bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1 Fungal Communities in Soybean Cyst Nematode-Infested Soil under 2 Long Term Corn and Soybean Monoculture and Crop Rotation

3

4 Noah Strom1, Weiming Hu2, Deepak Haarith3, Senyu Chen4, Kathryn Bushley1*

5

6 1Department of Plant and Microbial Biology, University of Minnesota, Saint Paul, MN, USA;

7 2Entomology and Nematology Department, University of Florida, Gainesville, FL, USA;

8 3Department of Plant Pathology, University of Minnesota, Saint Paul, MN, USA.

9 4Southern Research and Outreach Center, University of Minnesota, Waseca, MN, USA;

10 *Correspondence:

11 Kathryn Bushley

12 [email protected]

13

14 Abstract

15 Corn (Zea mays) and soybean (Glycine max) production forms an integral part of economies 16 worldwide, but yields are limited by biotic and abiotic factors associated with short rotations and long- 17 term monocultures. In this study, a long-term rotation study site with corn and soybean planted in 18 annual rotation, five-year rotation, and long-term monoculture was utilized to examine the relationships 19 between crop sequences, soil fungal communities, soybean cyst nematode (SCN, Heterodera glycines) 20 densities, soil properties, and crop yields. High throughput sequencing of the ITS1 region of fungal 21 rDNA revealed that soil fungal community structure varied significantly by crop sequence, with fungal 22 communities under five consecutive years of monoculture becoming progressively similar to 23 communities in long-term monoculture plots associated with their respective crop hosts. Total alpha 24 diversity was greater under corn, but patterns of diversity and relative abundance of specific functional 25 groups differed by crop host, with more pathotrophs proliferating under soybean and more saprotrophs 26 and symbiotrophs proliferating under corn. Soil phosphorus (P) varied significantly by crop sequence, 27 with lower levels of P corresponding with relative abundance of Glomerales, Paraglomerales, and 28 Sebacinales and higher levels of P corresponding with relative abundance of Mortierellales. Soil 29 density of the SCN was positively correlated with relative abundance and diversity of nematode- 30 trapping fungi and with relative abundance of many potential nematode egg parasites. These results 31 suggest several possible explanations for the improved yields associated with crop rotation, including 32 decreased pathogen pressure, modification of soil properties, and increased diversity of soil fungal 33 communities. Future research should investigate the potential of nematode-trapping fungi to regulate 34 SCN densities and examine the relationships between soil P and specific arbuscular mycorrhizal and 35 mortierellalean fungi associated with corn and soybean hosts. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

36 Keywords: Crop rotation, monoculture, soybean cyst nematode, nematophagous fungi, soil 37 phosphorus, mycorrhizae, metabarcoding

38 1. INTRODUCTION

39 Corn (Zea mays) and soybean (Glycine max) production forms an integral part of the economies of 40 multiple nations in both the eastern and western hemisphere (Hartman et al., 2011; Meade et al., 2016) 41 and plays a major role in the food security of developing nations (Hartman et al., 2011; Shiferaw et al., 42 2011). In the United States in 2016, corn and soybean accounted for 53% of total acreage planted to 43 principal crops and 49% of principal crop production (U.S. Department of Agriculture National 44 Agricultural Statistics Service, 2017). However, both crops suffer yield penalties under continuous 45 monoculture. Corn yield is thought to be limited primarily by levels of soil nitrogen (N), whereas the 46 main factor thought to limit soybean yield is disease pressure from plant pathogens, especially the 47 soybean cyst nematode (SCN, Heterodera glycines) (Gentry et al., 2013; Seifert et al., 2017). Other 48 soil properties, especially soil phosphorus (P), may be involved in yield declines of both crops (Bender 49 et al., 2015; Xin et al., 2017). Soil fungi, including arbuscular mycorrhizal fungi (AMF), interact with 50 plants, plant pathogens, and soil properties in complex ways that may affect crop yields (Francl and 51 Dropkin, 1985; Johnson et al., 1992; Tylka et al., 1991).

52 The SCN is responsible for the largest soybean yield declines, averaging 3.5 million metric tons 53 in yield losses in the U.S., annually (Koenning and Wrather, 2010). While crop rotation to corn is the 54 most widely adopted practice to reduce SCN populations, even five years of continuous corn cropping 55 cannot eliminate this pathogen (Porter et al., 2001). With few available sources of genetic resistance 56 to SCN (Liu et al., 2017) and the phasing out of chemical nematicides due to environmental and human 57 toxicity (Warnock et al., 2017), biocontrol of SCN using nematophagous fungi is being investigated as 58 an alternative to manage this pathogen (Chen and Dickson, 2012). Understanding how these fungi 59 respond to common corn-soybean crop sequences may help identify fungal taxa that could serve as 60 biocontrol agents.

61 Methods for characterizing soil fungal communities have undergone significant advances in 62 recent years. Early studies used denaturing gradient gel electrophoresis (DGGE) to demonstrate 63 seasonal shifts in bacterial and fungal communities in plant rhizospheres and to a lesser extent in bulk 64 soils (Gomes et al., 2003; Smalla et al., 2001; Smit et al., 2001), but this technique is known to 65 underestimate species diversity (Gomes et al., 2003). In the last decade, high throughput amplicon 66 sequencing has facilitated more complete characterization of agricultural soil microbial communities 67 than earlier culture-based and molecular approaches (Lindahl et al., 2013). Many of these studies have 68 demonstrated the effects of continuous monoculture cropping on the relative abundance of plant 69 pathogens (Bai et al., 2015; Liu et al., 2014b; Wu et al., 2016; Xiong et al., 2014), while others have 70 described the dynamics of beneficial microbes, including AMF (Martínez-García et al., 2015; Senés- 71 Guerrero and Schüßler, 2016) and antagonists of plant pathogens (Hamid et al., 2017; Xu et al., 2012), 72 including those with hypothesized roles in SCN-suppressive soils (Hu et al., 2017). High throughput 73 amplicon sequencing has also been used to examine the effects of tillage (Hu et al., 2017; Smith et al., 74 2016), organic farming practices (Hartmann et al., 2014), fertilizer use (Lin et al., 2012; Ramirez et al., 75 2010), crop rotation (Bainard et al., 2017), land-use, and soil properties (Lauber et al., 2008) on soil 76 microbial communities.

77 In addition to their effects on soil microbial communities, cropping sequences are known to 78 affect soil physical and chemical properties that could impact yield. For example, rotation of corn to 79 soybean has been shown to increase bioavailable N, which improves corn yield (Peterson and Varvel, bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

80 1989). Likewise, rotation of soybean to corn has been shown to improve soil aggregate stability and 81 water holding capacity and to increase soil levels of organic matter, P, K, Mg, and Ca, properties that 82 are associated with increased soybean yield (Perez-brandan et al., 2014). However, another study 83 found that soil P was higher under continuous soybean monoculture compared to continuous corn 84 monoculture at two sites in Minnesota, including our study site (Johnson et al., 1991). AMF, which 85 are canonically thought to enhance plant growth through the uptake and transport of soil P (Parniske, 86 2008), undergo community shifts related to corn-soybean crop sequences (Johnson et al., 1991). It has 87 been hypothesized that the AMF communities associated with corn and soybean becomes less 88 mutualistic under continuous monoculture of each host, potentially contributing to yield declines 89 (Johnson et al., 1992).

90 The population density of nematophagous fungi may also be affected by cropping sequences. 91 For example, the nematode-endoparasitic fungus, Hirsutella rhossiliensis, which is thought to have a 92 density-dependent relationship with SCN (Jaffee et al., 1992), was more commonly isolated from SCN 93 juveniles during soybean cropping years with higher SCN density (Chen and Reese, 1999). Likewise, 94 the SCN egg parasites, Metacordyceps chlamydosporia and Purpureocillium lilacinum, were shown to 95 increase in relative abundance in the soybean rhizosphere and in SCN cysts over continuous soybean 96 monoculture in fields in northeastern China (Hamid et al., 2017; Song et al., 2016a). Nematode- 97 trapping fungi Dactylellina ellipsospora (syn. Monacrosporium ellipsosporum) and Arthrobotrys 98 dactyloides (Syn: Drechslerella dactyloides) have been shown to have a density-dependent relationship 99 with the root knot nematode (Meloidogyne javanica) under greenhouse conditions (Jaffee et al., 1993). 100 However, little is known about the relationships between common corn-soybean crop sequences and 101 SCN-antagonistic fungi in bulk soil.

102 In addition to studying shifts in the abundance of fungal taxa, we also sought to address changes 103 in functional guilds of fungi across crop rotation sequences. The concept of an ecological guild dates 104 to Root (1967), who defined a guild as a group of species that consume the same resource in the same 105 way. Recent approaches have broadly classified fungi according to both their trophic mode and 106 ecological guild (Nguyen et al., 2016). In addition to these broad ecological categories, three guilds of 107 nematophagous fungi are recognized: i) near-obligate endoparasites of free-living nematodes, ii) 108 nematode egg parasites, and iii) nematode-trapping fungi (Chen and Dickson, 2012). A fourth group 109 of nematode-antagonistic fungi consisting of species that kill nematodes through the production of 110 antibiotics is often described (Chen and Dickson, 2012), but this group does not precisely meet the 111 definition of a guild, as its description does not specify a feeding mode. To our knowledge, this is the 112 first study to comprehensively investigate the bulk soil fungal community associated with corn and 113 soybean rotations from the standpoint of trophic modes and nematophagous guilds in addition to 114 individual taxa or operational taxonomic units (OTUs).

115 This study utilized a unique long-term research site in Waseca, MN, at which corn and soybean 116 have been planted under continuous long-term monoculture, annual rotation, and 5-year rotations since 117 1982. A previous study conducted at the same site showed that populations of SCN increased under 118 continuous soybean monoculture, whereas populations of the plant parasitic nematodes, Pratylenchus 119 (lesion nematode) and Helicotylenchus (spiral nematode), increased under continuous corn 120 monoculture (Grabau and Chen, 2016a, 2016b). These nematode populations were correlated with 121 yield declines in soybean and corn, respectively. However, it is unknown whether increases in SCN 122 density at this site are correlated with increases in the relative abundance of SCN-antagonistic fungi in 123 the bulk soil, knowledge that could have profound implications for biocontrol. The objectives of this 124 study were to (i) investigate the role of corn/soybean crop rotations and continuous monoculture in 125 shaping bulk soil fungal communities, and (ii) identify specific fungal taxa or functional guilds bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

126 correlated with corn and soybean yield, soil properties, and SCN density. We hypothesized that (i) 127 bulk soil fungal communities under consecutive years of monoculture would become progressively 128 more similar to those in long-term monoculture plots; (ii) crop rotation would result in greater alpha 129 diversity of fungal communities compared to continuous monoculture; and (iii) nematophagous fungi 130 would increase in diversity and relative abundance over continuous soybean monoculture, 131 corresponding with an increase in density of the SCN.

132 2. MATERIALS AND METHODS

133 2.1 Experimental design

134 The experimental design used in this study is described in previous studies conducted at the same site 135 (Grabau and Chen, 2016b, 2016a; Hu et al., 2018). Long-term rotation plots at the University of 136 Minnesota Southern Research and Outreach Center in Waseca, MN (44°04' N, 93°33' W), where corn 137 and soybean have been planted in continuous cropping sequences since 1982, were used to study the 138 effects of crop sequences on the bulk soil fungal community in 2015 and 2016. The crop sequences in 139 this study were continuous corn and soybean monocultures (Cc and Ss), 5-year rotations (C1-C5 and 140 S1-S5), annual rotations (Ca and Sa), and non-Bt corn and SCN-resistant soybean monocultures (Cn 141 and Sr) (Table S1). Since 2010, all treatment plots except for Cn and Sr plots were planted with Bt 142 corn (Dekalb 50-66) and SCN-susceptible soybean (Pioneer 92Y12). The Cn and Sr plots were planted 143 with non-Bt corn (Dekalb 50-67) and SCN-resistant soybean (Pioneer 92Y22), respectively, since 144 2010, but prior to 2010 all soybean plots were planted with SCN-susceptible soybeans (Table S1). 145 There were four replicates for each treatment, arranged in the field in a randomized complete block 146 design.

147 Experimental plots were 4.57 m wide by 7.62 m long and contained 6 rows of plants, each. 148 Plots were managed by conventional tillage practices consisting of fall chisel plowing and field 149 cultivation prior to planting. All crops were resistant to glyphosate (Roundup) herbicide, which was 150 used at a rate of 2.2 kg/ha to prevent weed growth. Corn and soybean plots were sprayed with Endigo 151 insecticide at a rate of 245 g/ha at midseason in 2015 for aphid control. No insecticides were used in 152 2016. Corn plots were fertilized with nitrogen as urea at a rate of 224.4 kg/ha at midseason in 2015 153 and 2016. No phosphorus-containing fertilizer was applied in 2015 and 2016, but P-K fertilizer was 154 applied at a rate of 80 and 120 lbs. per acre, respectively, to all plots in 2014.

155 2.2 Sample collection

156 Bulk soil samples were collected in the spring (at planting), midseason (2-3 months after planting), and 157 fall (at harvest). A total of 20 soil cores were taken at regular intervals from between the two central 158 rows of each plot using a 2.54 cm diameter probe sunk to a depth of 20 cm. Soil samples were stored 159 in a cold room at 4 oC until further processing on the same day. Soil cores from each plot were pooled, 160 pushed by hand through a 5 mm mesh screen to break apart cores into smaller aggregates, and 161 thoroughly mixed by hand. A subsample of 100 g of homogenized soil was used for SCN egg 162 quantification. Soil properties analysis was performed on 100 g of homogenized soil collected in the 163 spring by the Research Analytic Laboratory at the University of Minnesota. The remaining 164 homogenized soil (250 g) was stored in a resealable plastic bag at -80 oC for later processing.

165 2.3 SCN egg density quantification and yield measurement

166 The density of SCN eggs in the bulk soil was determined following the methods described in Hu, et al. 167 (2017). Briefly, cysts were separated from 100 cm3 of bulk soil by elutriation followed by bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

168 centrifugation in a 63% (w/v) sucrose solution. Cysts were crushed to release eggs (Niblack et al., 169 1993). Eggs were then collected in water and quantified by examining a subsample of egg suspension 170 with an inverted microscope. This number was used to calculate the total number of eggs in 100 cm3 171 of soil. Yield was measured from 20 feet of the two central rows of each experimental plot using a 172 plot combine.

173 2.4 DNA extraction, amplification, and sequencing

174 Fifty grams of soil from each 250 g soil sample was homogenized further in a coffee grinder, and DNA 175 was extracted from 0.25 g of this soil using a MoBio PowerSoilâ DNA Isolation Kit. Twenty 176 microliters of DNA per sample were submitted to the University of Minnesota Genomics Center (Saint 177 Paul, MN, USA) for amplification and sequencing. Additional samples to assess the quality and 178 accuracy of the amplification and sequencing were also included. These included a fungal mock 179 community, a synthetic mock community (Palmer et al., 2018), technical replicates consisting of 180 replicate DNA extractions from the same soil sample, and negative controls in which no soil was added 181 to the extraction kit. For a description of the preparation and analysis of mock communities, see the 182 Supplementary text.

183 A two-step dual-indexed amplification method was used for amplifying fungal internal 184 transcribed spacer 1 (ITS1) regions of fungal rDNA (Gohl et al., 2016). Paired-end sequencing (2 x 185 250 bp) was carried out using a MiSeq 600 cycle v3 kit on a total of four lanes. 2015 fall and midseason 186 bulk soil samples were sequenced on one lane, and spring 2015 bulk soil samples were sequenced on 187 a separate lane along with technical replicates and a pool of 51 other samples. All 2016 bulk soil PCR 188 products were pooled and sequenced on two lanes, with half of the second lane shared with a pool of 189 229 other samples. Fungal mock community samples were included in all four lanes. Illumina 190 sequencing on these four lanes resulted in a total 47,902,914 reads from bulk soil samples that passed 191 Illumina quality control. A total of 7 bulk soil samples, which had fewer than 20,000 reads after quality 192 control, were re-submitted for amplification and sequencing, resulting in an additional 744,408 quality 193 reads.

194 2.5 Bioinformatics Processing

195 Fastq files of reads that passed Illumina quality control were stored and processed at the Minnesota 196 Supercomputing Institute (MSI, Minneapolis, MN, USA). The 2016 bulk soil fastq files from identical 197 samples in separate lanes were merged using the fastq-cat.pl script in the Gopher-biotools package 198 (Garbe, 2015). Preprocessing and merging of reads, OTU clustering, filtering, and curating were 199 performed using the AMPtk v1.1.0 pipeline with default parameters, except where noted (Palmer et 200 al., 2018). The trim length for illumina reads was set to 250 bp for pre-processing. Processed and 201 merged reads were clustered in AMPtk using the "dada2" option, which first clusters sequences with 202 DADA2 v1.6 (Callahan et al., 2016) and then clusters the output iSeqs of DADA2 into OTUs using 203 USearch v9.2.64 (Edgar, 2013). The OTU table was filtered using AMPtk's default index bleed 204 filtering rate of 0.05% and curated using LULU v1.1.0, which merges "split" OTUs, resulting in better 205 alpha diversity estimates (Frøslev et al., 2017).

