Distinct responses of oomycete plant parasites according to their lifestyle in a landscape-scale metabarcoding survey

Oomycetes are an ubiquitous protistan lineage including devastating crop parasites. Although their ecology in agrosystems has been widely studied, little is known of their distribution in natural and semi-natural ecosystems. We provide here a baseline of the diversity and distribution of soil oomycetes, classified by lifestyles (biotrophy, hemibiotrophy and saprotrophy), at the landscape scale in temperate grassland and forest. From 600 soil samples, we obtained 1,148 Operational Taxonomy Units representing ∼20 million Illumina reads (region V4, 18S rRNA gene). We found a majority of hemibiotrophic plant pathogens, which are parasites spending part of their life cycle as saprotrophs after the death of the host. Overall both grassland and forest constitute an important reservoir of plant pathogens. In forests, relative abundances of obligate biotrophs and hemibiotrophs differed between regions and showed opposite responses to edaphic conditions and human-induced management intensification, suggesting different ecological requirements for these two functional guilds.


Introduction
Oomycetes are ubiquitous and widespread in terrestrial (Singer et al. 2016;Lara and Belbahri 2011; 27 Geisen et al. 2015), freshwater (Duffy et al. 2015) and marine ecosystems (Garvetto et al. 2018). As 28 protists, they are included in the superphylum Stramenopiles (or Heterokonta) in the superkingdom 29 Harosa (or "SAR") (Ruggiero et al. 2015). They include c. 2,000 known species, mostly in the two 30 crown groups, the Saprolegniales and Peronosporales (95% of the species) (Thines and Choi 2016). In 31 terrestrial ecosystems, oomycetes occur as pathogens (obligate biotrophs or hemibiotrophs) on plants 32 and other eukaryotes and, less commonly, as saprotrophs (Marano et al. 2016), the plant pathogens 33 representing more than 60% of the oomycete taxa (Thines and Kamoun 2010). Well-known examples 34 are the soil-borne downy mildews with genera like Phytophthora and Pythium and the white rusts 35 (Albugo) on plant leaves (Savory et al. 2015). The genus Pythium is one of the most important soil-36 borne plant pathogens, being ubiquitous and with an extremely wide host range, attacking the roots of 37 thousands of different plant species (Beakes and Thines 2016). Phytophthora, the "plant destroyer", is 38 responsible for the widespread rapid tree decline (Hayden et al. 2013) and for damages to important Variation partitioning among the three predictors indicated that ecosystem (grassland vs 149 forest), region (Alb, Hainich and Schorfheide) and year of sampling (2011 and 2017) together 150 accounted for 28.6% (adjusted R 2 ) of the total variation in oomycete beta diversity. 151 Ecosystem explained 16.3% of the variation, region 10.0%, and the year of sampling only 152 2.2%. Among the soil parameters that were measured in the framework of the Biodiversity 153 Exploratories (Table S4 & Table S5), carbon content (total, organic and inorganic) and 154 nitrogen (total N) were all co-correlated; they were tested separately, and based on a slightly 155 higher significance in the models, organic carbon was kept. Among the three co-correlated 156 soil components, sand was preferred over clay or silt, because of its recognized importance 157 for oomycetes occurrence. The soil and ecological parameters included in our analyses could 158 explain more variance in forests (R 2 10.5-25.3) than in grasslands (R 2 2.6-5.9) (Table S6). 159 The most parsimonious models identified an influence of soil type in grassland as well as in 160 forest (Table S6). Other explanatory factors in grassland, but only for specific lifestyles or 161 regions, were the LUI index, mowing intensity, the soil C/N ratio and the organic C content. 162 In addition to soil type, more significant explanatory factors were identified in forest: main 163 tree species, soil pH, sand content, forest management intensity. Other factors, selected by a 164 minority of the models, were forest developmental stage, organic C content, and C/N ratio 165 (Table S6). . 166 We further investigated the effects of these selected environmental parameters on each of the 167 oomycete lifestyles, by comparing their relative abundances in each ecosystem. 168 Hemibiotrophs and biotrophs (and to a lesser extent, also saprotrophs) showed opposite 169 responses to environmental parameters, both in grassland (Fig. S4) and forest (Fig. S5). Less 170 significant differences were found in grassland than in forest, since the OTUs in the former increasing saprotrophs (Fig. S4A), a tendency also partially reflected by the communities of 174 specific soil types that were either only present in Hainich or in Schorfheide (Fig. S4B). 175 Hemibiotrophs decreased at a higher land use intensity (LUI index) (Fig. S4C). An increasing 176 C/N ratio, indicative of soils poor in nitrogen, led to a relative decrease of hemibiotrophs and 177 an increase of biotrophs (Fig. S4D). In soils with high organic carbon content, biotrophs 178 decreased and saprotrophs increased (Fig. S4E). 179 In forest, hemibiotrophs and biotrophs together with saprotrophs showed opposite shifts in 180 relative abundances patterns. At the regional scale, Schorfheide differed from Hainich and 181 Alb by a decrease of hemibiotrophs and an increase of biotrophs and saprotrophs (Fig. S5A), 182 a trend mirrored by the soil types unique to each region (Fig. S5B) and further by the tree 183 species unique to Schorfheide, i.e. pine and oak (Fig. S5C). Hemibiotrophs were reduced, 184 while biotrophs and saprotrophs slightly increased with increasing forest management 185 intensity and increasing sand content (Fig. S5D & E). With an increasing C/N ratio, 186 hemibiotrophs decreased and biotrophs increased (as in grasslands) but also saprotrophs 187 increased (Fig. S5F). Hemibiotrophs decreased in more acidic soils (mostly present in 188 Schorfheide), while biotrophs and saprotrophs increased (Fig. S5G). 189

