Content uploaded by Benoit Laurent
Author content
All content in this area was uploaded by Benoit Laurent on May 26, 2020
Content may be subject to copyright.
Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Benoit Laurent
Phytobiomes Journal
1
1A richer community of Botryosphaeriaceae within a less
2diverse community of fungal endophytes in grapevines than in
3adjacent forest trees revealed by a mixed metabarcoding
4strategy.
5 Benoit Laurent1*, Marylise Marchand1*, Emilie Chancerel1, Gilles Saint-Jean1, Xavier Capdevielle1,
6 Charlotte Poeydebat1, Anthony Bellée2, Gwenaëlle Comont2, Laure Villate1, Marie-Laure Desprez-
7 Loustau1
81 BIOGECO, Univ. Bordeaux, INRAE, F-33610 Cestas, France
92 UMR1065 SAVE, ISVV, INRAE, Villenave d’Ornon, France.
10 *Corresponding author: benoit.laurent@inrae.fr
11
12 Keywords: heterogeneous landscape, fungal community, latent pathogens, Diplodia, Neofusicoccum,
13 Botryosphaeria, management practice
14
15 Funding: COTE Cluster of Excellence Grant Number ANR-10-LABX-45
16
Page 1 of 53
Benoit Laurent
Phytobiomes Journal
2
17 Abstract:
18 Botryosphaeriaceae are a diverse group of endophytic fungi colonizing the inner tissue of
19 many woody species. As opportunist pathogens, they have been increasingly involved in
20 diebacks worldwide. Nonetheless, the diversity of Botryosphaeriaceae, especially in
21 asymptomatic plants, remains largely unknown. Using an innovative and mixed strategy of
22 metabarcoding, this study aims to investigate the diversity of the fungal endophyte community,
23 with a focus on Botryosphaeriaceae which colonize grapevine and adjacent oak and pine trees in
24 a French landscape. These data were used to test if the differentiation between hosts is more
25 important than geographical effects for shaping the Botryosphaeriaceae communities and that
26 similarity is higher between communities of grapevine and oak - both Angiosperms - than
27 between oak and pine trees. We revealed a high level of diversity in Botryosphaeriaceae fungi,
28 in both grapevines and forest trees, with a greater richness for grapevines. Contrasting results
29 were obtained for the endophytic community, which was more diverse in forest trees. Our
30 results support the hypothesis that host factors prevail on geographic effects to explain the
31 diversity of Botryosphaeriaceae at the studied spatial scale. However, the features of the agro-
32 eco-system, such as management practices, were suggested to be more important than
33 phylogeny to structure the fungal community. This highlights the importance of management
34 practices for the microbiome of plant trees.
35
Page 2 of 53
Benoit Laurent
Phytobiomes Journal
3
36 INTRODUCTION
37 The fungal family of the Botryosphaeriaceae (Ascomycota, Dothideomycetes,
38 Botryosphaeriales) includes common and widespread pathogens of woody hosts, most of which
39 belong to the Lasiodiplodia, Neofusicoccum, Dothiorella, Diplodia and Botryosphaeria genera
40 (Slippers et al. 2017). Recent surveys define 23 different genera and more than 180 species from
41 culture (Slippers et al. 2017; Yang et al. 2017). The majority of these species, if not all, share the
42 ability to live endophytically and asymptomatically in plant tissues for long periods of time
43 (Slippers and Wingfield 2007). They can also switch to pathogenic ‘behaviour’ usually following a
44 period of stress experienced by the host, provoking various symptoms, such as leaf spots, fruit
45 and root rot, dieback and cankers (Slippers and Wingfield 2007). Hence, members of the
46 Botryosphaeriaceae are often described as latent or opportunistic pathogens (Slippers and
47 Wingfield 2007; Yang et al. 2017).
48 Botryosphaeriaceae fungi are being increasingly reported to damage agents on woody
49 hosts, and several new species and pathogen-host associations have been steadily described
50 (Alves et al. 2013; Slippers et al. 2017; Zhou 2017; Brodde et al. 2018; Díaz et al. 2018; Li et al.