206 Based on the recovery of taxa from our fungal mock community, a synthesis of four different 207 methods and a custom R script (available at https://github.com/stro0070/OTU-taxonomy-assignment) 208 was used for taxonomy assignment (Supplementary text). This new approach to assigning taxonomy 209 to OTUs resulted in a greater percentage of recovered taxa from our fungal mock community, as well 210 as greater accuracy and precision of taxonomy assignments compared to any single method. The OTU bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

211 table with taxonomy was filtered to include only fungal OTUs and members of the synthetic mock 212 community. The AMPtk pipeline resulted in 12,068 OTUs, of which 11,701 remained after filtering 213 to remove non-fungal OTUs and adding back synthetic mock community OTUs. The final OTU table 214 was filtered to remove OTUs with fewer than 10 reads across the entire table, after which 8,587 OTUs 215 remained, of which 8,491 were present in bulk soil samples. After filtering, sequence depth for bulk 216 soil samples ranged from 16,435 to 205,157 reads with an average sequence depth of 63,947. The 217 rarefaction curves for most samples began to level off by around 30,000 reads, indicating the sampling 218 depth was adequate to capture most of the fungal diversity (Figure S1).

219 Fungal taxonomy is currently undergoing substantial revision. As of the date of this writing, 220 some phylum names in the UNITE database are outdated, and subphyla names are not included 221 (UNITE Community, 2017). We decided to retain the UNITE database taxonomy assignments without 222 modification due to the complex nature of the effort that would be required to update them. Thus, we 223 report, for example, on the abundance of Glomeromycota rather than Glomeromycotina and on 224 Mortierellomycota instead of Mortierellomycotina (Spatafora et al., 2016). Results of taxonomy 225 assignments, fungal and synthetic mock communities, and technical replicates analysis are described 226 in the Supplementary text, Tables S2-5, and Figures S2 and S3.

227 FUNGuild v1.1.0 (Nguyen et al., 2016) was used to assign ecological guilds and trophic modes 228 to OTUs. Only OTUs that were assigned a trophic mode with a confidence ranking of "Probable" or 229 higher were used for statistical analyses of trophic modes. The nematophagous fungal guilds, 230 "endoparasites," "egg parasites," and "trapping fungi," were defined by performing literature searches 231 for SCN-parasitic fungi (keywords: "soybean cyst nematode", "biological control," "fungi," 232 "nematophagous", "trapping", "predator," "endoparasite", "egg parasite") (Hu et al., 2018). Taxa 233 belonging to each nematophagous guild are listed in Table S6.

234 2.6 Statistical analyses

235 All statistical analyses were performed in RStudio v 3.5.0 (R Core Team, 2018; RStudio Team, 2016). 236 All plots except for distance-based redundancy analysis were created in "ggplot2" (Wickham, 2009). 237 In all analyses in which ANOVA was used, Levene's test (Levene, 1960) was used to ascertain equal 238 variance, and all P-values associated with multiple hypothesis testing were corrected by the false 239 discovery rate (FDR) procedure (Benjamini and Hochberg, 1995).

240 2.6.1 Yield and SCN density

241 Significant differences in yield were detected using ANOVA followed by Tukey's HSD test (Tukey, 242 1949). Due to unequal variance that could not be corrected through transformation, the Kruskal-Wallis 243 test (Kruskal and Wallis, 1952) was used to compare SCN density by crop sequence and by season for 244 individual crop hosts. Post-hoc comparisons were performed using the "kruskal" function (Mendiburu, 245 2017). Correlations between cube root-transformed SCN density and susceptible soybean yield were 246 tested by fitting linear models (Yield ~ SCN Density) with the "lm" function in R (R Core Team, 2018) 247 followed by ANOVA.

248 2.6.2 Soil properties

249 The mean and standard error for soil properties associated with each crop sequence were calculated 250 using the "aggregate" function in the "stats" package in R (R Core Team, 2018). Significant effects of 251 crop sequence on individual soil properties were tested using ANOVA on the model Y ~ Block + 252 CropSequence. A square root transformation was used for Fe in 2015 and 2016 and for NO3 in 2016 bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

253 to meet the ANOVA assumption of equal variance. Individual soil properties were tested for 254 correlations with fungal community dissimilarity in spring 2015 and 2016 using the "adonis" function 255 with 9999 permutations in the R package, "vegan" (Oksanen et al., 2016). Vectors for each soil 256 property were generated using the "envit" function. Correlations between individual soil properties 257 and yield were tested using the "lm" function followed by ANOVA.

258 2.6.3 Differential abundance

259 Differential abundance analyses were performed at the phylum and species level, for trophic modes, 260 and for nematophagous guilds. Due to the high false discovery rate associated with using proportions 261 for differential abundance testing (Weiss et al., 2017), read counts were instead transformed to centered 262 log ratios (CLRs) (Aitchison, 1982; Martín-Fernández et al., 2003), which are associated with a much 263 lower false discovery rate (Gloor et al., 2017). This is the same approach used in the R package 264 "ALDEx2," (Gloor, 2018) and has been used in other metabarcoding studies (Lee et al., 2014; 265 Mcdonald et al., 2018). CLRs of taxa, trophic modes, and nematophagous guilds with equal variance 266 according to Levene's test were compared across crop sequences using ANOVA on the linear model, 267 Y ~ Block + CropSequence for each sampling timepoint (e.g. Spring 2015) and across seasons using 268 ANOVA on the linear model Y ~ Block + CropSequence + Season + CropSequence:Season for each 269 sampling year. For CLRs that failed Levene's test, the Kruskal-Wallis test (Kruskal and Wallis, 1952) 270 was used, instead. Post-hoc comparisons were made using Tukey's HSD test (Tukey, 1949) or the 271 "kruskal" function with FDR correction in the R package, "agricolae" (Mendiburu, 2017).

272 Spearman correlation tests were used to test for correlations between CLRs of taxa, trophic 273 modes, and nematophagous guilds and soybean and corn monoculture year, SCN density, and yield in 274 each SeasonYear (e.g. Fall 2016). Data from Sr and Cn plots were removed prior to testing for 275 correlations between taxa/guilds and yield, because these cultivars were a different genotype than 276 plants in the majority of treatments. In addition to removing data from Sr and Cn plots, data from Sa 277 and Ca plots were removed prior to testing for correlations between taxa and monoculture year of either 278 host, because annual rotation plots could not be assigned a numeric monoculture year.

279 2.6.4 Fungal community diversity

280 Beta diversity analyses were performed on the OTU table subset by each SeasonYear (e.g. Fall 2016). 281 Read counts were converted to relative abundances and square root transformed (Hellinger 282 transformation) (Legendre and Gallagher, 2001), and a Bray-Curtis dissimilarity matrix was 283 constructed using these transformed data (Bray and Curtis, 1957). Non metric multidimensional 284 scaling (NMDS) (Kruskal, 1964) was performed using the "metaMDS" function in vegan (Oksanen et 285 al., 2016). The R2 and P-values for the effect of crop sequence on fungal communities were calculated 286 using the "adonis" function in vegan with 9999 permutations on the formula, Y ~ Crop Sequence. 287 Additionally, the adonis function was used to parse the effects of block, season, and crop sequence on 288 fungal community dissimilarity data subset by year, using the formula Y ~ Block + Season + 289 CropSequence + Season:CropSequence. Distance-based redundancy analysis was performed using the 290 "capscale" function in vegan (Legendre and Anderson, 1999; Oksanen et al., 2016) using an OTU table 291 collapsed by order with the QIIME2 "collapse" plugin (Caporaso et al., 2010). Read counts were 292 Hellinger transformed prior to this analysis. The analysis was constrained by soil properties, spring 293 SCN density, and yield and scaled by "species".

294 Alpha diversity metrics were calculated using the "estimate_richness" function in the 295 "Phyloseq" package (McMurdie and Holmes, 2013). Effects of crop sequence and season on alpha bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

296 diversity were tested for individual years using ANOVA on the same model as for beta diversity. 297 Pairwise comparisons across crop sequences and seasons were made using Tukey's HSD test (Tukey, 298 1949). Spearman correlation tests were used to test for correlations between the diversity of 299 nematophagous guilds and trophic modes and SCN density, yield, and monoculture year of each crop 300 host. Rarefaction curves were generated using the "rarecurve" function in vegan (Oksanen et al., 2016).

301 3. DATA ACCESSION

302 The raw sequences from bulk soil samples were deposited into the NCBI database (Accession number: 303 PRJNA484933).

304 4. RESULTS

305 4.1 Soybean cyst nematode density and crop yields

306 Corn yields were significantly higher in years following soybean (C1 and Ca), especially in the first 307 year of corn after five years of soybean (C1) (Figure 1A). However, long-term corn monoculture (Cc) 308 yield was not significantly lower than corn yield from earlier years of corn monoculture (C1-C5) 309 (Figure 1A). By contrast, soybean did not have significantly higher yields in the first year following 310 five years of corn (S1) or in annual rotation with corn (Sa) compared to other soybean crop sequences 311 (Figure 1B). In 2015, SCN-resistant soybean (Sr) had significantly greater yields than SCN- 312 susceptible soybean (Ss), which had significantly lower yields than SCN-susceptible soybean under 313 shorter-term monocultures (S3 and S5) (Figure 1B).

314 SCN density differed significantly by crop sequence in 2015 and 2016 according to the Kruskal- 315 Wallis test (P < 0.0001) (Figure 1C). In general, SCN egg density decreased with increasing years of 316 corn monoculture and increased with increasing years of soybean monoculture (Figure 1C). SCN 317 density was also significantly lower in SCN-resistant soybean plots compared to other soybean plots 318 (Figure 1C). Significant (P < 0.01) negative correlations between 2016 soybean yield and 2016 319 midseason and fall SCN density were observed (Figure 1D). When SCN density data were subset by 320 crop host, a significant effect of season was observed for SCN-susceptible soybean in 2016 (P < 321 0.0001), with higher SCN density in fall compared to spring or midseason (Figure S4). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Corn Yield by Crop Sequence B Soybean Yield by Crop Sequence

2015 2016 2015 2016

16 a P < 0.001 P < 0.001 5.0 P < 0.001 P = 0.06 a b a 14 b abc a 4.5 ab a e cd bc bc ab cde cde c a ab ab 12 d d de d d d bc ab 4.0 ab b ab 10 c

3.5 8 Yield (metric tons / ha) Yield (metric tons / ha) Yield 6 3.0 C1 C2 C3 C4 C5 Cc Cn Ca C1 C2 C3 C4 C5 Cc Cn Ca S1 S2 S3 S4 S5 Ss Sr Sa S1 S2 S3 S4 S5 Ss Sr Sa C SCN Density by Crop Sequence D Susceptible Soybean Yield by SCN Density

2015 2015 2016 2500 ab bc bc a y = 4.1 − 0.0064, R2 = 0.0074, P = 0.792 y = 3.9 − 0.015, R2 = 0.039, P = 0.465 P < 0.001 y = 4.2 − 0.017, R2 = 0.07, P = 0.34 4.5 y = 4.2 − 0.062, R2 = 0.34, P = 0.0045 2000 y = 4.1 − 0.00097, R2 = 0.00013, P = 0.95 y = 4.3 − 0.06, R2 = 0.33, P = 0.0045 bc bcd ab 1500 4.4 abc bc 1000 cde 4.0 500 def efg fg fg 4.0 g g 0

2016 2500

Yield (metric tons / ha) Yield 3.5 P < 0.001 3.6 2000

SCN eggs / 100 cc soil a 1500 b−g abc ab 1000 a−d a−g a−e a−f c−g a−d a−f 0 5 10 15 0 5 10 15 fgh SCN Density (eggs / 100 cc soil) (cube root transormed) 500 d−h gh e−h h 0 Season Spring Midseason Fall 322 C1 C2 C3 C4 C5 Cc Cn Ca S1 S2 S3 S4 S5 Ss Sr Sa 323 Figure 1. Corn (A) and soybean (B) yields were compared using ANOVA on the linear model, Yield ~ Block + Crop 324 Sequence. Reported P-values are associated with the Crop Sequence term in this model. Different lowercase letters denote 325 significant differences in mean yield as determined by Tukey's honestly significant difference (HSD) test. SCN density 326 (C) was compared across crop sequences using the Kruskal-Wallis test on data pooled from all seasons (spring, midseason, 327 and fall) within a sampling year, and significance categories were calculated using the "kruskal" function in R at a 328 significance level of a = 0.05 with false discovery rate (FDR)-correction. Error bars display one standard error above and 329 below the mean. (D) Relationships between soybean yield and SCN density are described by linear models. P-values 330 associated with these models are FDR-corrected. Crop sequences include 5 consecutive years of corn after the 5th year of 331 soybean (C1-C5); 5 consecutive years of soybean after the 5th year of corn (S1-S5); soybean in annual rotation with corn 332 (Sa); corn in annual rotation with soybean (Ca); long-term Bt corn (Cc) and SCN-susceptible soybean (Ss) monocultures; 333 non-Bt Corn (Cn) and SCN-resistant soybean (Sr) monocultures.

334 4.2 Fungal community comparisons across crop sequences and seasons

335 Adonis, a non-parametric multivariate ANOVA used to identify sources of variation between 336 communities (Anderson, 2001; Oksanen, 2015), revealed significant effects of crop sequence and 337 season on fungal community dissimilarity, with crop sequence explaining the highest proportion of 338 dissimilarity in both 2015 and 2016 (Table 1). NMDS plots from each sampling timepoint displayed 339 a pattern or trend in which long-term continuous monoculture communities showed the clearest 340 separation by crop host, with long-term soybean monoculture (Ss and Sr) communities on opposite 341 sides of the NMDS plots from long-term corn monoculture (Cr and Cn) communities (Figure 2A). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

342 Between these extremes, communities generally became more similar to the long-term monoculture 343 communities with increasing years of monoculture of each crop. For example, communities from later 344 years of corn or soybean monoculture were more similar to their respective long-term monoculture 345 communities, with communities from S3, S4, and S5 mapping close to Ss and Sr and communities 346 from C3, C4, and C5 mapping close to those from Cc and Cn. Communities associated with first year 347 crop sequences (S1 and C1), however, were more similar to fifth year crop sequence communities 348 associated with the alternate crop (C5 and S5), especially in spring. These communities from first year 349 crop sequences became increasingly more similar to their respective host communities during the 350 growing season from spring to fall (Figure 2A).

2015 2016

R2 P-value R2 P-value Season 0.11 0.001** 0.057 0.001** CropSeq 0.17 0.001** 0.23 0.001** Season:CropSeq 0.11 0.998 0.10 1

351 Table 1. Adonis results. R2 and P-values were calculated based on the formula, Community Distance ~ Block + Season + 352 Crop Sequence + Season:Crop Sequence. Bray-Curtis distance was calculated using Hellinger-transformed OTU counts. 353 Significant effects are reported at **P < 0.01.

354 4.3 Soil properties

355 Of the soil properties tested, only P, Mn, and Cu were found to differ significantly by crop sequence 356 (Table S7). Phosphorus decreased under continuous corn monoculture and increased under continuous 357 soybean monoculture in both 2015 and 2016, while Mn and Cu showed the reciprocal pattern. 358 Although all measured soil properties except for Zn had significant relationships with fungal 359 community dissimilarity in at least one sampling year (Table S7 and Figure 2B), only pH, Fe, and 360 Mn had R2 values greater than 0.1 according to univariate adonis (Table S7). The vector for P on the 361 NMDS plot showed an association with soybean crop sequences (Figure 2B), and P levels under 362 continuous soybean (Ss and Sr) were over twice those observed under continuous corn (Cc) (Table 363 S7). Correlations between individual soil properties and yields of corn and soybean typically had 364 opposite signs for the two crops (Table S8). While significant relationships between soil properties 365 and corn yield were inconsistent between 2015 and 2016, soybean yields were consistently 366 significantly positively correlated with organic matter (OM), manganese (Mn), copper (Cu), and total 367 organic carbon (TOC), and significantly negatively correlated with pH (Table S8).