190
Our data stemmed from a thorough sampling across three regions spanning a South-North 191 gradient in Germany using taxon-specific primers. Based on the high sequencing depth 192 (saturation was reached) ( richness not yet taxonomically recorded or sequenced in the ITS1 database. Consistently with 196 previous studies on protists, a high alpha-diversity and a low beta-diversity were found endemicity of oomycetes, but thank to our thorough sampling, high sequencing depth and 199 the use of taxon-specific primers, we showed that almost all OTUs were shared between 200 ecosystems and regions. This implies that community assembly of oomycetes is not limited 201 by dispersal over countrywide distances. As a corollary, the remarkable differences in beta 202 We found that plant-associated hemibiotrophic oomycetes dominated in natural and semi-207 natural ecosystems, which thus constitute a reservoir of plant pathogens with the potential to 208 spread to neighbouring agrosystems. Pythium and Phytophthora, the two most infamous 209 destructive oomycete plant pathogens, together constituted 73% of the hemibiotrophs in 210 grassland and forest. Half of the OTUs and 67% of the hemibiotrophs were attributed to the 211 genus Pythium (Fig. 1), a result in line with previous oomycete community studies (Sapkota 212 and Nicolaisen 2015;Riit et al. 2016;Venter et al. 2017), and by Peronosporales being by far 213 the largest order in Oomycota, comprising more than 1,000 species (Thines 2014). The ability 214 to alternate between saprotrophy and parasitism likely favoured the hemibiotrophs over the 215 obligate biotrophic downy mildews were represented by Peronospora (3%, c. 400 described 217 species), which parasitizes a wide range of flowering plants, the Brassicaceae-infecting 10 The oomycete communities were more evenly distributed in grassland than in forest. 223 Grassland soil have a higher microbial biomass than forests (Dequiedt et al. 2011). In our 224 study, differences between grassland and forest, in terms of OTUs richness, alpha diversity 225 and evenness, were small. In a study comparing ecotypes across Wales, the relative 226 abundance of oomycetes compared to other Stramenopiles (algal Ochrophyta, mostly) 227 decreased from crops and grasslands to forests and bogs -the absolute numbers of oomycete 228 OTUs were not provided (George et al. 2019). This is in contrast with reports on Oomycetes 229 representing the dominant Stramenopiles in forest soils (Geisen et al. 2015). Although we 230 could not distinguish between soil-and air-borne species, the leaf-infecting white blister rusts 231 (mainly the genus Albugo), which are obligate angiosperm parasites (Beakes and Thines 232 2016), were nearly absent from your soil study, with only two OTUs attributed to Pustula. 233 Their absence in 600 soil samples confirmed that they mostly spread by air or possibly by 234 vertical transmission as seed endophytes (Ploch and Thines 2011). Our study, in assessing the 235 biogeography of oomycete pathogens at a national scale, also contributes to a better 236 understanding of the mechanisms of dispersal from natural reservoirs to agrosystems, and 237 perhaps could contribute to better reveal the evolutionary processes leading to the emergence 238 of new pathogens. 239

