel. abundance Max relative abundance PSH_1 23 3% 30% PSH_2 27 5% 34% PSH_3 21 1% 24% PSH_4 6 0% 1% PSH_5 34 5% 37% PSH_6 23 1% 6% PSH_7 24 4% 36% PSH_8 36 9% 38% PSH_9 A.fin 29 4% 34%

Table S9. Niche specific distribution in two orthogonal contrasts defined from the phylogenetic tree in Fig 1. PHS number are listed for the two pairs of sister-PSH. Model results with degrees of freedom (Df), sums of squares, mean square sums, F-values, R2- values and P(Pr)-values are shown. P-values below 0.05 are marked with an asterisk and those below 0.1 with •. Df SumsOfSqs MeanSqs F.Model R2 Pr(>F) PSH_2 vs. PSH_3 1 0.6380 0.63805 1.7717 0.03878 0.09950• PSH_7 vs. PSH_8 1 0.6884 0.68842 1.9116 0.04184 0.02985* Residual 42 15.1253 0.36013 0.91937 Total 44 16.4518 1.00000

7 Supplementary figures

IF irrigation + fertilization

I irrigation only

N F fertilization only 0 negative control

CC clear cut post-study

UM unmanipulated plots

reserve boundary

meteorological tower 21

laboratory 23 fence 24

1

2 3 13 12 11 4 16 17

road 0 50m

Figure S1 Schematic view of plots sampled at Ivantjärnsheden field station in Jädraås. Redrawn from (Axelsson & Bråkenhielm, 1980). The stand was naturally regenerated after tree-felling in 1957 and thinned before the onset of an experimental field study conducted between 1974 - 1990 (experiment Ih2 (9802)). In the study 30 x 30 m plots received one of three treatments (irrigation and fertilization (IF) in green, irrigation (I) in blue, un-manipulated control (O) in brown as well as clear cut plots (CU) striped), labeled with plot number. The clear-cut (grey stripes) was part of a second study following the initial experiment and represent one plot from each previous treatment. Forest has since started to regenerated in clear-cut plots.

8

Figure S2 A maximum likelihood tree based on an alignment of the rDNA LSU region of 273 fungal ASVs from the current study as well as reference sequences A. borealis (KF993708), A. finlayi (JF836022) and the uncultured lineage GS31 (KY687760) which represents a sister class of Archaeorhizomycetes. Bootstrap support values are calculated from 1000 iterations and shown on the Archaeorhizomycetes branch and all deeper branches. Collapsed clades represent Basidiomycota (including 106 ASVs), Ascomycota outside the Taphrinomycotina (111 ASVs) and Mucoromycota (11 ASVs). Forty-two ASV LSU regions representing Archaeorhizomycetes form a well-supported sister clade to GS31 and are highlighted by a box on the tree.

9 0.4 PRICO2.2 PRICO1.2

PRICO2.1 PRICO5.2 PRICO5.1 PRICO5.3 PRICO3.1 PRICO1.1

PRICO3.3 PRICO2.3 Country 0.2 PRICO3.2 jadraas24_B a Benin PRICO6.2 PRICO1.3 PRICO7.1 USA.EM5UFMUSA.EM5UFSUSA.EM1UFM jadraas03_Bjadraas12_E USA.EM1UFS jadraas13_Bjadraas03_E a Burkina Faso PRICO4.3 PRICO7.3 USA.EM7UFM PRICO6.1India1.4 USA.EM2UFM jadraas02_Ejadraas13_E jadraas01_Ojadraas02_O USA.EM2UFS jadraas23_Ejadraas21_E a India PRICO4.2 jadraas01_Ejadraas03_Ojadraas13_O Africa1.4 PRICO7.2 jadraas04_Bjadraas12_Bjadraas17_E jadraas24_E PRICO6.3 Africa2.2 jadraas17_Ojadraas21_O a Kyrgystan negCTL jadraas23_Bjadraas01_Bjadraas11_E Africa1.3 jadraas17_B jadraas16_Bjadraas16_E 0.0 India1.1 jadraas12_O Africa2.3 Africa1.2 jadraas23_O a India1.3 India2.4 jadraas24_Ojadraas04_Ojadraas11_O Puerto Rico jadraas04_Ejadraas16_O

nMDS 2 jadraas02_Bjadraas21_B jadraas11_B a Sweden Africa1.1 Tajik2.3 India1.2 Africa2.4 India3.1 a USA India2.2 posCTL Tajik1.4 Tajik2.4 Tajik2.2India2.3Tajik1.3 Africa2.1 a India3.2 pos India2.5 −0.2 India3.3 India2.1 a neg Tajik1.1 Tajik1.2Tajik2.1 India3.5