368 bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Crop Sequence Effect

Spring2015 Mid2015 Fall2015 0.2 0.2

C3 Cn C5 0.1 0.0 C5 CropSeq 0.1 Cn C5 C3 S5 Cc C3 S1 C2 Cc C4 C2 C4 C1 C1 Ca S4 S3 Sr C4 S2 S1 Sa S2 Cc Sa C2 Sa 0.0 Ca −0.2 Cn Ca C2 Ss Ss S4 S1 S2 0.0 S5 C1 S5 C1 C3 S3 S3 S4 Sr Ss C4 R2=0.33 −0.1 Sr R2=0.35 −0.4 P=0.001 P=0.001 R2=0.26 C5 −0.1 P=0.090 Cc −0.2 −0.1 0.0 0.1 0.2 −0.4 −0.2 0.0 0.2 −0.4 −0.2 0.0 0.2 Cn S1 Spring2016 Mid2016 Fall2016

MDS2 S2 S3 0.10 Cc C4 C3 S4 Cn S1 C2 0.1 0.1 C3 S5 C5 S1 0.05 Sa C3 Cn C4 C4 Cc C2 Ss Cc Sa Ca Cn C5 C5 S1 Ca Sr S2 S3 C2 Sa 0.00 S5 0.0 C1 Ca 0.0 S2 Ca S3 S4 Sr S3 S2 S5 Ss Ss C1 Ss S4 Sa S5 R2=0.36 Sr −0.05 R2=0.36 C1 Sr R2=0.35 S4 P=0.001 P=0.001 P=0.001 −0.1 −0.3 −0.2 −0.1 0.0 0.1 0.2 −0.3 −0.2 −0.1 0.0 0.1 0.2 −0.3 −0.2 −0.1 0.0 0.1 0.2 MDS1 B Soil Properties Vectors

Spring2015 Spring2016 0.3 0.3 pH pH

0.2 0.2 CropSeq SCN Zn C3 C:N C1 S2 0.1 COM C2 C3 Cn N C5 C:N 0.1 C4 SCN C2 S3 Cc S1 C2 Sa Cn C3 S4 Cc S1 S2 C5 SaC 0.0 Ca C4 S5 C4 Ss 0.0 N OM S5 S3 MDS2 Cu S2 C5 Ss Sr Ca Zn C1 Cu Sr Cc Sr −0.1 S3 K S5 Ss K −0.1 S4 S4 C1 Cn Ca Mn S1 Sa −0.2 Mn −0.2 Fe P Fe P

−0.4 −0.2 0.0 0.2 0.4 −0.4 −0.2 0.0 0.2 MDS1 369 370 Figure 2. (A) Non-metric multidimensional scaling (NMDS) plots showing the effect of crop sequence on fungal 371 communities in each SeasonYear. R2 and P-values were derived from adonis on the formula, Bray-Curtis Distance ~ Crop 372 Sequence. (B) NMDS plots of fungal communities with environmental data vectors. Vectors indicate the weight and 373 direction of soil properties and soybean cyst nematode (SCN) density in relation to fungal communities. Crop sequences 374 include 5 consecutive years of corn after the 5th year of soybean (C1-C5); 5 consecutive years of soybean after the 5th year 375 of corn (S1-S5); soybean in annual rotation with corn (Sa); corn in annual rotation with soybean (Ca); long-term Bt corn 376 (Cc) and SCN-susceptible soybean (Ss) monocultures; non-Bt Corn (Cn) and SCN-resistant soybean (Sr) monocultures.

377 4.4 Alpha diversity

378 Chao1 estimates of fungal OTUs varied significantly by crop sequence in 2015 and by season in both 379 years (Table 2). Although there were no pairwise significant differences in Chao1 between crop 380 sequences according to Tukey's HSD test, Chao1 was significantly greater under corn monoculture 381 when compared between crop hosts (Table S9) and tended to increase under continuous corn bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

382 monoculture and decrease under continuous soybean monoculture (Figure 3A, Table S10). Numbers 383 of observed OTUs assigned to two different trophic modes by FUNGuild (Nguyen et al., 2016) 384 significantly differed by crop sequence and season (Table 2) and displayed consistent patterns in 385 relation to crop host, with saprotrophs and symbiotrophs having greater diversity under corn compared 386 to soybean, especially under long-term monoculture (Figure 3B, S9). Symbiotroph diversity also 387 showed significant negative correlations with SCN density and corn yield at one sampling timepoint, 388 each (Table S10). Richness of OTUs assigned to nematode-trapping fungi and egg parasite guilds 389 varied significantly by crop sequence in both years and by season in 2015 (Table 2). Fifth-year and 390 long-term soybean monoculture plots (S5 and Ss) and first-year corn monoculture plots (C1) typically 391 had the most trapping fungi and egg parasite OTUs, whereas long-term corn monoculture plots (Cn 392 and Cc) typically had the fewest (Figure 3C). The diversity of nematode-trapping and egg parasitic 393 guilds was significantly correlated with SCN egg density in spring of both years and negatively 394 correlated with corn monoculture year in multiple timepoints, each (Table S10).

2015 Trapping Egg ANOVA Chao1 Saprotrophs Pathotrophs Symbiotrophs fungi parasites Endoparasites Season <0.001*** <0.001*** <0.001*** < 0.001*** <0.001*** 0.02* 0.67 CropSeq 0.024* 0.0010** 0.38 <0.001*** 0.0059** <0.001*** 0.56 Season:CropSeq 0.95 0.67 0.87 0.16 0.0051** 0.94 0.70 Season Spring 824a 98a 27a 35b 3.95a 5.38b 0.55a Mid 842a 94a 24a 62a 3.58a 5.69ab 0.5a Fall 618b 72b 21b 36b 2.39b 6.09a 0.44a 2016 Trapping Egg ANOVA Chao1 Saprotrophs Pathotrophs Symbiotrophs fungi parasites Endoparasites Season <0.001*** <0.001*** <0.001*** <0.001*** 0.058 0.31 0.24 CropSeq 0.12 0.044* 0.19 <0.001*** <0.001*** <0.001*** 0.023* Season:CropSeq 0.98 0.98 0.66 0.85 0.99 0.98 0.34 Season Spring 969b 103b 26b 53b 3.69a 5.53a 0.52a Mid 1018b 106b 28b 67a 3.55a 5.69a 0.36a Fall 1161a 117a 35a 68a 4.28a 5.86a 0.5a 395 Table 2. Alpha diversity metrics across seasons and crop sequences. Chao1 is total alpha diversity of the bulk soil fungal 396 communities. Diversity metrics are also reported for individual nematophagous guilds and for three trophic modes. Only 397 OTUs assigned a trophic mode with a confidence ranking of "Probable" or "Highly Probable" were used for this analysis. 398 Significant P-values associated with ANOVA on the model Diversity ~ Block + Season + Crop Sequence + Season:Crop 399 Sequence are reported at *P < 0.05, **P < 0.01, and ***P < 0.001. Different lowercase letters indicate significant 400 differences in means according to Tukey's honestly significant difference (HSD) test. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A 2015 B 2015 2015 C 2015 2015 Chao1 Saprotrophs Symbiotrophs Trapping fungi Egg parasites 12.5 1600 P = 0.024 160 P = 0.001 P < 0.001 10.0 P = 0.006 P < 0.001 a abc a a ab a abc a ab a ab ab a a abc a a a a a 10.0 a abc ab a a abc abc bc a−d a−e a−e 7.5 ab a abc a a a 1200 a abc abc abc 100 a−d a a a 120 abc a a c abc ab ab ab ab ab b ab ab a a−e a a b−e cde ab ab ab ab 7.5 abc ab ab abc cde 5.0 b−e 800 de 80 bc c 50 e e 5.0 2.5 400 40 2.5 0 0.0 1 2 3 4 5 34 N/R A 1 2 3 4 5 34 N/R A 1 2 3 4 5 34 N/R A 1 2 3 4 5 34 N/R A 1 2 3 4 5 34 N/R A

2016 2016 2016 2016 2016

Chao1 Chao1 Saprotrophs Symbiotrophs Trapping fungi Egg parasites P = 0.115 P = 0.044 160 P < 0.001 P < 0.001 10 P < 0.001 Observed OTUs a Observed OTUs a a a a−e abc abc a a

a a ab a a a ab abc abc abc 150 a a a 9 a a a−e abc abc ab a a a a abc a−f 8 a a a 120 a−d a a−f 1500 a a a a a−f abc bc abc bc abc abc a−d abc abc cdabc a a a−f b−f a a a bc c a def 6 bcd a 120 a bc c bc 6 80 c−f d c−f ef f 1000 90 3 4 40

60 0 2 1 2 3 4 5 35 N/R A 1 2 3 4 5 35 N/R A 1 2 3 4 5 35 N/R A 1 2 3 4 5 35 N/R A 1 2 3 4 5 35 N/R A MonocultureYear MonocultureYear MonocultureYear 401 Host Corn Soybean 402 Figure 3. Alpha diversity of fungal community, trophic modes, and nematophagous guilds by crop host and monoculture 403 year. (A) Total fungal community alpha diversity, estimated by Chao1. (B) Observed OTUs classified as saprotrophs and 404 symbiotrophs according to FUNGuild. (C) Observed OTUs classified as nematode-trapping fungi or nematode egg 405 parasites. The x-axis denotes monoculture year of corn and soybean: 1-5 = years of 1-5 of continuous soybean and corn 406 monoculture; 34,35 = long-term corn and soybean monocultures of 34 and 35 years; N/R = continuous non-Bt Corn and 407 SCN-resistant soybean monocultures since 2010. A = corn-soybean annual rotation plots. Data from spring, midseason, 408 and fall within each year were pooled prior to analysis. P-values are associated with the Crop Sequence term according to 409 ANOVA on the linear model, Diversity ~ Block + Season + Crop Sequence + Season:Crop Sequence. Different letters 410 denote significant differences in means according to Tukey's honestly significant difference (HSD) test.

411 4.5 Fungal community profiles

412 Averaging across all sampling timepoints and crop sequences, a majority of reads belonged to OTUs 413 assigned to phylum Ascomycota (56%), with Basidiomycota (14%), Mortierellomycota (13%), and 414 Chytridomycota (2%) having the next most abundant read counts (Figure 4A). A significant portion 415 of reads (12%) were assigned to kingdom Fungi but not to any fungal phylum. Fewer than 1% of reads, 416 each, were assigned to Blastocladiomycota, Calcarisporiellomycota, Entomophthoromycota, 417 Entorrhizomycota, Glomeromycota, Kickxellomycota, Monoblepharomycota, Mucoromycota, 418 Neocallimastigomycota, Olpidiomycota, Rozellomycota, and Zoopagomycota.

419 Seventy-nine percent of sequence reads were clustered into OTUs that could not be assigned a 420 functional guild or trophic mode at a confidence ranking of "Probable" or higher by FUNGuild 421 (Nguyen et al., 2016). Of sequence reads belonging to OTUs that were assigned a "Probable" or 422 "Highly Probable" trophic mode, 76% were classified as Saprotrophs, 12% as Pathotrophs, 5% as 423 Symbiotrophs, 3% as Pathotroph-Saprotrophs, and 1% or less each as Pathotroph-Symbiotrophs, 424 Saprotroph-Symbiotrophs, Pathotroph-Saprotroph-Symbiotrophs, and Saprotroph-Pathotroph- 425 Symbiotrophs. (Figure 4B). A list of taxa assigned to each trophic mode is given in Table S11. A 426 majority of reads belonging to OTUs assigned to ecological guilds with "Probable" or "Highly 427 Probable" assignments were classified as Undefined Saprotrophs (66%), followed by Plant Pathogens 428 (10%), Undefined Saprotroph Wood Saprotrophs (3%), and Arbuscular Mycorrhizal fungi (3%), with 429 each other category making up 2% or less of the total. (Figure S5). A list of taxa assigned to each 430 guild is given in (Table S12). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

431 In terms of nematophagous fungal guilds, 3.7% of total reads belonged to OTUs classified as 432 nematode egg parasites, while a much smaller proportion belonged to OTUs classified as nematode- 433 trapping fungi (0.07%) and nematode endoparasites (0.01%) (Figure 4C). Four species belonging to 434 the nematode endoparasite guild, nine species belonging to the nematode egg parasite guild, and six 435 genera containing twenty-one species of nematode-trapping fungi were detected in our sequencing data 436 (Figure 5 and Table S13). Additional BLAST searches of OTUs identified as "Clonostachys" and 437 "Purpureocillium" against the UNITE database (UNITE Community, 2017) revealed that known egg- 438 parasitic taxa, Clonostachys rosea and P. lilacinum, respectively, were their most likely identities, 439 confirming that these OTUs belonged in the egg parasite guild.

440 4.6 Differential abundance by crop sequence

441 While it is impossible to ascertain true shifts in absolute abundance of taxa through metabarcoding, 442 CLRs (centered log ratios) can be used to identify shifts in relative abundance with an acceptably low 443 false discovery rate (Gloor et al., 2017). Using the CLR transformation, two main patterns were 444 observed regarding crop sequence effects on the relative abundance of fungal taxa, trophic modes, and 445 nematophagous guilds: 1) an increase in relative abundance over years of soybean monoculture and a 446 decline over years of corn (soybean-associated fungi), and 2) an increase in relative abundance over 447 years of corn monoculture and a decline over years of soybean (corn-associated fungi) (Figures 6, 7, 448 S6-8, Table S14). CLRs associated with annual rotation plots were often intermediate to those from 449 long-term monoculture plots associated with either crop host.

450 Phyla, trophic modes, and nematophagous guilds were identified that showed soybean- 451 associated or corn-associated patterns. The most consistent pattern observed at the phylum level was 452 for Glomeromycota, which was more abundant under corn monoculture in all but one sampling 453 timepoint (Figure 6A). In terms of trophic modes, symbiotrophs showed a corn-associated pattern, 454 whereas pathotrophs showed a soybean-associated pattern (Figure 6B,C). Among nematophagous 455 guilds, nematode-trapping fungi displayed the clearest pattern, with the highest relative abundance 456 associated with later years of soybean monoculture, including both susceptible (Ss) and SCN-resistant 457 (Sr) soybean plots, and with early years of corn monoculture (Figure 6D). These patterns were also 458 supported by significant spearman rank correlation tests showing significant positive correlations of 459 relative abundance of these phyla, trophic modes, or nematophagous guilds with increasing years of 460 soybean or corn monoculture at one or more sampling timepoints.

461 At the genus and species level, many fungi showed soybean-associated patterns of relative 462 abundance (Figures 7, S7, and S8). Although Glomeromycota generally showed higher relative 463 abundance under corn (Figure 6A), one AMF species, Rhizophagus irregularis, showed a soybean- 464 associated pattern in midseason of both years (Figure S7, S8). Soybean-associated taxa also included 465 several potential soybean pathogens, such as Septoria arundinacea, Fusicolla merismoides (Syn. 466 Fusarium merismoides), and Dactylonectria macrodidyma (Malapi-Wight et al., 2015) (Figures 7, S7, 467 and S8). Several fungi that have been shown to be capable of colonizing both soybean roots and SCN 468 cysts also exhibited a soybean-associated pattern. These included Paraphoma radicina (Stiles and 469 Glawe, 1989) and the soybean pathogens Corynespora cassiicola (Carris et al., 1986), Cadophora 470 gregata (Syn. Phialophora gregata) (Carris et al., 1986), and Fusarium solani (Stiles and Glawe, 471 1989). Increasing years of soybean monoculture saw greater relative abundance for non-pathogenic 472 soybean endophytes, such as Plectosphaerella sp. (Impullitti and Malvick, 2013), and 473 Pseudorobillarda sojae (Uecker and Kulik, 1986). One species of Tremellomycete yeast, Hannaella 474 oryzae, which commonly inhabits plant phyllospheres (Landell et al., 2014), showed a soybean- 475 associated pattern at four sampling timepoints (Figures 7, S7, and S8). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Fungal phyla

2015 2016 Phylum 100 Aphelidiomycota Ascomycota Basidiomycota Blastocladiomycota 75 Calcarisporiellomycota Chytridiomycota Entomophthoromycota Entorrhizomycota 50 Glomeromycota Kickxellomycota Monoblepharomycota Mortierellomycota 25 Mucoromycota Neocallimastigomycota Olpidiomycota Rozellomycota 0 Zoopagomycota Unidentified.Fungi C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa B Fungal trophic modes

2015 2016

100

Trophic mode 75 Pathotroph Pathotroph.Saprotroph Pathotroph.Saprotroph.Symbiotroph 50 Pathotroph.Symbiotroph Saprotroph Saprotroph.Pathotroph.Symbiotroph 25 Saprotroph.Symbiotroph Symbiotroph

0 mean relative abundance % abundance mean relative

C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C Nematophagous fungal guilds

2015 2016

6 4

3 Nematophagous guild 4 Nematode−trapping fungi 2 Endoparasites Egg parasites 2 1

0 0

C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa Crop Sequence 476 477 Figure 4. Fungal community profiles by crop sequence for each year. Data from each season (spring, midseason, and fall) 478 were pooled prior to plotting. Each bar represents the average proportion of (A) phyla, (B) trophic modes, or (C) 479 nematophagous guilds associated with each crop sequence in a given year. Only OTUs that were assigned a trophic mode 480 with a confidence ranking of "Probable" or higher were included in the trophic mode profile plots. Crop sequences include 481 5 consecutive years of corn after the 5th year of soybean (C1-C5); 5 consecutive years of soybean after the 5th year of corn 482 (S1-S5); soybean in annual rotation with corn (Sa); corn in annual rotation with soybean (Ca); long-term Bt corn (Cc) and 483 SCN-susceptible soybean (Ss) monocultures; non-Bt Corn (Cn) and SCN-resistant soybean (Sr) monocultures. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Nematode−trapping fungi

2015 2016 0.20 0.15

0.15 taxonomy 0.10 Arthrobotrys Dactylella 0.10 Dactylellina Hohenbuehelia 0.05 Monacrosporium 0.05 Orbilia

0.00 0.00

C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa B Nematode egg−parasitic fungi

2015 2016

taxonomy 6 4 Setophoma_terrestris Clonostachys_sp 3 4 Metacordyceps_chlamydosporia Lecanicillium_lecanii 2 Cylindrocarpon_sp Fusarium_neocosmosporiellum 2 1 Fusarium_oxysporum Fusarium_solani Purpureocillium_sp 0 0

mean relative abundance % abundance mean relative C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C Endoparasitic fungi

2015 2016

0.04 0.02 taxonomy 0.03 Haptocillium_balanoides Hirsutella_minnesotensis 0.02 0.01 Hirsutella_rhossiliensis Catenaria_anguillulae 0.01

0.00 0.00

C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa Crop Sequence 484 485 Figure 5. Nematophagous fungal guild profiles by crop sequence in each year. Data from each season (spring, midseason, 486 and fall) were pooled prior to creating profile plots. Each bar represents the average proportion of (A) nematode-trapping 487 fungal taxa, (B) nematode egg-parasitic taxa, or (C) nematode endoparasitic taxa associated with each crop sequence in a 488 given year. Crop sequences include 5 consecutive years of corn after the 5th year of soybean (C1-C5); 5 consecutive years 489 of soybean after the 5th year of corn (S1-S5); soybean in annual rotation with corn (Sa); corn in annual rotation with 490 soybean (Ca); long-term Bt corn (Cc) and SCN-susceptible soybean (Ss) monocultures; non-Bt Corn (Cn) and SCN- 491 resistant soybean (Sr) monocultures.