240
In forests, hemibiotrophs and obligate biotrophs showed opposite responses to a number of 241 environmental factors, suggesting different ecological requirements of both functional groups. 242 In summary, relative abundance of hemibiotrophs was decreased in nitrogen-poor, sandy and 243 acidic soils, which are the type of soils common in Schorfheide, planted with pines and oaks 244 being carbon-dependent, increased with C/N ratio in forest (Fig. S5F) and with organic C 268 content in grassland (Fig. S4E). Thus, in order to successfully enhance soil suppressiveness, it 269 is necessary to understand how particular management practices will differentially influence 270 each key component of biodiversity (Löbmann et al. 2016;Schlatter et al. 2017). Here, we 271 show that a more intensive forest management, as recorded in the Biodiversity Exploratories 12 (including estimation of harvested trunks, invasive tree species and cut dead wood) will not 273 increase the abundance of the hemibiotrophs. 274

275
More puzzling was the presence in soil of "water-moulds" (Saprolegniales, 21%), although 276 their constant and widespread occurrence in soils in addition to aquatic habitats has been 277 documented in the past (Dick and Newby 1961;Johnson et al. 2002). Some species of 278 Saprolegnia are known as fish or fish eggs parasites (saprolegniosis in fisheries), but are also 279 saprotrophs (Johnson et al. 2002). Some species of Atkinsiella are known for the damages 280 they inflict on marine crustaceans, but there are also parasites of terrestrial invertebrates 281 (Beakes and Thines 2016). The supposed predominance of the genus Aphanomyces (Beakes 282 and Thines 2016) is not fully confirmed by our study (6% of the OTUs). Species of the genus 283 can be saprotrophs or hemibiotrophs, so in our classification it is reported as "undetermined 284 lifestyle". 285

286
The distribution of oomycete communities is only partially explained in our study by 287 ecosystem, region or year of collection (only 29% of the variation explained). We 288 hypothesized that plant diversity and evenness and the presence of suitable plant hosts 289 (especially for oomycetes with a narrow host range) may be explanatory factors that were not 290 recorded: vegetation is a major determinant of the spatial distribution of soil pathogens 291 (Gómez-Aparicio et al. 2012). Microbial ecology studies suffer from a lack of systematic 292 basic information on community distribution which hampers predicting effects of 293 anthropogenic environmental changes. Providing a detailed baseline data on the occurrence of 294 oomycete taxa, the ecology and the distribution of their lifestyles is an important contribution 13 to the understanding of ecological processes and ecosystem functioning, a prerequisite for 296 subsequent analyses linking them to the distribution and diversity of potential plant hosts. 297 Soil cores from each plot were sieved (2 mm mesh size), mixed, homogenised and 311 immediately frozen for further analysis. Soil DNA was extracted from 400 mg of soil, 3-to 6-312 times, using the DNeasy PowerSoil Kit (Qiagen GmbH, Hilden, Germany) following the 313 manufacturer's protocol, to obtain a sufficient amount to be shared between research groups of 314 the Biodiversity Exploratories. to 534 bp. The first PCR was conducted with the primers S1777F (non-specific -5' 15 TCTTCATCGDTGTGCGAGC 3'). The second PCR was conducted with barcoded primers 345 S1786StraF (specific for Stramenopiles -5' GCGGAAGGATCATTACCAC 3') -and the 346 58SOomR as before. The barcodes consisted in eight-nucleotide-long sequences appended to 347 the 5'-ends of both the forward and the reverse primers, because tagging only one primer leads  (Table S1). 360

Materials and Methods
For the amplification, we incorporated 1 µl of 1:10 soil DNA template for the first PCR and 1 361 µl of the resulting amplicon as a template for a following semi-nested PCR. We employed the     Table S1. Combinations of barcodes used in this study, with the corresponding soil samples. 671 Code for sample names: AE= Alb, HE=Hainich, SE=Schorfheide. G=grassland sites -to be 672 replaced by F in forest sites (the same barcodes were applied since the grassland and forest 673 samples were amplified and sequenced separately). 674 Table S2. Database of the abundance of each OTU per sample. The taxonomic assignment 675 (supergroup, class, order, family, genus and species) is provided according to the best hit by 676 BLAST (PR2 database), with the % of similarity given. Functional traits (lifestyle and 677 substrate) were estimated for each genus following Table S3. 678 Table S3. References for the functional traits of the oomycetes genera identified in this study. 679