51 2018; Zlatković et al. 2018). For example, the frequency of grapevine trunk disease has been on
52 the rise worldwide (Bruez et al. 2012; Fontaine et al. 2016), with Diplodia seriata, D. mutila,
53 Neofusicoccum parvum, N. australe, N. luteum, Lasiodiplodia theobromae and Botryosphaeria
54 dothidea being the most frequent and pathogenic Botryosphaeriaceae species associated with
55 this disease (Urbez-Torres 2011; Chethana et al. 2016). Interestingly, some of the species are
56 not restricted to grapevine: B. dothidea has an important impact on pistachio production in
Page 3 of 53
Benoit Laurent
Phytobiomes Journal
4
57 China and USA, and also affects native trees, such as acacia in Africa (Marsberg et al. 2016).
58 Similarly, Diplodia seriata and D. mutila have been isolated from fruit rot and canker of apple
59 trees in Northern and Southern America and are becoming new threats for apple production
60 (Úrbez-Torres et al. 2016; Crespo et al. 2018; Díaz et al. 2019). These two Diplodia species have
61 also been described as causal agents of olive and nut tree dieback in the USA (Úrbez-Torres et
62 al. 2013; Moral, Morgan and Michailides 2019) and Europe (Kaliterna et al. 2012), as well as
63 many other native and non-native angiosperm and gymnosperm trees (Alves et al. 2013; Phillips
64 et al. 2013; Zlatković et al. 2018). Neofusiccocum luteum, N. parvum and N. australe have been
65 frequently isolated from coniferous trees (Alves et al. 2013), and Lasiodiplodia theobromae has
66 been isolated from 252 different host genera, including both angiosperms and gymnosperms
67 (U.S. National Fungus Collections Fungus-Host Database, checked the 03rd of April, 2019).
68 Diplodia sapinea has never been found on grapevine, but is probably one of the most damaging
69 pathogens of conifers. First reported as an agent of pine damage in the southern hemisphere in
70 the early 20th century, it has been more recently associated to many outbreaks in Europe and
71 worldwide (Stanosz et al. 2002; Bihon et al. 2011; Fabre et al. 2011; Luchi et al. 2014;
72 Decourcelle et al. 2015; Paez and Smith 2018; Brodde et al. 2019). In only rare cases, D. sapinea
73 has been isolated from angiosperm species (Zlatković et al. 2017).
74 Climate change is likely to be a major factor linked to the emergence of diseases
75 associated with Botryosphaeriaceae fungi, because it increases the intensity and frequency of
76 stress experienced by trees, and it creates better growing conditions for thermophilic species
77 (Desprez-Loustau et al. 2006; Sturrock et al. 2011). Additionally, the occurrence of
Page 4 of 53
Benoit Laurent
Phytobiomes Journal
5
78 Botryosphaeriaceae fungi in asymptomatic plants - including seeds - explain why these
79 pathogens may be unnoticed during trade, favouring their worldwide dissemination and
80 diffusion through nursery trade (Decourcelle et al. 2015; Marsberg et al. 2016). The
81 opportunistic “behaviour” of the Botryosphaeriaceae may also explain why the role of
82 Botryosphaeriaceae has been overlooked in several decline syndromes, especially in grapevine,
83 since the same species were detected or isolated from both healthy and declining plants (Úrbez-
84 Torres et al. 2006; Bruez et al. 2014). The broad host range of at least some Botryosphaeriaceae
85 species (although host range is limited to the level of knowledge about the species) suggests
86 that there is no barrier to disease transmission between native and non-native plant species, or
87 cultivated/urban and natural settings (Burgess et al. 2006; Stanosz et al. 2007; Sakalidis et al.