492 bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

493 Many nematophagous fungi showed soybean-associated patterns of relative abundance 494 (Figures 7, S7, and S8). These included the nematode-trapping fungi Orbilia auricolor, Arthrobotrys 495 scaphoides, Athrobotrys polycephala, and Arthrobotrys xiangyunensis, of which the latter two were 496 also significantly positively correlated with SCN density at one and two sampling points, respectively 497 (Figure S8). Among potential egg-parasites, Clonostachys sp. showed a strong soybean-associated 498 pattern and was also significantly correlated with SCN density at three sampling timepoints (Figures 499 7B, S8). Several species of Mortierella, including Mortierella elongata, Mortierella alpina, and 500 Mortierella polygonia, some isolates of which have been shown to inhibit SCN egg hatch (Juba et al., 501 2004), showed soybean-associated patterns and significant correlations with SCN density (Figures S7, 502 S8). Didymella americana (syn. Peyronellaea americana), which has been isolated from nematode 503 cysts (Aveskamp et al., 2010), was positively correlated with soybean monoculture year in midseason 504 2016, while an OTU less precisely identified as Didymellaceae sp. showed a soybean-associated 505 pattern and significant correlations with SCN density at multiple timepoints (Figure S8). One 506 basidiomycete species, Coprinellus micaceus (syn. Coprinus micaceus) was positively correlated with 507 SCN density in fall 2016 and with soybean monoculture year in spring 2016 (Figures 7B and S8). 508 Some members of Coprinellus are known to produce nematicidal compounds either as a defense 509 against herbivory or, for nematophagous species, as part of a feeding strategy (Degenkolb and 510 Vilcinskas, 2016; Plaza et al., 2016)

511 A contrasting group of fungi showed a corn-associated pattern (Figures 7, S7, and S8). Among 512 symbiotrophic taxa, these included AMF identified as Paraglomus brasilianum and Glomus indicum, 513 as well as the dark septate endophyte Periconia macrospinosa (Doran et al., 1984). A corn-associated 514 pattern was also observed for OTUs identified as members of the family, Serendipitaceae, which is in 515 the order Sebacinales (Figures 7, S7, and S8). Increasing years of corn monoculture saw greater 516 relative abundance of several saprotrophic taxa, including Helicosporium pallidum, Cyphellophora sp. 517 (Tedersoo et al., 2014), Cladophialophora chaetospira, Exophiala equina, Talaromyces verruculosus, 518 Coniochaeta ligniaria (Jiménez et al., 2017), Collarina aurantiaca, Myrmecridium phragmitis, 519 Humicola grisea, Myrmecridium schulzeri (Tedersoo et al., 2014), and Apodus sp. (Tedersoo et al., 520 2014). The plant pathogens Cladosporium, Plenodomus biglobosus (Marin-Felix et al., 2017), 521 Parastagonospora avenae, Parastagonospora nodorum, Phaeosphaeriopsis sp., Setophoma terrestris, 522 Drechslera sp., Ceratocystis sp., and Phaeosphaeria sp. all showed corn-associated patterns, as did 523 one Tremellomycete yeast, a member of Cystofilobasidiales (Fell et al., 1999).

524 No significant correlations were found between the relative abundance of specific taxa and 525 soybean yield at any sampling timepoint, but a few taxa showed significant correlations with corn yield 526 in spring and midseason 2016 (Figure S8). In general, taxa that correlated positively with corn yield 527 had greater relative abundance under continuous soybean monoculture. One of these taxa was 528 classified as Glomus aggregatum, an AMF associated with soybean monoculture in midseason 2016, 529 and another was identified as Mortierella sp. (Figure S7 and S8). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Glomeromycota relative abundance by crop sequence

2015 2016 4

3 2.5 2.0 2 1.5 1 P = 0.03 P = 6e−04 P = 1e−08 P = 0.07 P = 0.02 1.0 0 P = 0.02

B Symbiotroph relative abundance by crop sequence

2015 2016

2.0 1.5 1.5 1.0 1.0 0.5 P = 0.066 0.5 P = 0.004 0.0 P = 0.034 P = 0.046 P = 0.090 P = 0.243 Season −0.5 0.0 Spring C Pathotroph relative abundance by crop sequence Mid 2015 2016 Fall P = 0.066 3.0 P = 0.004 4 P = 0.150 P = 0.003 P = 0.314 2.5 P = 0.042 Centered log ratio (CLR) Centered log ratio 3 2.0

2 1.5

1.0 1

D Nematode−trapping fungi relative abundance by crop sequence

2015 2016

300 P = 3e−04 P = 0.003 P = 0.175 150 P = 0.001 P = 0.852 P = 0.311 200 100

100 50

0 0

C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa 530 Crop Sequence 531 Figure 6. Differential abundance of (A) Glomeromycota, (B) symbiotrophs, (C) pathotrophs, and (D) nematode-trapping 532 fungi across crop sequences. The phylum, trophic modes, and nematophagous guild shown all had centered log ratios 533 (CLRs) that differed significantly by crop sequence according to ANOVA or the Kruskal-Wallis test (P < 0.05 after false 534 discovery rate (FDR) correction) in multiple sampling timepoints. FDR-corrected P-values associated with differential 535 abundance tests are shown on each plot. For consistency, data from all timepoints are shown, even if differences were non- 536 significant. Error bars show one standard error above and below the mean. Crop sequences include 5 consecutive years of 537 corn after the 5th year of soybean (C1-C5); 5 consecutive years of soybean after the 5th year of corn (S1-S5); soybean in 538 annual rotation with corn (Sa); corn in annual rotation with soybean (Ca); long-term Bt corn (Cc) and SCN-susceptible 539 soybean (Ss) monocultures; non-Bt Corn (Cn) and SCN-resistant soybean (Sr) monocultures. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Fall 2016 B Fall 2016 Clonostachys_sp* Volutella_ciliata Neonectria_sp Neonectria_sp Ramularia_sp Clonostachys_sp* Plectosphaerella_sp Lasiodiplodia_sp Lectera_sp Septoria_arundinacea Lectera_sp Path Plectosphaerella_sp Path Septoria_arundinacea Bipolaris_sp Chytridiomycota_sp Septoria_sp Triparticalcar_sp Metarhizium_rileyi Bipolaris_papendorfii Plectosphaerella_cucumerina Exserohilum_turcicum Drechslera_sp Epicoccum_sorghinum Coniochaeta_sp Drechslera_sp Path Coniochaetaceae_sp Sap Cladosporium_sp Sym Hannaella_oryzae Path Hannaella_oryzae Cladosporium_sp Sap Kabatiella_zeae Sym Dioszegia_zsoltii Fusarium_solani* Coniochaeta_sp Pluteus_podospileus Fusarium_solani* Path−Sap Didymellaceae_sp Fusarium_sp Path−Sap Didymellaceae_sp Nectriaceae_sp Ceratocystis_sp Ceratocystis_sp Colletotrichum_sp Colletotrichum_graminicola Path−Sym Path−Sym Colletotrichum_gloeosporioides Colletotrichum_graminicola Corynespora_cassiicola Ogataea_sp Tetracladium_sp Tetracladium_sp Dactylonectria_macrodidyma Arthrobotrys_polycephala* Pseudombrophila_hepatica Conocybe_vinaceobrunnea Helicosporium_sp Thielaviopsis_basicola Tritirachium_oryzae Ogataea_sp Arthrobotrys_xiangyunensis* Fusicolla_merismoides Plenodomus_biglobosus rho Myrmecridium_schulzeri Exophiala_equina Saccharomycetales_sp mean CLR Collarina_aurantiaca 0.5 Corticium_sp 6 Corynespora_cassiicola 0.0 Exophiala_alcalophila Dactylonectria_macrodidyma 4 −0.5 Sap Nigrospora_oryzae Sap Fusicolla_merismoides Collarina_aurantiaca 2 Plenodomus_biglobosus Helicosporium_sp Skeletocutis_chrysella 0 Coprinellus_micaceus Exophiala_equina Clitopilus_sp Humicola_grisea Clavaria_sp Keissleriella_yonaguniensis Trichoderma_koningiopsis Trichoderma_tomentosum Candida_haemulonis Conlarium_sp Exophiala_opportunistica Kickxellomycota_sp Exophiala_sp Phaeosphaeriopsis_sp Cladophialophora_chaetospira Exophiala_opportunistica Talaromyces_verruculosus Exophiala_sp Phaeosphaeriopsis_sp Cladophialophora_chaetospira Sistotrema_sp Sap−Sym Cantharellales_sp Sap−Sym Cantharellales_sp Periconia_macrospinosa Cadophora_gregata Periconia_macrospinosa Cadophora_gregata Phylloporus_sp GS24_sp Sym Glomus_indicum Sym Paraglomus_sp Ceratobasidium_sp Paraglomus_brasilianum Archaeosporaceae_sp Paraglomerales_sp Paraglomerales_sp Glomeraceae_sp Lasiosphaeriaceae_sp Paraphoma_radicina Conioscypha_sp Sporidiobolus_pararoseus Cystofilobasidiales_sp Pseudorobillarda_sojae Setophoma_terrestris* Ophiostomataceae_sp Pseudorobillarda_sojae Unident Pseudotrichia_mutabilis Ophiostomataceae_sp Symmetrospora_symmetrica Paraphoma_radicina Solicoccozyma_terrea Unident Lasiosphaeria_sp Symmetrospora_symmetrica Neocallimastigomycota_sp Neocallimastigomycota_sp Cystofilobasidiales_sp Serendipitaceae_sp Setophoma_terrestris* Phaeosphaeria_sp Conioscypha_sp Serendipitaceae_sp Corn_Year Soybean_Year 540 C1 C2 C3 C4 C5 Cc Cn S1 S2 S3 S4 S5 Ss Sr Ca Sa FallSCNDensity 541 Figure 7. Differential abundance of fungal taxa by crop sequence in fall 2016. (A) Centered log ratios (CLRs) of fungal 542 taxa that differed significantly by crop sequence according to ANOVA or the Kruskal-Wallis test (P < 0.05 after false 543 discovery rate (FDR) correction). (B) Taxa that significantly correlated with SCN density or monoculture year of corn or 544 soybean, according to Spearman's rank correlation test. Trophic modes assigned to each taxon are given in the left-hand 545 columns. Path = pathotroph; Sap = saprotroph; Sym = symbiotroph; Unident = unidentified trophic mode. Within each 546 trophic mode category, taxa are ordered from top to bottom by strength of correlation with soybean monoculture year. 547 Asterisks denote taxa belonging to nematophagous guilds, as defined in this study. Crop sequences include 5 consecutive 548 years of corn after the 5th year of soybean (C1-C5); 5 consecutive years of soybean after the 5th year of corn (S1-S5); 549 soybean in annual rotation with corn (Sa); corn in annual rotation with soybean (Ca); long-term Bt corn (Cc) and SCN- 550 susceptible soybean (Ss) monocultures; non-Bt Corn (Cn) and SCN-resistant soybean (Sr) monocultures.

551 bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

552 4.7 Differential abundance by season

553 Changes in relative abundance by season were generally more significant than by crop sequence, most 554 likely due to the fact that data from multiple crop sequences were combined for these analyses, resulting 555 in a larger sample size. The number of phylum-level discoveries ranged from 6 fungal phyla in 2016 556 soybean plots to 13 fungal phyla in 2016 corn plots (Figure S9). The number of species-level 557 discoveries ranged from 77 taxa in 2016 soybean plots to 337 taxa in 2016 corn plots (Table S15). 558 Unlike differential abundance by crop sequence, in which only two main patterns were evident, taxa, 559 trophic modes, and nematophagous guilds exhibited a variety of seasonal shifts and showed similar 560 seasonal patterns regardless of crop host (Figure S9-S11).

561 There was some consistency between shifts in relative abundance of taxa by crop sequence and 562 shifts observed across seasons. For example, several species of corn-associated AMF, as well as the 563 plant pathogen, S. terrestris, which increased in relative abundance over continuous corn monoculture, 564 also increased over the growing season under corn, having greater relative abundance in midseason or 565 fall compared to spring (Table S15). Soybean-associated taxa, including many potentially 566 nematophagous fungi, like Plectosphaerella cucumerina, Clonostachys sp., D. americana, Fusarium 567 oxysporum, F. solani, M. chlamydosporia, and several species of Mortierella (Hu et al., 2018), had 568 greater relative abundance at either midseason or fall under soybean in 2015, and several others, 569 including A. scaphoides and Coprinopsis candidolanata, decreased over the corn growing season in 570 2016 (Table S15). The soybean-associated AMF, R. irregularis, increased from spring to midseason 571 under soybean in 2015 (Table S15). Similarly, the soybean endophyte, P. sojae, and the soybean 572 pathogen, C. gregata, increased in relative abundance throughout the soybean growing season in 2015 573 and 2016 (Table S15). Other soybean-associated plant pathogens, like S. arundinacea, which 574 consistently decreased under continuous corn monoculture, also decreased throughout a single corn 575 growing season in 2016 (Table S15).

576 4.8 Correspondence between fungal taxa, yield, SCN density, and soil properties

577 Distance-based redundancy analysis showed that a combination of crop yield, SCN density, and soil 578 properties explained, on average, 57% and 52% of variation between fungal communities at the ordinal 579 level in corn plots and soybean plots, respectively. Corn yield and soil P varied with fungal 580 communities in the same direction in 2015 and 2016 and corresponded with the relative abundance of 581 Mortierellales, an order containing phosphate-solubilizing fungi (Li et al., 2018) (Figure 8A,B). A 582 spearman correlation test revealed that the relative abundance of this order was positively correlated 583 with soil P in 2015 (Table 3). While not immediately apparent on the redundancy analysis plots, the 584 relative abundance of Sebacinales, an order containing ectomycorrhizal and endophytic species (Weiß 585 et al., 2011, 2016), and two glomeromycete orders, Glomerales and Paraglomerales, were significantly 586 negatively correlated with soil P (Table 3). SCN density and yield corresponded with soybean- 587 associated fungal communities in opposite directions in 2016 (Figure 8C,D). Hypocreales, an order 588 containing many nematophagous and entomopathogenic fungi (Castillo Lopez et al., 2014; Ghayedi 589 and Abdollahi, 2013; Manzanilla-Lopez, Rosa H; Esteves, Ivania; Finetti-Sialer, Mariella; Hirsch, 590 Penny; Ward, Elaine, Devonshire, Jean; and Hidalgo-Diaz, 2013; Stirling, 2014), corresponded with 591 SCN density in both corn and soybean distance-based redundancy analysis plots (Figure 8). bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Corn Spring 2015 Corn Spring 2016

0.8 Yield 0.6

OM

N Cantharell pH Mn SordariC Agaric 0.4 N 0.2 Cu K Cu Chaetothyri MortierellPOMC Coniochaet SebacinZn Fe Hypocre Glomer Cystofilobasidi SCN Cystofilobasidi Saccharomycet CN SCN Heloti Filobasidi Agaric Mortierell Hypocre Mn Coniochaet

0.0 Glomer

− 0.2 SordariK Zn CAP2 (7.67%) Fe CAP2 (11.3%) Peziz P CN pH

− 0.6 Yield − 0.4 Cantharell

−0.5 0.0 0.5 −0.5 0.0 0.5

CAP1 (20.6%) CAP1 (17.8%)

Soybean Spring 2015 Soybean Spring 2016

Cantharell 0.6 CN 0.4 SCN

0.4 pH K Cu Agaric Chaetothyri Cystofilobasidi Hypocre SCN Cu ParaglomerZn CN Capnodi Fe K Zn 0.0 Mortierell

0.2 pH Tremell C Mn N OM Coniochaet Capnodi Chaetothyri Yield YieldGlomer 0.0 Peziz Mortierell Cystofilobasidi Agaric Fe P − 0.4

CAP2 (5.76%) Hypocre Sordari Mn CAP2 (8.73%) OM P C N − 0.8 − 0.4

−0.6 −0.2 0.2 0.4 0.6 −0.5 0.0 0.5

CAP1 (17.3%) CAP1 (16.7%) 592 593 Figure 8. Distance-based redundancy analysis (dbRDA) of fungal communities for (A) corn plots in 2015, (B) corn plots 594 in 2016, (C) soybean plots in 2015, and (D) soybean plots in 2016. Relative position of fungal orders in the biplots is based 595 on Bray-Curtis dissimilarity of square root transformed relative abundance at the ordinal level. Vectors indicate the weight 596 and direction of soil properties, soybean cyst nematode (SCN) density, and crop yields associated with fungal communities. 597 The dbRDA axes describe the percentage of the variation explained by each axis while being constrained to account for 598 differences in soil properties, SCN density, and yield. Order names were shortened by removing "ales" from the end of 599 each name. Orders that mapped near the plot center or that overlapped, including Glomerales, Paraglomerales, and 600 Sebacinales on some plots, were omitted to aid readability. C = carbon; Cu = copper; Fe = iron; K = potassium; Mn = 601 manganese; N = nitrogen; OM = organic matter; P = phosphorus; CN = carbon:nitrogen; SCN = soybean cyst nematode.