PRICO4.1 India3.4

−0.4 −0.2 0.0 0.2 nMDS 1

Figure S3 Non-parametric multidimensional scaling (nMDS) ordination displaying communities (fungal ASVs only, positive control sequence (itASV_1) removed; 4,461 ASVs / 1,680,044 sequence reads included) from all 96 samples sequenced on the ugit_102 IonTorrent sequencing run (plot stress = 0.24 based on a maximum of 200 random starts). Prior to ordination, the itASV by sample count matrix was standardized to per- sample relative proportions, and a dissimilarity matrix was calculated using the Bray-Curtis dissimilarity index. Samples in the current study (in yellow) was sequenced on an IonTorrent chip together with samples from several other studies including unpublished samples from Benin, Burkina Faso, India and Kyrgystan, as well as published data from Puerto Rico (Urbina et al. 2016) and IN, USA (Rosling et al. 2016). Positive (pos) and negative (neg) controlls were also included. Samples generally cluster by study site. However jadraas24_B clusters with samples from Puerto Rico possibly indicating tag-swithich affecting this particular sample.

10

10% 100%

9% 90%

8% 80%

7% 70%

6% 60%

5% 50%

4% 40% % of fungal community fungal of % 3% 30%

2% 20% Cumulative % of Archaeorhizomycetes community Archaeorhizomycetes of % Cumulative 1% 10%

0% 0% 28 32 38 31 30 42 36 48 62 51 63 34 35 44 52 PSH 9 PSH 24 PSH 19 PSH 18 PSH 20 PSH 13 PSH 11 PSH 10 PSH 15 PSH 27 * PSH 8 PSH * 2 PSH * 5 PSH * 7 PSH * 1 PSH * 3 PSH * 6 PSH * 4 PSH *

* PSH_9 A. fin A. PSH_9 * Figure S4 Relative sequence read abundance of the 34 most abundant Archaeorhiozmycetes PSHs, delimited based on long- and short- environmental amplicon sequences from soil sampled in a sounth central Swedish pine forest. These represent half of the total number of PSHs delimited at the site. Primary axis illustrates percentage of each PSH out of the total fungal community based on reads across all 36 samples. PSHs highlighted in dark red and indicated with an * before its labele are supported by the combined dataset of long amplicon data from the “phylogenetic” dataset and short ampicon data from the “ecological” dataset. Secondary axis illustrates cumulative percentage of each PSH as a propostion of the total Archaeorhizomycetes community at the site, based on sequence reads in the ecological dataset.

11 0.6

0.4

perc_fungal_reads_archy 0.2

0.0

B E O horizon

Figure S5 Proportion (%) of reads assigned to class Archaeorhizomycetes out of all reads assigned to kingdom Fungi in the “ecological” dataset. Based on 12 samples from each of the soil horizons, organic soil (O), elluvial mineral soil (E) and illuvial mineral soil (B). Individual samples are visualized as grey dots.

12

a)

b)

Figure S6 Comparisons across horizons for a) number of Archaeorhizomycetes itASVs and b) number of Archaeorhizomycetes PSHs across soil horizons. These variables were significantly different based on 2-way ANOVA (Table S5).

13

Figure S7 Diversity of Archaeorhizomycetes in mid-Sweden Pine forest podzol soil illustrated in maximum likelihood tree based on environmental ribosomal long and short amplicon ASVs representing 68 phylogenetic species hypotheses (PSH) including both described species A. borealis and A. finlayi. The sister taxa GS31 and four Taphrinamycotina species as outgroup. Nodes are cartooned to visualize PSHs represented by more than one ASV. All Archaeorhizomycetes ASV from “phylogenetic” and “ecological” dataset are labled at the nodes. Bootstrap support are calculated from 1000 iterations.

14

Figure S8 Results from nMDS analysis based on the relative abundance of 68 delaminated Archaeorhizomycetes PSHs (red circles labeled PSH_number, A.fin, A.bor or Xnumber (for single itASVs)). Stress value for the nMDS is 0.2058808.

15

18%

16% PSH_7

14% PSH_8

12%

10%

8%

6%

4%

2%

0% O E B 45% 40% 35% 30% 25%

PSH_7 20% 15% 10% 5% 0% 0% 5% 10% 15% 20% 25% 30% 35% 40% PSH_8

Figure S9 Relative abundance of PSH_7 vs PSH_8 a) average for each horizon with SEM in error bars and b) relative abundance of the two sister PSHs across all 36 samples included in the current study.

Cited references

Axelsson, B., & Bråkenhielm, S. (1980). Investigation sites of the Swedish coniferous forest project - biological and physiological features. In T. Persson (Ed.), Structure and Function of Northern Coniferous Forests - An Ecosystem Study (pp. 25 - 64). Berlings.

16