88 2011; Begoude Boyogueno et al. 2012; Marsberg et al. 2016; Mehl and Roux 2017; Zlatković et
89 al. 2018). For example, no gene flow restriction was observed between Botryosphaeria australis
90 populations from Eucalyptus globulus plantations and from native eucalypt forests in Australia
91 (Burgess et al. 2006); nor was it found between Lasiodiplodia species populations in Congo from
92 Theobroma cacao, the cultivated cacao tree, and from the native Terminalia spp timber trees
93 (Begoude Boyogueno et al. 2012).
94 This ability to perform host shifts or to infect many different host species may explain
95 the invasive success of Botryosphaeriaceae fungi, enabling them to spread in heterogeneous
96 landscapes. In this study, we focus on a mosaic landscape formed by vineyards and forests in
97 Nouvelle-Aquitaine, France. This region is the second largest wine-growing region in France with
98 216,000 ha of vineyards (especially Bordeaux vineyards), and it is among the most forested in
Page 5 of 53
Benoit Laurent
Phytobiomes Journal
6
99 Europe with 2.8 million ha of forest, almost 80% of which comprises maritime pine (Pinus
100 pinaster) plantations. Several surveys have reported Botryosphaeriaceae fungi on various hosts
101 in the region, especially Diplodia sapinea from pine trees; Diplodia corticola, Diplodia mutila, N.
102 parvum and B. dothidea in rarer cases from diverse hardwoods (Fabre et al., 2011, French Forest
103 Health service database); and N. parvum, B. dothidea, D. mutila, D. seriata, D. intermedia,
104 Lasiodiplodia viticola and Spencermartinsia viticola from grapevine (Larignon et al. 2001;
105 Comont et al. 2016; Nivault 2017).
106 A previous study in the same region used metabarcoding to investigate the fungal
107 communities present in the air and leaves within vineyards and adjacent forests, showing strong
108 differentiation between grapevine and forest tree fungal communities. Some results have
109 further suggested that the major driver of differentiation between vineyards and forest patches
110 was not dispersal but rather selective pressures associated with host plant, microclimate and
111 management practices (Fort et al. 2016). How these findings, obtained from three neighbouring
112 sites and for the total foliar fungal communities, can be generalized to a larger spatial scale and
113 all groups of fungi is questionable. Our study aims to address this question by analysing
114 Botryosphaeriaceae diversity in comparison to the whole fungal community within a
115 heterogeneous landscape consisting of vineyard and forest patches in South-West France.
116 For the purpose of this study, we developed and tested an innovative DNA
117 metabarcoding strategy that produced one dataset using Botryosphaeriaceae-specific primers
118 and a second dataset using the “universal” fungal barcode (Schoch et al. 2012). The diversity of
119 Botryosphaeriaceae fungi was investigated in adjacent vineyards and forest plots by sampling
Page 6 of 53
Benoit Laurent
Phytobiomes Journal
7
120 both healthy and symptomatic twigs from grapevines, pines and oaks, the three main species in
121 the ecosystems, with high environmental and economic significance. Our first objective was to
122 assess the diversity of Botryosphaeriaceae in grapevines and forests trees at the regional scale,
123 in terms of species richness and species composition. These data could then be used to test
124 several hypotheses: i) differentiation between hosts is more important than geographic effects,
125 both for the whole fungal community and for Botryosphaeriaceae, as observed by Fort et al.
126 (2016), and ii) in this case, similarity is higher between fungal communities of grapevine and oak
127 -both Angiosperms - than between oak and pine trees, according to a phylogenetic signal in the
128 host range.