602 bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Order 2015 2016 Glomerales -0.28* -0.54*** Paraglomerales -0.37** -0.27* Sebacinales -0.44*** -0.38** Mortierellales 0.40** 0.12 603 Table 3. Spearman correlations (rho) between centered 604 log ratios (CLRs) of four fungal orders and soil 605 phosphorus in spring of 2015 and 2016. Significant 606 relationships are reported at FDR-corrected P-values of 607 *P < 0.05, **P < 0.01, and ***P < 0.001.

608 5. DISCUSSION

609 Utilization of a long-term research site at which crops have been planted in continuous rotations and 610 monoculture since 1982 allowed us to gain unique insights into the effects of crop sequences on soil 611 fungi. Shifts in the composition of fungal communities correlated with years of continuous 612 monoculture of soybean or corn, providing further evidence for the role of plant hosts in shaping soil 613 microbial communities (Berg et al., 2005; Kuske et al., 2002; Peiffer et al., 2013). Overall soil fungal 614 alpha diversity was higher under corn monoculture, but the diversity of specific functional guilds 615 showed contrasting patterns, with saprotrophs and symbiotrophs increasing in diversity under 616 continuous corn and nematode-trapping and egg parasitic fungi increasing in diversity under 617 continuous soybean. Several potential predators and opportunistic pathogens of SCN were identified 618 whose relative abundance was positively correlated with SCN density and with years of continuous 619 soybean monoculture. These predominantly included members of nematode-trapping and egg- 620 parasitic guilds but also included fungi that may immobilize or kill nematodes through the production 621 of nematicidal compounds. Soil chemical properties, most notably P, shifted over continuous 622 monoculture of both crop hosts and corresponded with shifts in fungal taxa, including several fungal 623 orders involved in P solubilization and acquisition (Mortierellales, Glomerales, Paraglomerales, 624 Sebacinales). Changes in soil properties, SCN density, and the relative abundance of specific fungal 625 groups, like AMF and plant-pathogenic fungi, suggest multiple causes of yield declines under 626 continuous monoculture of soybean and corn.

627 5.1 Yield and soybean cyst nematode density

628 The results of corn yield (Figure 1A) were similar to those of a metanalysis of hundreds of thousands 629 of fields in the United States (Seifert et al., 2017) and to a previous study at our site in which a yield 630 penalty was observed during the second year of corn monoculture, with insignificant yield penalties 631 with additional years of corn (Grabau and Chen, 2016b). Decreased nitrogen availability has been 632 identified as the main limiting factor in corn yields under continuous corn cropping (Gentry et al., 633 2013). However, in our study total N actually increased under continuous corn monoculture (Table 634 S7), most likely due to N fertilization with urea in the corn plots. The most significant change in any 635 soil property was seen for P, which decreased under corn monoculture and increased under soybean 636 monoculture, a result similar to one observed in a previous study at the same site (Johnson et al., 1991). 637 While a significant correlation between soil P and corn yield was not detected (Table S8), corn yield 638 and soil P corresponded with fungal communities in the same direction (Figure 8A,B), raising the 639 possibility that corn yield was in fact limited, at least in part, by available P. One notable difference 640 between our results and those of Grabau and Chen (2016b) was that non-Bt corn yields were lower in 641 2015 and higher in 2016 compared to long-term continuous Bt-corn (Figure 1A). One possible 642 explanation for this observation is that pressure from plant parasitic nematodes or insect pests was bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

643 higher in 2015, allowing Bt corn to out-perform non-Bt corn (Grabau and Chen, 2016b). However, 644 this hypothesis cannot be tested, because measurements of plant-parasitic nematodes or insect pests 645 other than SCN were not taken for this study.

646 Soybean yields in our study (Figure 1B) were somewhat less consistent with previous reports. 647 In the same metanalysis of corn and soybean fields in the United States, soybean yields declined each 648 year for up to five years of consecutive monoculture, without the yield penalty leveling off as for corn 649 (Seifert et al., 2017). A similar pattern was observed in a previous study at our site, with the lowest 650 soybean yields observed for long-term SCN-susceptible soybean monoculture (Ss) (Grabau and Chen, 651 2016a). By contrast, our results showed relatively stable yields in years 1 through 5 of continuous 652 soybean monoculture (Figure 1B). Our findings were consistent, however, with Grabau and Chen's 653 (2016a) study in that SCN-resistant soybean (Sr) had higher yields than SCN-susceptible soybean in 654 long-term monoculture, especially in 2015 (Figure 1B), the year in which SCN density was greater 655 (Figure 1C). Interestingly, this is the same year in which Bt corn outperformed non-Bt corn, which 656 could suggest that weather conditions or some other factor associated with this growing season 657 encouraged reproduction of plant parasitic nematodes or insect pests affecting both soybean and corn. 658 SCN has been proposed to be the single most limiting factor for soybean yield (Seifert et al., 2017) and 659 has previously been shown to be negatively correlated with soybean yields at this site (Grabau and 660 Chen, 2016b). Our results showing significant negative correlations between SCN egg density and 661 susceptible soybean yields (Figure 1D) corroborate these findings.

662 Overall, the pattern of SCN density observed in this experiment (Figure 1C) supports the 663 practice of multi-year rotations with corn to manage SCN populations. SCN levels in soybean annual 664 rotation plots were comparable to those in long-term soybean monoculture and only declined 665 significantly after four or five years of corn monoculture in 2015 and 2016, respectively (Figure 1C), 666 a more extreme result than an earlier report at the same site that showed significant declines in SCN 667 density within the first year of corn (Porter et al., 2001). SCN may become more problematic with a 668 warming climate that favors longer growing seasons, resulting in more generations of this polycyclic 669 pathogen per year (Boland et al., 2004; Chen, 2011). The practice of annual rotation, which resulted 670 in a significant improvement in corn yields (Figure 1A), may no longer be optimal for improving 671 yields of soybean due to its failure to regulate SCN. Alternatively, it is possible for long-term soybean 672 monoculture to lead to the build-up of microbes that antagonize SCN, resulting in SCN-suppressive 673 soil (Chen, 2007). While our study showed that this build-up occurs, it cannot provide evidence that 674 soil fungi helped to control SCN densities, due to the nature of the study design. However, a nearby 675 field at the same experimental site with a similar long-term soybean monoculture history showed 676 evidence of biological suppression of SCN (Hu et al., 2017), and it is possible that such suppression is 677 also occurring in the long-term soybean monoculture plots used in our study.

678 5.2 Fungal communities shaped by long-term monoculture

679 Analysis of fungal community structure revealed that with increased years of monoculture, soil fungal 680 communities become more similar to those of the long-term monoculture plots associated with their 681 respective crop hosts (Figure 2A). This finding is in agreement with other studies showing this same 682 phenomenon for other crops, including coffee (Zhao et al., 2018) vanilla (Xiong et al., 2014), 683 Pseudostellaria heterophylla (Wu et al., 2016), and soybean (Bai et al., 2015). These shifts in bulk 684 soil communities may be related to shifts in soil properties that favor the proliferation of certain groups 685 of fungi (Lauber et al., 2008), or they could be due to the "rhizosphere effect" in which certain groups 686 of microbes proliferate in the soils associated with plant roots (Hiltner, 1904; Kristin and Miranda, 687 2013; Lapsansky et al., 2016). Our research suggests that many factors, including soil properties, plant bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

688 hosts, and even plant-parasitic nematodes like the SCN help shape the fungal community over 689 continuous monoculture. These results add to the growing body of evidence that long-term 690 monoculture has a dramatic impact on soil microbial communities.

691 5.3 Alpha diversity affected by crop sequence

692 Plant species richness is a predictor of soil fungal richness (Maltz et al., 2008; Tedersoo et al., 693 2014), and a greater variety of root exudates can support a more diverse fungal community (Broeckling 694 et al., 2008). We expected, therefore, that annual rotation (Ca and Sa) would result in increased soil 695 fungal alpha diversity. However, total alpha diversity was more strongly associated with corn 696 monoculture than with crop rotation (Figure 3A and Table S9). This pattern was driven, in part, by 697 the diversity of symbiotrophic and saprotrophic fungi (Figure 3B). Our result showing increased 698 diversity of symbiotrophs under corn monoculture is consistent with an earlier study at the same site 699 documenting higher diversity of AMF under corn monoculture compared to soybean (Johnson et al., 700 1991). This result may imply that corn has fewer selective barriers to AMF colonization than soybean. 701 Alternatively, a more diverse AMF community may be the result of greater below-ground biomass in 702 corn (Johnson et al., 1991). A larger root system may support more AMF, overall, resulting in greater 703 detection of diverse AMF in our metabarcoding analysis. The increased diversity of saprotrophs could 704 also be related to biomass, as a greater amount of crop residue, which promotes the proliferation of 705 diverse saprotrophic fungi (Ma et al., 2013), remains after corn harvest compared to soybean harvest 706 (Doran et al., 1984). Our data showing both greater diversity of saprotrophic fungi (Figure 3B and 707 Table S9) and greater amounts of organic matter (Table S7) under corn monoculture support this 708 hypothesis.

709 While overall alpha diversity was higher under corn monoculture (Table S9), nematode- 710 trapping and egg-parasitic nematophagous guilds showed increased alpha-diversity under soybean 711 monoculture, correlating with higher egg densities of SCN (Figure 3C, Tables S9 and S10). This 712 finding is consistent with other ecological studies that show an increase in bacterial and fungal diversity 713 in response to increased availability of limited resources (Cline et al., 2018; She et al., 2018). Further 714 research is needed to understand whether these fungi contribute to controlling SCN populations or if 715 their merely act as opportunistic SCN parasites without substantially reducing SCN density.

716 Our findings regarding total alpha diversity (Figure 3A and Table S9) suggest another 717 potential mechanism through which crop rotation benefits soybean. General suppression of soil-borne 718 plant pathogens relies on a diverse fungal community that limits the establishment of pathogenic 719 microbes in the soil (Garbeva et al., 2011). It is possible that corn monoculture helps create such a 720 community by increasing the taxonomic and functional diversity of soil fungi, a pattern that may also 721 contribute to soil fertility (Mäder et al., 2002). It is also possible that increased AMF diversity under 722 corn monoculture contributes, in part, to higher soybean yields following years of corn (Johnson et al., 723 1991, 1992).

724 Seasonal variation, including an increase in bulk soil fungal alpha diversity (Chao1) from 725 midseason to fall in 2016 (Table 2), was consistent with the pattern observed for alpha diversity of 726 fungal communities in SCN cysts in 2015 and 2016 at the same study site (Hu et al., 2018). It is 727 possible that these related patterns reflect an ecological relationship between bulk soil and SCN cyst 728 fungal communities, given that soil may act as a reservoir for microbial taxa (Mendes et al., 2014) from 729 which certain groups of microbes can be selectively filtered by niche environments (Kristin and 730 Miranda, 2013), like SCN cysts. Unexpectedly, the opposite pattern was observed in 2015, in which 731 alpha diversity in the fall was significantly lower than midseason (Table 2). It is possible that the bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

732 lower alpha diversity in fall 2015 was due to abnormal weather or environmental conditions that 733 introduced a sampling bias. The fall of 2015 was one of the hottest on record, with temperatures above 734 80o F, no precipitation the week prior to sampling, and significantly higher solar radiation compared 735 to fall 2016 (Regents of the University of Minnesota, 2018). It is possible that some fungi or fungal 736 DNA that would otherwise have been present in the soil was destroyed by excessive heat, dryness, and 737 solar radiation at this sampling timepoint.

738 5.4 Arbuscular mycorrhizal fungi

739 Earlier studies at our experimental site found that unique communities of AMF proliferated under 740 continuous corn and soybean monoculture and that the proliferation of specific AMF taxa was 741 correlated with yield declines in their respective crop hosts (Johnson et al., 1991, 1992). Johnson 742 (1992) hypothesized that one reason for these yield declines was that AMF that proliferated under 743 monoculture were less mutualistic than AMF sustained by crop rotation. While it is probable that our 744 metabarcoding approach failed to capture the full diversity of AMF (Kohout et al., 2014; Stockinger et 745 al., 2010), our results nonetheless supported the idea that corn and soybean host unique AMF 746 communities (Figures 7, S7, and S8). In further support of Johnson's (1992) hypothesis is the 747 observation that OTUs identified as Glomeraceae sp., which proliferated under corn, were negatively 748 correlated with corn yield at one timepoint, whereas OTUs identified as Glomus aggregatum, which 749 showed a soybean-associated pattern, were positively correlated with corn yield (Figure S8). These 750 findings support the idea that soybean-associated AMF are more mutualistic on corn than are corn- 751 associated AMF.

752 However, it may be possible for AMF to be correlated with corn yield declines without being 753 the cause of these declines. Our results suggest an alternative hypothesis that specific AMF proliferated 754 under corn monoculture as an adaptation to lower soil P (Table S7). In this model, both AMF and soil 755 P are correlated with yield declines in corn. However, it is possible that yield declines would be more 756 severe without AMF, which help the plant take up the limited soil P. This model is in agreement with 757 studies showing that corn yields and uptake of P are higher if the preceding crop is corn rather than a 758 non-mycorrhizal crop, such as canola (Miller, 2000). In order to address these contrasting hypotheses, 759 further studies testing the efficiency with which soybean and corn-associated AMF uptake and 760 translocate P and affect the yield both crop hosts are warranted.

761 5.5 Nematophagous fungi

762 Of the three nematophagous guilds that were examined, nematode-trapping fungi had the 763 strongest relationships with crop sequence and SCN density (Figures 3C, 6D and S6). A pioneering 764 experiment on fungal-nematode interactions showed that nematode-trapping fungi build up in soil with 765 a large quantity of free-living nematodes and may provide a form of biological control against plant 766 parasitic nematodes under such conditions (Linford, 1937). Subsequent studies failed to corroborate 767 this conclusion (Stirling, 2014). However, a density-dependent host-parasite relationship, which is an 768 important trait for biocontrol agents of plant-parasitic nematodes (Jaffee et al., 1992), has been 769 observed experimentally for several species of trapping fungi, including D. ellipsospora (syn. M. 770 ellipsosporum), A. dactyloides (Syn: D. dactyloides) (Jaffee et al., 1993), and Dactylellina haptotyla 771 (Jaffee, 2003). Trapping fungi that produce adhesive knobs (e.g. D. ellipsospora) or constricting rings 772 (e.g. A. dactyloides) have been shown to respond to an increase in nematodes as a food source more 773 than trapping fungi that produced adhesive networks (e.g. Arthrobotrys oligospora) (Jaffee, 2003). 774 However, our findings did not corroborate these results. Instead, we observed increased relative 775 abundance of adhesive-network forming fungi, including Arthrobotrys superba, A. polycephala (Yu et bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

776 al., 2014), and A. xiangyunensis (Liu et al., 2014a) under consecutive years of soybean monoculture 777 (Figures 7, S7, and S8). Of the trapping fungi we detected, only A. polycephala and A. xiangyunensis 778 showed significant positive correlations with SCN egg density (Figures 7B and S8).

779 One of the more interesting findings was that the relative abundance of nematode-trapping 780 fungi in SCN-resistant soybean plots was nearly as great as in long-term susceptible soybean 781 monoculture plots (Figure 6D). One possible explanation for this observation is that the plant host, 782 rather than SCN density, is responsible for the build-up of trapping fungi in the bulk soil (Bordallo et 783 al., 2002). However, another possible explanation is that the abundance of trapping fungi DNA in bulk 784 soils of SCN-resistant soybean plots is due to a soil memory effect (Lapsansky et al., 2016). From 785 1982 until 2010, these plots were planted with SCN-susceptible soybean and therefore likely had a 786 sustained build-up of SCN cysts and, possibly, predators of SCN, like nematode-trapping fungi. These 787 fungi could remain active as soil saprobes or merely be present as large quantities of spores even after 788 a decline in SCN cyst density.

789 The nematode egg parasite guild is difficult to define for several reasons. Dozens of species of 790 fungi have been isolated from SCN cysts (Chen and Chen, 2002), but far fewer have been tested for 791 their ability to parasitize eggs in vitro. It is, therefore, possible that many of these fungi grow 792 saprotrophically on dead nematode eggs and are not true parasites that infect living eggs. For this 793 study, we chose to define the egg parasite guild as fungal species that have been isolated from nematode 794 cysts and been shown to parasitize nematode eggs in an in vitro bioassay. Thus, we have taken a 795 conservative approach, and it is possible that this guild includes far more taxa than we included. In 796 addition to the difficulty in defining the egg parasite guild, it is also difficult to draw conclusions about 797 the reasons for changes in relative abundance or alpha diversity of this guild over various crop 798 sequences (Figure 3C and Table S14). Nematode egg parasites often occupy multiple ecological 799 niches. For example, isolates of F. solani can act both as plant pathogens (Rupe et al., 1997) and soil 800 saprotrophs (O’Donnell et al., 2008), and some isolates have been shown to parasitize SCN eggs (Chen 801 and Chen, 2003). It is possible that egg parasite diversity increases under soybean monoculture 802 (Figure 3C) because the plant host, rather than the SCN, is hosting these pathogens. Nonetheless, it 803 is intriguing that many fungi that have been isolated from SCN cysts showed positive correlations with 804 SCN density. For example, the well-characterized nematode egg parasite P. lilacinum (Song et al., 805 2016b) was below the limits of detection in many corn monoculture plots but was consistently detected 806 in long-term susceptible soybean monoculture plots (Ss) (Figure 5B). Likewise, potential SCN egg 807 parasites C. cassiicola, P. radicina, Clonostachys sp., and D. macrodidyma all showed consistently 808 positive correlations with SCN density across multiple timepoints (Figures 7B and S8). Interestingly, 809 two Mortierella species, M. alpina and M. elongata, which were shown to have strong correlations 810 with SCN density when quantified from SCN cysts in our same study system (Hu et al., 2018), also 811 showed significant correlations with SCN density in bulk soil (Figure S8). Some Mortierella species 812 isolated from SCN cysts have demonstrated a relatively high egg parasitism index, a measure of the 813 percentage of eggs colonized within a cyst (Chen and Chen, 2002). Collectively, these findings suggest 814 that some Mortierella species may be parasitic on SCN. However, an alternative hypothesis supported 815 by our results is that the soybean plant promotes the proliferation of Mortierella species to improve P 816 acquisition (Table S9). Future research testing both of these hypotheses is warranted.

817 Endoparasites of SCN, such as H. rhossiliensis and Hirsutella minnesotensis, are commonly 818 isolated from SCN J2s in soybean fields in Minnesota (Liu and Chen, 2000), and the percentage of J2s 819 parasitized by H. rhossiliensis has been shown to be affected by crop sequence in a culture-based study 820 at our experimental site (Chen and Reese, 1999). However, our study failed to detect a clear association 821 between this nematophagous guild and crop sequence or SCN density. This may be due to the fact that bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

822 these fungi make up an extremely small proportion of the entire bulk soil fungal community (Figures 823 4C and 5C), causing their DNA to be below the limits of detection in most bulk soil samples.

824 5.6 Relationships between soil properties, crop sequences, and fungal communities

825 Even though a majority of soil properties did not differ significantly by crop sequence, the 826 general trend showed that most soil properties increased either under soybean or corn monoculture, 827 with OM, Fe, Mn, Cu, total N, and TOC associated with corn and pH and P associated with soybean 828 (Table S7). Many of these changes are likely associated with inputs from fertilizer or from the crops, 829 themselves. Corn, a plant with greater overall biomass than soybean, contributes more organic matter 830 and carbon to the soil (Karlen et al., 1994). While soybean was expected to be associated with higher 831 levels of N due to the effect of nitrogen-fixing rhizobia (Cooper, 2007), the urea fertilizer used in corn 832 plots resulted in greater total N under corn (Table S7). Shifts in pH across crop sequences may also 833 be related to urea fertilizer, which makes soil more acidic (Lungu and Dynoodt, 2008).

834 However, no phosphorus-containing fertilizer was applied to the plots in this study, so a 835 biological cause for differences in P levels is likely. The decline in P under corn monoculture may be 836 related to the associated increase in relative abundance of Glomeromycota and Serendipitaceae sp. 837 (Figures 6, 7, and S6-S8). AMF substantially increase the ability of plants to acquire P, but in doing 838 so deplete P from surrounding soil (Zhang et al., 2010). When a portion of the corn plant is removed 839 at harvest, this P is not replaced. Sebacinales, the order containing Serendipitaceae sp., appeared to 840 display the same pattern as the glomeromycete orders (Figure 8 and Table S9), a result that may be 841 reflective of a similar ecological function. One well-studied member of Sebacinales, Serendipita 842 indica (syn. Piriformospora indica), is associated with increased plant levels of N, K, and P (Kumar et 843 al., 2012) and has been shown to actively transport P to corn roots (Yadav et al., 2010). Collectively, 844 these findings suggest that the corn-associated fungal community may help corn to adapt to a low-P 845 environment. Conversely, the correspondence between the soybean-associated order Mortierellales 846 and soil P (Figure 8 and Table S9) may be a result of these fungi converting inorganic or organic 847 phosphates into the bioavailable form detected in our assay (Zhang et al., 2011). Higher P levels 848 brought about by the activity of mortierellalean fungi may explain, in part, the improved corn yield 849 following years of soybean. However, it is also possible that the positive correlation between the 850 relative abundance of Mortierella sp. and corn yield (Figure S8) is related to other mechanisms, such 851 as the impact of Mortierella sp. on plant growth-promoting hormones (Li et al., 2018).

852 6. CONCLUSIONS

853 In this study, we showed that continuous soybean and corn monoculture is associated with significant 854 changes in bulk soil fungal communities, with specific functional groups of fungi proliferating under 855 one or the other crop host. Our hypothesis that fungal communities would become progressively 856 similar to corresponding communities in long-term monoculture plots was supported. However, the 857 hypothesis that crop rotation would result in greater fungal diversity was not supported. Instead, alpha 858 diversity was found to increase in response to corn monoculture rather than to crop rotation. The 859 greatest impact of corn monoculture on soybean yield was arguably its effect on SCN density, which 860 decreased over consecutive years of corn and increased over consecutive years of soybean. Our 861 hypothesis that the diversity and relative abundance of nematophagous fungi would track the 862 population density of SCN was supported for nematode-trapping fungi and for many potential 863 nematode egg parasites, suggesting that members of these guilds hold promise as biocontrol agents 864 against SCN. Relationships between corn monoculture, soil P, and AMF and Sebacinales hint at the 865 adaptive role of these fungi in P acquisition, whereas relationships between soybean monoculture, bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

866 Mortierellales, and soil P may help explain the increased yields of corn in rotation with soybean. Future 867 research should test nematode-trapping and nematode egg-parasitic isolates for their potential to 868 control SCN populations and address the roles of corn and soybean-associated AMF, sebacinalean, and 869 mortierellalean fungi in P solubilization and uptake. Our findings serve to illustrate the complexity of 870 relationships between crop sequences, soil fungal communities, plant parasitic nematodes, and soil 871 properties and provide fundamental knowledge that can be used to guide future strategies to manage 872 SCN and improve corn and soybean yields.

873 AUTHOR CONTRIBUTIONS

874 NS assisted in collection of soil samples and DNA isolations, processed Illumina metabarcoding data, 875 developed bioinformatics and statistical approaches, analyzed the data, and wrote the manuscript. KEB 876 co-conceived of, supervised and provided guidance on the research, and edited the manuscript. WH 877 helped collect cysts and soil samples, provided training and analytical pipelines for analyzing the data, 878 and edited the manuscript. DH helped collect cysts and soil samples and edited the manuscript. SC 879 co-conceived of and supervised the project, maintained the long-term research study plot site, and 880 provided expertise on SCN biology.

881 FUNDING

882 This research was supported by United States Department of Agriculture (USDA) National Institute of 883 Food and Agriculture (NIFA) grant 2015-67013-23419.

884 ABBREVIATIONS

885 AMF: arbuscular mycorrhizal fungi

886 CLR: centered log ratio

887 ITS1: internal transcribed spacer 1

888 OTU: operational taxonomic unit

889 SCN: soybean cyst nematode

890 ACKNOWLEDGEMENTS

891 Fungal cultures and tissue for mock communities were generously supplied by Todd Burnes, Peter 892 Kennedy, Senyu Chen, Harold Corby Kistler, and Timothy James. The authors would like to thank 893 Wayne Gottschalk, Cathryn Johnson, Jeff Ballman, and undergraduate workers for their help in 894 collecting soil samples and Bryani Lee for her assistance in processing soil samples and extracting 895 DNA. We also thank Nicholas Dunn at the Minnesota Supercomputing Institute and Jon Palmer for 896 help with implementing AMPtk, and Gabriel Al-Ghalith for help with statistical analysis using centered 897 log ratios. We would also like to acknowledge Daryl Gohl, Allison MacLean, and Corbin Dirkx at the 898 University of Minnesota Genomics Center for sharing their expertise and methods development in 899 amplification and sequencing of fungal DNA for metabarcoding experiments.

900 Conflict of interest statement: The submitted work was carried out in the absence of any personal, 901 professional or financial relationships that could potentially be construed as a conflict of interest. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

902 SUPPLEMENTARY MATERIAL

903 Supplementary Materials and Methods.

904 Supplementary code, find.longest.taxonomy.R, can be found in the following GitHub repository: 905 https://github.com/stro0070/OTU-taxonomy-assignment

906 Figure S1 | Rarefaction curves for bulk soil samples.

907 Figure S2 | Community profiles of fungal mock community.

908 Figure S3 | Principle coordinates analysis of technical replicates.

909 Figure S4 | SCN Density by season for each crop host.

910 Figure S5 | Community profiles of ecological guilds assigned by FUNGuild.

911 Figure S6 | Spearman correlations between phyla, trophic modes, and nematophagous guilds and SCN 912 density, monoculture year, and yield.

913 Figure S7 | Differential abundance of taxa by crop sequence.

914 Figure S8 | Spearman correlations between taxa and SCN density, monoculture year, and yield.

915 Figure S9 | Phylum differential abundance by season.

916 Figure S10 | Trophic mode differential abundance by season.

917 Figure S11. Differential abundance of nematophagous guilds by season

918 Table S1 | Corn and soybean crop sequences

919 Table S2 | Mock community 5 composition.

920 Table S3 | Taxonomy assignments.

921 Table S4 | Mock community taxa identified by bioinformatics pipeline.

922 Table S5 | OTUs detected in synthetic mock community sample.

923 Table S6 | Nematophagous guilds.

924 Table S7 | Mean values for soil properties within crop sequences

925 Table S8 | Correlations between soil properties and corn and soybean yields

926 Table S9 | Alpha diversity metrics by crop host

927 Table S10 | Spearman correlations between alpha diversity and SCN, monoculture year, and yield

928 Table S11 | Taxonomic composition of trophic modes assigned by FUNGuild. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

929 Table S12 | Taxonomic composition of ecological guilds assigned by FUNGuild.

930 Table S13 | Nematophagous taxa detected and their assignment to nematophagous guilds.

931 Table S14 | Differential abundance of fungal phyla, trophic modes, and nematophagous guilds across 932 crop sequences

933 Table S15 | Differential abundance of taxa by season.

934 REFERENCES

935 Aitchison, J. (1982). The statistical analysis of compositional data. J. R. Stat. Soc. Ser. B 44, 139–177.

936 Anderson, M. J. (2001). A new method for non-parametric multivariate analysis of variance. Austral 937 Ecol. 26, 32–46. doi:10.1111/j.1442-9993.2001.tb00081.x.

938 Aveskamp, M. M., de Gruyter, J., Woudenberg, J. H. C., Verkley, G. J. M., and Crous, P. W. (2010). 939 Highlights of the Didymellaceae: A polyphasic approach to characterise Phoma and related 940 pleosporalean genera. Stud. Mycol. 65, 1–60. doi:10.3114/sim.2010.65.01.

941 Bai, L., Cui, J., Jie, W., and Cai, B. (2015). Analysis of the community compositions of rhizosphere 942 fungi in soybeans continuous cropping fields. Microbiol. Res. 180, 49–56. 943 doi:10.1016/j.micres.2015.07.007.

944 Bainard, L. D., Navarro-Borrell, A., Hamel, C., Braun, K., Hanson, K., and Gan, Y. (2017). Increasing 945 the frequency of pulses in crop rotations reduces soil fungal diversity and increases the proportion 946 of fungal pathotrophs in a semiarid agroecosystem. Agric. Ecosyst. Environ. 240, 206–214. 947 doi:10.1016/j.agee.2017.02.020.

948 Bender, R. R., Haegele, J. W., and Below, F. E. (2015). Nutrient uptake, partitioning, and 949 remobilization in modern soybean varieties. Agron. J. 107, 563–573. doi:10.2134/agronj14.0435.

950 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful 951 approach to multiple testing. J. R. Stat. Soc. 57, 289–300. doi:10.2307/2346101.

952 Berg, G., Zachow, C., Lottmann, J., Costa, R., Smalla, K., and Go, M. (2005). Impact of plant species 953 and site on rhizosphere-associated fungi antagonistic to Verticillium dahliae Kleb. Appl. Environ. 954 Microbiol. 71, 4203–4213. doi:10.1128/AEM.71.8.4203.

955 Boland, G. J., Melzer, M. S., Hopkin, A., Higgins, V., and Nassuth, A. (2004). Climate change and 956 plant diseases in Ontario. Can. J. Plant Pathol. 26, 335–350. doi:10.1080/07060660409507151.

957 Bordallo, J. J., Lopez-Llorca, L. V., Jansson, H. B., Salinas, J., Persmark, L., and Asensio, L. (2002). 958 Colonization of plant roots by egg-parasitic and nematode-trapping fungi. New Phytol. 154, 491– 959 499. doi:10.1046/j.1469-8137.2002.00399.x.

960 Bray, J., and Curtis, J. (1957). An ordination of the upland forest communities of southern Wisconsin. 961 Ecol. Monogr. 27, 325–349.

962 Broeckling, C. D., Broz, A. K., Bergelson, J., Manter, D. K., and Vivanco, J. M. (2008). Root exudates 963 regulate soil fungal community composition and diversity. Appl. Environ. Microbiol. 74, 738– bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

964 744. doi:10.1128/AEM.02188-07.

965 Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., and Holmes, S. P. (2016). 966 DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581– 967 583. doi:10.1038/nmeth.3869.

968 Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F. D., Costello, E. K., et al. 969 (2010). QIIME allows analysis of high-throughput community sequencing data. Nat. Publ. Gr. 7, 970 335–336. doi:10.1038/nmeth0510-335.

971 Carris, L. M., Glawe, D. A., and Gray, L. E. (1986). Isolation of the soybean pathogens Corynespora 972 cassiicola and Phialophora gregata from cysts of Heterodera glycines in Illinois. Mycologia 78, 973 503–506.

974 Castillo Lopez, D., Zhu-Salzman, K., Ek-Ramos, M. J. ulissa, and Sword, G. A. (2014). The 975 entomopathogenic fungal endophytes Purpureocillium lilacinum (formerly Paecilomyces 976 lilacinus) and Beauveria bassiana negatively affect cotton aphid reproduction under both 977 greenhouse and field conditions. PLoS One 9, e103891. doi:10.1371/journal.pone.0103891.

978 Chen, F., and Chen, S. (2002). Mycofloras in cysts, females, and eggs of the soybean cyst nematode in 979 Minnesota. Appl. Soil Ecol. 19, 35–50.

980 Chen, S. (2007). Suppression of Heterodera glycines in soils from fields with long-term soybean 981 monoculture. Biocontrol Sci. Technol. 17, 125–134. doi:10.1080/09583150600937121.

982 Chen, S. (2011). Soybean cyst nematode management guide. Univ. Minnesota Ext. Available at: 983 http://www.extension.umn.edu/agriculture/soybean/soybean-cyst-nematode/EFANS-Soybean- 984 SoybeanCystNematode-WebQuality.pdf [Accessed November 30, 2018].

985 Chen, S., and Chen, F. (2003). Fungal parasitism of Heterodera glycines eggs as influenced by egg age 986 and pre-colonization of cysts by other fungi. J. Nematol. 35, 271–277.