129 Materials and Methods
130 Sampling
131 Samples were collected from 28 sites between the 20th July and the 4th August 2017, in an area
132 in South West France spanning ~14,000 km² (Table 1, Supplementary file 1), with a distance
133 between sites ranging from ~5 to ~220 km and belonging to different appellations. Appellations
134 are delimitated geographical regions with specific geological, climate and grapevine growing
135 conditions. Between three to five sites were sampled for each appellation, for a total of seven
136 different appellations (Supplementary file 1). Sites were selected based on the occurrence of
137 adjacent forests and vineyards. Sites with approximatively equivalent surface area of forests and
138 vineyards were preferred, as well as those with forest plots composed of both pine and oak
139 trees. Within each vineyard plot, lignified twigs from the previous growing season of about 0.5
Page 7 of 53
Benoit Laurent
Phytobiomes Journal
8
140 to 2 cm in diameter were cut using sterilized pruning shears from three symptomatic and three
141 asymptomatic grapevines, located randomly in the patch (distance between vines of 3 to ~50
142 meters). Symptomatic grapevines were chosen on the basis of foliar symptoms linked to
143 “grapevine trunk disease”, i.e., dark-red leaf discoloration causing eventual dry out of the
144 tissues between the veins and the margin of the leaf or when the plant had undergone a severe
145 form of the disease and was dying (Larignon et al. 2001). Lignified twigs from 0.5 to 1.5 cm in
146 diameter from the previous growing season were cut using sterilized pruning shears for
147 grapevine. Pine trees and oak trees were sampled from neighbouring forest patches, at a
148 distance ranging from ~10 meters to few hundred meters from the vineyard plot, either at the
149 edge or in more central parts of the plot. Lignified twigs from 1 to 3 cm in diameter were cut
150 from low lying branches of the crowns from 2 to 8 meters high using sterilized pole pruner.
151 Symptomatic trees were found at lower frequency than for grapevine and when found,
152 symptomatic trees exhibited cankers, resinosis, shoot blight and/or dieback (Supplementary file
153 1). Grapevine, oak and pine twig samples were placed in individual and sterile plastic bags. In
154 total, 352 twigs were collected and stored at -20 °C until DNA extraction. For each sample, the
155 presence of brownish/dark necrosis in the internal tissue of the living twig, and the symptomatic
156 status of the sampled plant were noted.
157 DNA extraction and PCR protocols
158 Sample preparation was conducted inside a confined laboratory dedicated to environmental
159 DNA preparation. Twigs were superficially sterilized using 70% ethanol. The dead and living bark
160 of each sample was removed using a sterile scalpel and ~3 mm long wood pieces of ~8 mm
Page 8 of 53
Benoit Laurent
Phytobiomes Journal
9
161 diameter were collected in 96-well plates using sterile secateurs. Wood pieces were then
162 lyophilised overnight and ground with sterile metal beads (2 per well) using Retsch MM400 until
163 fine powder was obtained. Total DNA was then extracted from wood powder using Invisorb®
164 DNA Plant HTS 96 Kit/ C (Stratec) following the manufacturer’s instructions. The wood fungal
165 community in these samples was studied by amplifying a partial sequence of the internal
166 transcribed spacer 1 (ITS1) using the fungal specific ITS1F (Gardes and Bruns 1993) and ITS2
167 primers (White et al. 1990; Supplementary file 2a). The Botryosphaeriaceae community was
168 specifically targeted by amplifying 348 bp of the 28s large subunit of the ribosomal RNA gene,
169 referred hereafter as LSU, using the BotSp_LSU_F and BotSp_LSU_R primers, designed for the
170 purpose of this study using Geneious® 10.2.2 (Supplementary file 2a). These newly designed
171 primers were tested against known Botryosphaeriaceae species, as well as other fungi isolated
172 from grapevine, pine and oak tissues in a preliminary experiment, in order to validate their
173 specificity (Supplementary file 2a).