987 Chen, S., and Dickson, D. W. (2012). “Biological control of plant-parasitic nematodes,” in Practical 988 Plant Nematology, eds. R. H. Manzanilla-López and N. Marbán-Mendoza (Guadalajara, Jalisco, 989 Mexico: Colegio de Postgraduados and Mundi-Prensa, Biblioteca Básica de Agricultura.), 761– 990 811.

991 Chen, S., and Reese, C. D. (1999). Parasitism of the nematode Heterodera glycines by the fungus 992 Hirsutella rhossiliensis as influenced by crop sequence. J. Nematol. 31, 437–444.

993 Cline, L. C., Hobbie, S. E., Madritch, M. D., Buyarski, C. R., Tilman, D., and Cavender-Bares, J. M. 994 (2018). Resource availability underlies the plant-fungal diversity relationship in a grassland 995 ecosystem. Ecology 99, 204–216. doi:10.1002/ecy.2075.

996 Cooper, J. E. (2007). Early interactions between legumes and rhizobia : disclosing complexity in a 997 molecular dialogue. 103, 1355–1365. doi:10.1111/j.1365-2672.2007.03366.x.

998 Degenkolb, T., and Vilcinskas, A. (2016). Metabolites from nematophagous fungi and nematicidal 999 natural products from fungi as alternatives for biological control. Part I: metabolites from 1000 nematophagous ascomycetes. Appl. Microbiol. Biotechnol. 100, 3799–3812. doi:10.1007/s00253- bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1001 015-7234-5.

1002 Doran, J. W., Wilhelm, W. W., and Power, J. F. (1984). Crop residue removal and soil productivity 1003 with no-till corn, sorghum, and soybean. Soil Sci. Soc. Am. J. 48, 640–645. 1004 doi:10.2136/sssaj1984.03615995004800030034x.

1005 Edgar, R. C. (2013). UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat. 1006 Methods 10, 996–8. doi:10.1038/nmeth.2604.

1007 Fell, J. W., Roeijmans, H., and Boekhout, T. (1999). Cystofilobasidiales, a new order of 1008 basidiomycetous yeasts. Int. J. Syst. Bacteriol. 49, 907–913.

1009 Francl, L. J., and Dropkin, V. H. (1985). Glomus fasciculatum, a weak pathogen of Heterodera 1010 glycines. J. Nematol. 17, 470–5.

1011 Frøslev, T. G., Kjøller, R., Bruun, H. H., Ejrnæs, R., Brunbjerg, A. K., Pietroni, C., et al. (2017). 1012 Algorithm for post-clustering curation of DNA amplicon data yields reliable biodiversity 1013 estimates. Nat. Commun. 8. doi:10.1038/s41467-017-01312-x.

1014 Garbe, J. R. (2015). Gopher-biotools. (Version 1.0) [Computer program]. Available at 1015 https://bitbucket.org/jgarbe/gopher-biotools.

1016 Garbeva, P., Hol, W. H. G., Termorshuizen, A. J., Kowalchuk, G. A., and De Boer, W. (2011). 1017 Fungistasis and general soil biostasis - A new synthesis. Soil Biol. Biochem. 43, 469–477. 1018 doi:10.1016/j.soilbio.2010.11.020.

1019 Gentry, L. F., Ruffo, M. L., and Below, F. E. (2013). Identifying factors controlling the continuous 1020 corn yield penalty. Agron. J. 105, 295–303. doi:10.2134/agronj2012.0246.

1021 Ghayedi, S., and Abdollahi, M. (2013). Biocontrol potential of Metarhizium anisopliae (Hypocreales: 1022 Clavicipitaceae), isolated from suppressive soils of the Boyer-Ahmad region. J. Plant Prot. Res. 1023 53, 1–7. doi:10.2478/jppr-2013-0025.

1024 Gloor, G. B. (2018). ALDEx2: ANOVA-like differential expression tool for compositional data. 1025 doi:10.1186/2049-2618-1-12.

1026 Gloor, G. B., Macklaim, J. M., Pawlowsky-Glahn, V., and Egozcue, J. J. (2017). Microbiome datasets 1027 are compositional: And this is not optional. Front. Microbiol. 8, 1–6. 1028 doi:10.3389/fmicb.2017.02224.

1029 Gohl, D. M., Vangay, P., Garbe, J., MacLean, A., Hauge, A., Becker, A., et al. (2016). Systematic 1030 improvement of amplicon marker gene methods for increased accuracy in microbiome studies. 1031 Nat. Biotechnol. 34, 942–948. doi:10.1038/nbt.3601.

1032 Gomes, N. C. M., Fagbola, O., Costa, R., Rumjanek, N. G., Buchner, A., Mendona-hagler, L., et al. 1033 (2003). Dynamics of fungal communities in bulk and maize rhizosphere soil in the tropics. Appl. 1034 Environ. Microbiolgy 69, 3758–3766. doi:10.1128/AEM.69.7.3758.

1035 Grabau, Z. J., and Chen, S. (2016a). Determining the role of plant-parasitic nematodes in the corn- 1036 soybean crop rotation yield effect using nematicide application: II. Soybean. Agron. J. 108, 1168– bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1037 1179. doi:10.2134/agronj2015.0432.

1038 Grabau, Z. J., and Chen, S. (2016b). Determining the role of plant-parasitic nematodes in the corn– 1039 soybean crop rotation yield effect using nematicide application: I. Corn. Agron. J. 108, 782–793. 1040 doi:10.2134/agronj2015.0432.

1041 Hamid, M. I., Hussain, M., Wu, Y., Zhang, X., Xiang, M., and Liu, X. (2017). Successive soybean- 1042 monoculture cropping assembles rhizosphere microbial communities for the soil suppression of 1043 soybean cyst nematode. FEMS Microbiol. Ecol. 93, 1–10. doi:10.1093/femsec/fiw222.

1044 Hartman, G. L., West, E. D., and Herman, T. K. (2011). Crops that feed the World 2. Soybean- 1045 worldwide production, use, and constraints caused by pathogens and pests. Food Secur. 3, 5–17. 1046 doi:10.1007/s12571-010-0108-x.

1047 Hartmann, M., Frey, B., Mayer, J., Mäder, P., and Widmer, F. (2014). Distinct soil microbial diversity 1048 under long-term organic and conventional farming. ISME J. 9, 1177–1194. 1049 doi:10.1038/ismej.2014.210.

1050 Hiltner, L. (1904). Über neuere erfahrungen und probleme auf dem gebiete der bodenbakteriologie 1051 unter besonderer berücksichtigung der gründüngung und brache. Arb. der Dtsch. Landwirtsch. 1052 Gesellschaft 98, 59–78.

1053 Hu, W., Samac, D. A., Liu, X., and Chen, S. (2017). Microbial communities in the cysts of soybean 1054 cyst nematode affected by tillage and biocide in a suppressive soil. Appl. Soil Ecol. 119, 396–406. 1055 doi:10.1016/j.apsoil.2017.07.018.

1056 Hu, W., Strom, N., Rajendran, D. B., Chen, S., and Bushley, K. E. B. (2018). Mycobiome of Cysts of 1057 the Soybean Cyst Nematode under Long Term Crop Rotation. Front. Microbiol. 9, 386. 1058 doi:10.3389/FMICB.2018.00386.

1059 Impullitti, A. E., and Malvick, D. K. (2013). Fungal endophyte diversity in soybean. J. Appl. Microbiol. 1060 114, 1500–1506. doi:10.1111/jam.12164.

1061 Jaffee, B. (2003). Correlations between most probable number and activity of nematode-trapping fungi. 1062 Phytopathol. 93, 1599-1605. 93, 1599–1605. 1063 doi:http://dx.doi.org/10.1094/PHYTO.2003.93.12.1599.

1064 Jaffee, B., Phillips, R., Muldoon, A., and Mangel, M. (1992). Density-dependent host-pathogen 1065 dynamics in soil microcosms. Ecology 73, 495–506.

1066 Jaffee, B., Tedford, E., and Muldoon, A. (1993). Tests for density-dependent parasitism of nematodes 1067 by nematode-trapping and endoparasitic fungi. Biol. Control 3, 329–336.

1068 Jiménez, D. J., Hector, R. E., Riley, R., Lipzen, A., Kuo, R. C., Amirebrahimi, M., et al. (2017). Draft 1069 genome sequence of Coniochaeta ligniaria NRRL 30616, a lignocellulolytic fungus for 1070 bioabatement of inhibitors in plant biomass hydrolysates. Genome Announc. 5, e01476-16. 1071 doi:10.1128/genomeA.01476-16.

1072 Johnson, N., Copeland, P., Crookston, R., and Pfleger, F. (1992). Mycorrhizae: possible explanation 1073 for yield decline with continuous corn and soybean. Agron. J. 84, 387–390. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1074 Johnson, N., Pfleger, F. L., Crookston, R. K., Simmons, S. R., and Copeland, P. J. (1991). Vesicular– 1075 arbuscular mycorrhizas respond to corn and soybean cropping history. New Phytol. 117, 657–663. 1076 doi:10.1111/j.1469-8137.1991.tb00970.x.

1077 Juba, J., Meyer, S., Humber, R., Liu, X. Z., Nitao, J., and Huettel, R. (2004). Activity of fungal culture 1078 filtrates against soybean cyst nematode and root-knot nematode egg hatch and juvenile motility. 1079 Nematology 6, 23–32. doi:10.1163/156854104323072883.

1080 Karlen, D. L., Varvel, G. E., Bullock, D. G., and Cruse, R. M. (1994). “Crop rotations for the 21st 1081 century,” in Advances in Agronomy, ed. D. Sparks (San Diego, CA: Academic Press), 1–45. 1082 doi:https://doi.org/10.1016/S0065-2113(08)60611-2.

1083 Koenning, S. R., and Wrather, J. A. (2010). Suppression of soybean yield potential in the continental 1084 United States by plant diseases from 2006 to 2009. Plant Heal. Prog. Available at: 1085 http://www.plantmanagementnetwork.org/pub/php/research/2010/yield/.

1086 Kohout, P., Sudová, R., Janoušková, M., Čtvrtlíková, M., Hejda, M., Pánková, H., et al. (2014). 1087 Comparison of commonly used primer sets for evaluating arbuscular mycorrhizal fungal 1088 communities: Is there a universal solution? Soil Biol. Biochem. 68, 482–493. 1089 doi:10.1016/j.soilbio.2013.08.027.

1090 Kristin, A., and Miranda, H. (2013). The root microbiota-a fingerprint in the soil? Plant Soil 370, 671– 1091 686. doi:10.1007/s11104-013-1647-7.

1092 Kruskal, J. B. (1964). Multidimensional scaling by optimizing goodness of fit to a nonmetric 1093 hypothesis. Psychometrika 29, 1–27. doi:10.1007/BF02289565.

1094 Kruskal, W. H., and Wallis, W. A. (1952). Use of ranks in one-criterion variance analysis. J. Am. Stat. 1095 Assoc. 47, 583–621.

1096 Kumar, V., Sarma, M. V. R. K., Saharan, K., Srivastava, R., Kumar, L., Sahai, V., et al. (2012). Effect 1097 of formulated root endophytic fungus Piriformospora indica and plant growth promoting 1098 rhizobacteria fluorescent pseudomonads R62 and R81 on Vigna mungo. World J. Microbiol. 1099 Biotechnol. 28, 595–603. doi:10.1007/s11274-011-0852-x.

1100 Kuske, C. R., Ticknor, L. O., Miller, M. E., Dunbar, J. M., Davis, J. A., Barns, S. M., et al. (2002). 1101 Comparison of soil bacterial communities in rhizospheres of three plant species and the 1102 interspaces in an arid grassland. Appl. Environ. Microbiol. 68, 1854–1863. 1103 doi:10.1128/AEM.68.4.1854-1863.2002.

1104 Landell, M. F., Brandão, L. R., Barbosa, A. C., Ramos, J. P., Safar, S. V. B., Gomes, F. C. O., et al. 1105 (2014). Hannaella pagnoccae sp. nov., a tremellaceous yeast species isolated from plants and soil. 1106 Int. J. Syst. Evol. Microbiol. 64, 1970–1977. doi:10.1099/ijs.0.059345-0.

1107 Lapsansky, E. R., Milroy, A. M., Andales, M. J., and Vivanco, J. M. (2016). Soil memory as a potential 1108 mechanism for encouraging sustainable plant health and productivity. Curr. Opin. Biotechnol. 38, 1109 137–142. doi:10.1016/j.copbio.2016.01.014.

1110 Lauber, C. L., Strickland, M. S., Bradford, M. A., and Fierer, N. (2008). The influence of soil properties 1111 on the structure of bacterial and fungal communities across land-use types. Soil Biol. Biochem. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1112 40, 2407–2415. doi:10.1016/j.soilbio.2008.05.021.

1113 Lee, S. C., Tang, M. S., Lim, Y. A. L., Choy, S. H., Kurtz, Z. D., Cox, L. M., et al. (2014). Helminth 1114 colonization is associated with increased diversity of the gut microbiota. PLoS Negl. Trop. Dis. 8. 1115 doi:10.1371/journal.pntd.0002880.

1116 Legendre, P., and Anderson, M. J. (1999). Distance-based redundancy analysis: testing multispecies 1117 responses in multifactorial ecological experiments. Ecol. Monogr. 69, 1–24. doi:10.1890/0012- 1118 9615(1999)069[0001:DBRATM]2.0.CO;2.

1119 Legendre, P., and Gallagher, E. D. (2001). Ecologically meaningful transformations for ordination of 1120 species data. Oecologia 129, 271–280. doi:10.1007/s004420100716.

1121 Levene, H. (1960). “Robust tests for equality of variances,” in Contributions to probability and 1122 statistics: essays in honor of Harold Hotelling, ed. I. Olkin (Stanford University Press), 278–292.

1123 Li, F., Chen, L., Redmile-Gordon, M., Zhang, J., Zhang, C., Ning, Q., et al. (2018). Mortierella 1124 elongata’s roles in organic agriculture and crop growth promotion in a mineral soil. L. Degrad. 1125 Dev., 1642–1651. doi:10.1002/ldr.2965.

1126 Lin, X., Feng, Y., Zhang, H., Chen, R., Wang, J., Zhang, J., et al. (2012). Long-term balanced 1127 fertilization decreases arbuscular mycorrhizal fungal diversity in an arable soil in north China 1128 revealed by 454 pyrosequencing. Environ. Sci. Technol. 46, 5764–5771. doi:10.1021/es3001695.

1129 Lindahl, B. D., Nilsson, R. H., Tedersoo, L., Abarenkov, K., Carlsen, T., Kjøller, R., et al. (2013). 1130 Fungal community analysis by high-throughput sequencing of amplified markers--a user’s guide. 1131 New Phytol. 199, 288–99. doi:10.1111/nph.12243.

1132 Linford, M. B. (1937). Stimulated activity of natural enemies of nematodes. Science 85, 123–124.

1133 Liu, S., Kandoth, P. K., Lakhssassi, N., Kang, J., Colantonio, V., Heinz, R., et al. (2017). The soybean 1134 GmSNAP18 gene underlies two types of resistance to soybean cyst nematode. Nat. Commun. 8, 1135 14822. doi:10.1038/ncomms14822.

1136 Liu, S., Su, H., Su, X., Zhang, F., Liao, G., and Yang, X. (2014a). Arthrobotrys xiangyunensis, a novel 1137 nematode-trapping taxon from a hot-spring in Yunnan Province, China. Phytotaxa 174, 89–96. 1138 doi:10.11646/phytotaxa.174.2.3.

1139 Liu, X. Z., and Chen, S. Y. (2000). Parasitism of Heterodera glycines by Hirsutella spp. in Minnesota 1140 soybean fields. Biol. Control 19, 161–166. doi:10.1006/bcon.2000.0855.

1141 Liu, X., Zhang, J., Gu, T., Zhang, W., Shen, Q., Yin, S., et al. (2014b). Microbial community diversities 1142 and taxa abundances in soils along a seven-year gradient of potato monoculture using high 1143 throughput pyrosequencing approach. PLoS One 9, 1–10. doi:10.1371/journal.pone.0086610.

1144 Lungu, O., and Dynoodt, R. (2008). Acidification from long-term use of urea and its effect on selected 1145 soil properties. African J. Food, Agric. Nutr. Dev. 8, 63–76. doi:va07009.

1146 Ma, A., Zhuang, X., Wu, J., Cui, M., Lv, D., Liu, C., et al. (2013). Ascomycota members dominate 1147 fungal communities during straw residue decomposition in arable soil. PLoS One 8, 1–9. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1148 doi:10.1371/journal.pone.0066146.

1149 Mäder, P., Fliessbach, A., Dubois, D., Gunst, L., Fried, P., and Niggli, U. (2002). Soil fertility and 1150 biodiversity in organic farming. Science 296, 1694–1697. doi:10.1126/science.1071148.

1151 Malapi-Wight, M., Salgado-Salazar, C., Demers, J., Veltri, D., and Crouch, J. A. (2015). Draft genome 1152 sequence of Dactylonectria macrodidyma, a plant-pathogenic fungus in the Nectriaceae. Genome 1153 Announc. 3, 2014–2015. doi:10.1128/genomeA.00278-15.Copyright.