174 Library construction was done using a two-step PCR amplification. The first PCR amplification
175 was carried out using locus specific primers, preceded by particular molecular sequences
176 (TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG for the forward primers and
177 TCTCGTGGGCTCGGAGATGTGTATAAGAGACAG for the reverse primers). In order to limit the
178 stochastic biases introduced by the PCR (Zinger et al. 2019), amplification was replicated twice
179 in a final volume of 20 µl for each DNA sample. Negative controls for 96-well plate sample
180 preparation, DNA extraction and PCR were included. The PCR mix was composed of 4 µl of 5X
181 Hot Firepol® Blend Master Mix (Solis Biodyne), 0.4 µl of each primer at 10 µM, 5 µl of template
Page 9 of 53
Benoit Laurent
Phytobiomes Journal
10
182 DNA at ~20 ng/µl and 10.2 µl of pure-grade water for a final reaction volume of 20 µl. PCR
183 conditions consisted of a denaturation step at 95 °C for 15 minutes, followed by 40 cycles of
184 denaturation at 95 °C for 20 sec, primer hybridization at 55 °C for 30 sec, and elongation at 72 °C
185 for 30 sec. Duplicates of PCR products from each primer set were pooled and a second
186 amplification was performed to add the indexes (1 different index per sample) and the Illumina
187 adapters. This PCR amplification was carried out in a final volume of 20 μl. Each PCR mix
188 included 4 μl of PCR1 products, 4 µl of 5X Hot Firepol® Blend Master Mix (Solis Biodyne), 4 μl of
189 S index 2 µM, and 4 μl of N index 2 µM (Nextera Set2 indexes – Illumina), and was
190 supplemented with ultra-pure water. The PCR reaction conditions are as follows: denaturation
191 step at 95 °C for 15 min followed by 12 cycles comprising denaturation at 95 °C for 15 s,
192 hybridization of the primers at 60 °C for 1 min, and elongation at 72 °C for 1 min followed by a
193 final extension at 72 °C for 10 min. ITS1 and LSU libraries were then pooled separately in
194 equivolume and purified with CLEAN NA (Proteigene) beads (1.8X ratio). The ITS1 and LSU final
195 pools were then checked on Tapestation (Agilent Technologies) to verify the size of the libraries
196 and quantified in qPCR with the Kapa libraries quantization kit (Roche). The ITS1 and LSU pools
197 were sequenced in different Illumina MiSeq runs, using MiSeq Sequencing Reagent Kit v3 (600
198 cycles) and v2 (500 cycles) respectively. Reads produced in this analysis are retrievable from the
199 NCBI database of the BioProject accession n° PRJNA508496.
200 Mock community preparation and sequencing
201 In order to evaluate the family-targeted metabarcoding approach, DNA from two mock
202 communities was tested. The first mock community was composed of six Botryosphaeriaceae
Page 10 of 53
Benoit Laurent
Phytobiomes Journal
11
203 species: Diplodia mutila, Diplodia seriata, Diplodia corticola, Diplodia sapinea, Dothiorella
204 viticola and Neofusicoccum parvum. The second also included eight species, including two
205 Botryosphaeriaceae: Diplodia intermedia, Botryosphaeria dothidea, Epicoccum nigrum, Eutypa
206 lata, Alternaria sp., Aureobasidium pullulans, Cladosporium sp. and Phoma sp. (Supplementary
207 file 2b). For each mock community, DNA extracted from all species in pure culture was mixed
208 with 1 µl of DNA concentrated at 0.8 ng/µl for each species. Each DNA mixing was done in
209 duplicate. All DNA mixtures were amplified following the procedure described above for both
210 Botryosphaeriaceae family using BotSp_LSU_F and BotSp_LSU_R primers and for fungal
211 community using ITS1F and ITS2 primers. Amplification duplicates of each DNA mixture were
212 pooled prior to sequencing, giving a total of four datasets: DNA mixtures for the two mock
213 communities, amplified either with the ITS marker or with the Botryosphaeriaceae specific LSU
214 marker. A first PCR was done using the same locus specific primers and using similar PCR
215 conditions mentioned in “DNA extraction and PCR protocols”. PCR products were purified with
216 AMPure beads (1.8 X ratio) and a second PCR was done with 4μl of purified PCR1 products, 12 μl
217 of 2X KAPA HiFi HotStart Ready Mix, 4 μl of S index, 4 μl of N index (Nextera Set2 index -
218 Illumina) and 4.5 µl of ultra-pure water. The PCR conditions comprised: a denaturation step at
219 95 °C for 3 min followed by 12 cycles comprising denaturation at 95 °C for 30 s; hybridization of
220 the primers at 60 °C for 30 s; elongation at 72 °C for 30 s followed by a final extension at 72 °C
221 for 5 min. The purification of the libraries was carried out using AMPUre magnetic beads (ratio
222 of 0.9 X). Libraries are quantified and standardised at 10 ng/μl before pooling in equivolume.