1154 Maltz, M. R., Treseder, K. K., and McGuire, K. L. (2008). Links between plant and fungal diversity in 1155 habitat fragments of coastal sage scrub. Kearney Rep., 1–8. doi:10.5061/dryad.13s7s.

1156 Manzanilla-Lopez, Rosa H; Esteves, Ivania; Finetti-Sialer, Mariella; Hirsch, Penny; Ward, Elaine, 1157 Devonshire, Jean; and Hidalgo-Diaz, L. (2013). Pochonia chalamydosporia: advances and 1158 challanges to improve its performance as a biological control agent of sedentary endo-parasitic 1159 nematodes. J. Nematol. 45, 1–7.

1160 Marin-Felix, Y., Groenewald, J. Z., Cai, L., Chen, Q., Marincowitz, S., Barnes, I., et al. (2017). Genera 1161 of phytopathogenic fungi: GOPHY 1. Stud. Mycol. 86, 99–216. 1162 doi:10.1016/j.simyco.2017.04.002.

1163 Martín-Fernández, J. A., Barceló-Vidal, C., and Pawlowsky-Glahn, V. (2003). Dealing with zeros and 1164 missing values in compositional data sets using nonparametric imputation. Math. Geol. 35, 253– 1165 278. doi:10.1023/A:1023866030544.

1166 Martínez-García, L. B., Richardson, S. J., Tylianakis, J. M., Peltzer, D. A., and Dickie, I. A. (2015). 1167 Host identity is a dominant driver of mycorrhizal fungal community composition during 1168 ecosystem development. New Phytol. 205, 1565–1576. doi:10.1111/nph.13226.

1169 Mcdonald, D., Hyde, E., Debelius, J. W., Morton, J. T., Gonzalez, A., Ackermann, G., et al. (2018). 1170 American gut: an open platform for citizen science. mSystems 3, 1–28.

1171 McMurdie, P. J., and Holmes, S. (2013). Phyloseq: An R package for reproducible interactive analysis 1172 and graphics of microbiome census data. PLoS One 8, e61217.

1173 Meade, B., Puricelli, E., Mcbride, W., Valdes, C., Hoffman, L., Foreman, L., et al. (2016). Corn and 1174 soybean production costs and export competitiveness in Argentina, Brazil, and the United States. 1175 Econ. Res. Serv. EIB-154, 1–52. Available at: www.ers.usda.gov/publications/eib-economic- 1176 information-bulletin/eib-154%0ADownload.

1177 Mendes, L. W., Kuramae, E. E., Navarrete, a a, van Veen, J. a, and Tsai, S. M. (2014). Taxonomical 1178 and functional microbial community selection in soybean rhizosphere. Isme J 8, 1577–1587. 1179 doi:10.1038/ismej.2014.17.

1180 Mendiburu, F. de (2017). Agricolae: statistical procedures for agricultural research. R package version 1181 1.2-8. Available at: https://cran.r-project.org/package=agricolae.

1182 Miller, M. H. (2000). Arbuscular mycorrhizae and the phosphorus nutrition of maize: A review of 1183 Guelph studies. Can. J. Plant Sci. 80, 47–52. doi:10.4141/P98-130. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1184 Nguyen, N. H., Song, Z., Bates, S. T., Branco, S., Tedersoo, L., Menke, J., et al. (2016). FUNGuild: 1185 An open annotation tool for parsing fungal community datasets by ecological guild. Fungal Ecol. 1186 20, 241–248. doi:10.1016/j.funeco.2015.06.006.

1187 Niblack, T., Heinz, R., Smith, G., and Donald, P. (1993). Distribution, Density, and Diversity of 1188 Heterodera glycines in Missouri. J. Nematol. 25, 880–886.

1189 O’Donnell, K., Sutton, D. A., Fothergill, A., McCarthy, D., Rinaldi, M. G., Brandt, M. E., et al. (2008). 1190 Molecular phylogenetic diversity, multilocus haplotype nomenclature, and in vitro antifungal 1191 resistance within the Fusarium solani species complex. J. Clin. Microbiol. 46, 2477–2490. 1192 doi:10.1128/JCM.02371-07.

1193 Oksanen, J. (2015). Multivariate analysis of ecological communities in R: vegan tutorial. R Doc., 43. 1194 doi:10.1016/0169-5347(88)90124-3.

1195 Oksanen, J., Blanchet, F., Kindt, R., Legendre, P., Minchin, P., O’Hara, R., et al. (2016). Vegan: 1196 community ecology package. R package version 2.3-4. Available at: https://cran.r- 1197 project.org/package=vegan.

1198 Palmer, J., Lindner, D., and Drive, G. P. (2018). Non-biological synthetic spike-in controls and the 1199 AMPtk software pipeline improve fungal high throughput amplicon sequencing data. bioRxiv, 1– 1200 31.

1201 Parniske, M. (2008). Arbuscular mycorrhiza: The mother of plant root endosymbioses. Nat. Rev. 1202 Microbiol. 6, 763–775. doi:10.1038/nrmicro1987.

1203 Peiffer, J. A., Spor, A., Koren, O., Jin, Z., Tringe, S. G., Dangl, J. L., et al. (2013). Diversity and 1204 heritability of the maize rhizosphere microbiome under field conditions. Proc. Natl. Acad. Sci. U. 1205 S. A. 110, 6548–53. doi:10.1073/pnas.1302837110.

1206 Perez-brandan, C., Arzeno, J. L., Huidobro, J., Conforto, C., Grümberg, B., Hilton, S., et al. (2014). 1207 The effect of crop sequences on soil microbial, chemical and physical indicators and its 1208 relationship with soybean sudden death syndrome (complex of Fusarium species ). Spanish J. 1209 Agric. Res. 12, 252–264.

1210 Peterson, T. A., and Varvel, G. E. (1989). Crop yield as affected by rotation and nitrogen rate. III. 1211 Corn. Agron. J. 81, 735–738.

1212 Plaza, D. F., Schmieder, S. S., Lipzen, A., Lindquist, E., and Künzler, M. (2016). Identification of a 1213 novel nematotoxic protein by challenging the model mushroom Coprinopsis cinerea with a 1214 fungivorous nematode. Genes|Genomes|Genetics 6, 87–98. doi:10.1534/g3.115.023069.

1215 Porter, P. M., Chen, S. Y., Reese, C. D., and Klossner, L. D. (2001). Population response of soybean 1216 cyst nematode to long term corn – soybean cropping sequences in Minnesota. Agron. J. 93, 619– 1217 626. doi:10.2134/agronj2001.933619x.

1218 R Core Team (2018). R: A language and environment for statistical computing. Available at: 1219 https://www.r-project.org/.

1220 Ramirez, K. S., Lauber, C. L., Knight, R., Bradford, M. A., and Fierer, N. (2010). Consistent effects bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1221 of nitrogen fertilization on soil bacterial communities in contrasting systems. Ecology 91, 3414– 1222 3463. doi:doi:10.1890/10-0426.1.

1223 Regents of the University of Minnesota (2018). Historic Weather Reports. South. Res. Outreach Cent. 1224 - Waseca, MN. Available at: https://sroc.cfans.umn.edu/weather-sroc/historic-reports [Accessed 1225 August 5, 2018].

1226 RStudio Team (2016). RStudio: Integrated Development Environment for R. Available at: 1227 http://www.rstudio.com/.

1228 Rupe, J. C., Robbins, R. T., and Gbur, E. E. (1997). Effect of crop rotation on soil population densities 1229 of Fusarium solani and Heterodera glycines and on the development of sudden death syndrome 1230 of soybean. Crop Prot. 16, 575–580. doi:10.1016/S0261-2194(97)00031-8.

1231 Seifert, C. A., Roberts, M. J., and Lobell, D. B. (2017). Continuous corn and soybean yield penalties 1232 across hundreds of thousands of fields. Agron. J. 109, 541–548. doi:10.2134/agronj2016.03.0134.

1233 Senés-Guerrero, C., and Schüßler, A. (2016). A conserved arbuscular mycorrhizal fungal core-species 1234 community colonizes potato roots in the Andes. Fungal Divers. 77, 317–333. 1235 doi:10.1007/s13225-015-0328-7.

1236 She, W., Bai, Y., Zhang, Y., Qin, S., Feng, W., Sun, Y., et al. (2018). Resource availability drives 1237 responses of soil microbial communities to short-term precipitation and nitrogen addition in a 1238 desert shrubland. Front. Microbiol. 9, 1–14. doi:10.3389/fmicb.2018.00186.

1239 Shiferaw, B., Prasanna, B. M., Hellin, J., and Bänziger, M. (2011). Crops that feed the world 6. Past 1240 successes and future challenges to the role played by maize in global food security. Food Secur. 1241 3, 307–327. doi:10.1007/s12571-011-0140-5.

1242 Smalla, K., Wieland, G., Buchner, A., Zock, A., Parzy, J., Roskot, N., et al. (2001). Bulk and 1243 rhizosphere soil bacterial communities studied by denaturing gradient gel electrophoresis: plant- 1244 dependent enrichment and seasonal shifts revealed. Appl. Environ. Microbiol. 67, 4742–4751. 1245 doi:10.1128/AEM.67.10.4742.

1246 Smit, E., Leeflang, P., Gommans, S., Van Den Broek, J., Van Mil, S., and Wernars, K. (2001). Diversity 1247 and seasonal fluctuations of the dominant members of the bacterial soil community in a wheat 1248 field as determined by cultivation and molecular methods. Appl. Environ. Microbiol. 67, 2284– 1249 2291. doi:10.1128/AEM.67.5.2284-2291.2001.

1250 Smith, C. R., Blair, P. L., Boyd, C., Cody, B., Hazel, A., Hedrick, A., et al. (2016). Microbial 1251 community responses to soil tillage and crop rotation in a corn/soybean agroecosystem. Ecol. 1252 Evol. 6, 8075–8084. doi:10.1002/ece3.2553.

1253 Song, J., Li, S., Xu, Y., Wei, W., Yao, Q., and Pan, F. (2016a). Diversity of parasitic fungi from 1254 soybean cyst nematode associated with long-term continuous cropping of soybean in black soil. 1255 Acta Agric. Scand. Sect. B — Soil Plant Sci. 66, 432–442. doi:10.1080/09064710.2016.1158862.

1256 Song, Z., Shen, L., Zhong, Q., Yin, Y., and Wang, Z. (2016b). Liquid culture production of 1257 microsclerotia of Purpureocillium lilacinum for use as bionematicide. Nematology 18, 719–726. 1258 doi:10.1163/15685411-00002987. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1259 Spatafora, J. W., Chang, Y., Benny, G. L., Lazarus, K., Smith, M. E., Berbee, M. L., et al. (2016). A 1260 phylum-level phylogenetic classification of zygomycete fungi based on genome-scale data. 1261 Mycologia 108, 1028–1046. doi:10.3852/16-042.

1262 Stiles, C., and Glawe, D. (1989). Colonization of soybean roots by fungi isolated from cysts of 1263 Heterodera glycines. Mycologia 81, 797–799.

1264 Stirling, G. R. (2014). Biological control of plant-parasitic nematodes: soil ecosystem management in 1265 sustainable agriculture. 2nd ed. , ed. G. R. Stirling Wallingford: CABI 1266 doi:10.1079/9781780644158.0000.

1267 Stockinger, H., Krüger, M., and Schüßler, A. (2010). DNA barcoding of arbuscular mycorrhizal fungi. 1268 New Phytol. 187, 164. doi:10.1111/j.1469-8137.2010.03262.x.

1269 Tedersoo, L., Bahram, M., Põlme, S., Kõljalg, U., Yorou, N., Wijesundera, R., et al. (2014). Global 1270 diversity and geography of soil fungi. Science 346, 1052–1053. doi:10.1126/science.aaa1185.

1271 Tukey, J. W. (1949). Comparing individual means in the analysis of variance. Biometrics 5, 99–114.

1272 Tylka, G. L., Hussey, R. S., and Roncadori, R. W. (1991). Interactions of vesicular-arbuscular 1273 mycorrhizal fungi, phosphorus, and Heterodera glycines on Soybean. J. Nematol. 23, 122–133.

1274 U.S. Department of Agriculture National Agricultural Statistics Service (2017). Acreage. Available at: 1275 https://www.usda.gov/nass/PUBS/TODAYRPT/acrg0615.pdf.

1276 Uecker, F., and Kulik, M. (1986). Pseudorobillarda sojae, a new pycnidial coelomycete from soybean 1277 stems. Mycologia 78, 449–453.

1278 UNITE Community (2017). Full UNITE+INSD dataset. Version 01.12.2017. UNITE Community. 1279 Available at: https://doi.org/10.15156/BIO/587474 [Accessed March 2, 2018].

1280 Warnock, N. D., Wilson, L., Patten, C., Fleming, C. C., Maule, A. G., and Dalzell, J. J. (2017). 1281 Nematode neuropeptides as transgenic nematicides. PLoS Pathog. 13. 1282 doi:10.1371/journal.ppat.1006237.

1283 Weiß, M., Sýkorová, Z., Garnica, S., Riess, K., Martos, F., Krause, C., et al. (2011). Sebacinales 1284 everywhere: Previously overlooked ubiquitous fungal endophytes. PLoS One 6. 1285 doi:10.1371/journal.pone.0016793.

1286 Weiß, M., Waller, F., Zuccaro, A., and Selosse, M. A. (2016). Sebacinales - one thousand and one 1287 interactions with land plants. New Phytol. 211, 20–40. doi:10.1111/nph.13977.

1288 Weiss, S., Xu, Z. Z., Peddada, S., Amir, A., Bittinger, K., Gonzalez, A., et al. (2017). Normalization 1289 and microbial differential abundance strategies depend upon data characteristics. Microbiome 5, 1290 27. doi:10.1186/s40168-017-0237-y.

1291 Wickham, H. (2009). ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag.

1292 Wu, L., Chen, J., Wu, H., Wang, J., Wu, Y., Lin, S., et al. (2016). Effects of consecutive monoculture 1293 of Pseudostellaria heterophylla on soil fungal community as determined by pyrosequencing. Sci. 1294 Rep. 6, 26601. doi:10.1038/srep26601. bioRxiv preprint doi: https://doi.org/10.1101/516575; this version posted January 9, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1295 Xin, X., Qin, S., Zhang, J., Zhu, A., Yang, W., and Zhang, X. (2017). Yield, phosphorus use efficiency 1296 and balance response to substituting long-term chemical fertilizer use with organic manure in a 1297 wheat-maize system. F. Crop. Res. 208, 27–33. doi:10.1016/j.fcr.2017.03.011.

1298 Xiong, W., Zhao, Q., Zhao, J., Xun, W., Li, R., Zhang, R., et al. (2014). Different continuous cropping 1299 spans significantly affect microbial community membership and structure in a vanilla-grown soil 1300 as revealed by deep pyrosequencing. Microb. Ecol. 70, 209–218. doi:10.1007/s00248-014-0516- 1301 0.

1302 Xu, L., Ravnskov, S., Larsen, J., Nilsson, R. H., and Nicolaisen, M. (2012). Soil fungal community 1303 structure along a soil health gradient in pea fields examined using deep amplicon sequencing. Soil 1304 Biol. Biochem. 46, 26–32. doi:10.1016/j.soilbio.2011.11.010.

1305 Yadav, V., Kumar, M., Deep, A. K., Kumar, H., Sharma, R., Tripathi, T., et al. (2010). A phosphate 1306 transporter from the root endophytic fungus Piriformospora indica plays a role in phosphate 1307 transport to the host plant. J. Biol. Chem. 285, 26532–26544. doi:10.1074/jbc.M110.111021.

1308 Yu, Z., Mo, M., Zhang, Y., and Zhang, K. (2014). “Taxonomy of nematode-trapping fungi from 1309 Orbiliaceae, Ascomycota,” in Nematode-trapping fungi, eds. K.-Q. Zhang and K. D. Hyde 1310 (Dordrecht Heidelberg New York London: Springer), 41–210. doi:10.1007/978-94-017-8730-7.

1311 Zhang, F., Shen, J., Zhang, J., Zuo, Y., Li, L., and Chen, X. (2010). Rhizosphere processes and 1312 management for improving nutrient use efficiency and crop productivity. Implications for China. 1313 1st ed. Elsevier Inc doi:10.1016/S0065-2113(10)07001-X.

1314 Zhang, H., Wu, X., Li, G., and Qin, P. (2011). Interactions between arbuscular mycorrhizal fungi and 1315 phosphate-solubilizing fungus (Mortierella sp .) and their effects on Kostelelzkya virginica 1316 growth and enzyme activities of rhizosphere and bulk soils at different salinities. Biol. Fertil. Soils 1317 47, 543–554. doi:10.1007/s00374-011-0563-3.

1318 Zhao, Q., Xiong, W., Xing, Y., Sun, Y., Lin, X., and Dong, Y. (2018). Long-term coffee monoculture 1319 alters soil chemical properties and microbial communities. Sci. Rep., 1–11. doi:10.1038/s41598- 1320 018-24537-2.

1321