223 The final pool was then checked on Tapestation to verify the size of the libraries and quantified
Page 11 of 53
Benoit Laurent
Phytobiomes Journal
12
224 in qPCR (LC480; Roche) with the KAPA Library Quantification Kit (Roche). Sequencing was
225 conducted on Illumina MiSeq using Nano kit v2 producing paired 250 nucleotide long reads.
226 Database construction
227 A dedicated database was constructed composed of fungal ITS sequences retrieved from the
228 general release for fungi of the UNITE Community database (UNITE general FASTA release for
229 Fungi Version 18.11.2018. UNITE Community) including 71,042 “fungal species hypotheses” as
230 defined by the UNITE community (i.e., terminal fungal taxa represented by at least one ITS
231 sequence at a distance of 1.5% of any other sequences, among which 97% singletons). Of the
232 71,042 entries, only 66,647 which provide information at least at the phylum level were kept.
233 Our database also included LSU sequences from diverse fungi, retrieved from the NCBI
234 Reference Sequences (RefSeq; O’Leary et al., 2016) using the search criteria txid4751[Organism]
235 (i.e.,fungi) AND "large subunit ribosomal" [All Fields] NOT "whole genome" [All Fields] and only
236 keeping sequences with sizes ranging from 400 bp to 5000 bp (n=611 sequences). Furthermore,
237 LSU and ITS sequences issued from a recent phylogenetic reappraisal of the Botryosphaeriales
238 order were added, including 18 genera from the Botryosphaeriaceae family and 7 other genera
239 from 6 other families (Yang et al. 2017). Of these, the 159 sequences corresponding to the LSU
240 of Botryosphaeriales species were trimmed using Geneious® 10.2.2 software in order to match
241 the regions amplified by the Bot_sp LSU primer pair. Redundant sequences were removed
242 (n=84). When similar sequences were associated with different taxa, the different taxon names
243 were kept in the sequence header. The prefix Bot_LSU_TaxID_[1:75] was also added to the
244 header for tracking the sequence in the whole database. For the 472 ITS sequences retrieved
Page 12 of 53
Benoit Laurent
Phytobiomes Journal
13
245 from Yang et al. (2017), only the ITS2 primer was used for trimming, because the ITS1F primer
246 was not systematically found in the sequences. Similarly, only the 145 sequences that were
247 unique after trimming were added to the database with proper taxa in the header, including the
248 prefix Bot_ITS_TaxID_[1:145]. Overall, the final fungal database constructed for this study is
249 composed of 67,478 entries, later referred to as “Database Sequences”. This fasta database was
250 deposited at the INRA data repository and is accessible at https://doi.org/10.15454/JEOT4O.
251 Sequence processing and taxonomic assignment
252 The processing of metabarcoding data was inspired by the Si5 bioinformatic strategy as
253 proposed by Pauvert et al. 2019, which they claim exhibits the best balance between sensitivity
254 and accuracy parameters among the 360 bioinformatic pipelines that were tested. More
255 precisely, only the forward reads produced by the sequencer were used and filtered using
256 DADA2 and the filterAndTrim option (Callahan et al. 2016) with the following parameters:
257 minimal read length=100 nucleotides, sequences with more than maximum number of N
258 allowed in the reads = 0, maximal expected errors = 1, discard reads that match the phiX
259 genome (viral genome used to test contamination) = TRUE. Sequences were dereplicated and
260 Amplified Sequence Variants (ASVs) were constructed using the dada option after error rate
261 estimation. Chimeric sequences were filtered out and the remaining ASVs were blasted against
262 our database using blastn (v2.6.0+, Camacho et al. 2009). For each ASV, the blast results were
263 sorted according to bit-score and only the best hit, or several best hits if sharing the same bit-
264 score, was/were kept. After manual inspections, ASVs with no hit or having a bit-score inferior
265 to 300 for ITS and 350 for LSU were discarded. For the LSU dataset, ASVs showing similar bit-
Page 13 of 53
Benoit Laurent
Phytobiomes Journal
14
266 scores for the same Database Sequence(s) were pooled and their read counts were added. For
267 the ITS dataset, ASVs assigned to the same genus after blast were pooled and their read counts
268 were added together. The taxonomic units resulting from these operations will be later referred
269 to as “clades” for both ITS and LSU analyses. Reads were occasionally obtained in the negative
270 controls. These reads could arise from common laboratory contamination or minor cross
271 contamination between samples. Because most reads were found in low numbers in negative
272 samples and they were assigned to abundant and frequent taxa in the field samples
273 (Supplementary file 2c), we suggest that those sequences were a result of minor instances of
274 cross-contamination between samples of highly abundant species, and they were therefore not
275 removed in downstream analysis (Supplementary file 2c). The online version of FUNGuild
276 (Nguyen et al. 2016, http://www.stbates.org/guilds/app.php) was used to determine the trophic
277 mode and the ecological guild of the ASVs identified after ITS amplification.
278 Diversity indices and statistical analysis
279 Two indicators describing fungal communities were used. The first one consisted in the number
280 of clades, or clade richness, assessed with a resolution at the genus or upper taxonomic level for
281 the ITS amplicons (i.e., the wood fungal community), and at the species or genus level for the
282 LSU amplicons (i.e., the Botryosphaeriaceae community). The second indicator of diversity was
283 the number of ASV, or ASV richness, which accounted for putative different species, populations
284 or genotypes within a clade. Generalised linear mixed-effects models (GLMM) were used to test
285 for the effects of small geographical region (‘appellation’), host (grapevine, oak or pine) and
286 wood symptomatic status (dead/necrosis or no necrosis) on clade richness and ASV richness
Page 14 of 53
Benoit Laurent
Phytobiomes Journal
15
287 (Poisson distribution), as well as on the frequency of the Botryosphaeriaceae clades (Binomial
288 distribution). In the clade and ASV richness models, we added the ‘host x wood symptomatic
289 status’ and ‘host x appellation’ interactions as fixed effects to test whether the effects of
290 symptomatic status and appellation on the fungi community depended on the host. The
291 sampling site nested within the appellation was included in the models as a random factor to
292 account for observations clustering. Significance of fixed effects was tested using Type II ANOVA
293 (in R environment) with alpha set at 0.01. We evaluated the fit of the model to the data by
294 calculating the percentage of variance explained by fixed effects (marginal coefficient of
295 determination R²m calculated using the delta method) and by fixed plus random effects
296 (conditional coefficient of determination R²c calculated using the delta method) (Nakagawa and
297 Schielzeth, 2013). Finally, post-hoc multiple comparisons of means were performed for each
298 significant factor using Tukey’s Honestly Significant Difference method. All analyses were done
299 using the R programming language (version 3.5.3; R Core Development Team, 2013) and lme4
300 (Bates et al., 2014), MuMIn (Barton, 2018) and multcomp (Hothorn et al., 2008) packages.
301 Results
302 Evaluation of the targeted metabarcoding approach with mock communities
303 The results of sequencing the two mock communities using the fungal ITS and
304 Botryosphaeriaceae-targeted LSU are shown in Table 2. A total of 24 and 13 ASVs corresponding
305 to 7 and 5 clades (with more than 4 reads) and belonging to Botryosphaeriaceae were identified
306 after ITS and LSU sequencing of mock 1 respectively. With the LSU marker, five among the six
Page 15 of 53
Benoit Laurent
Phytobiomes Journal
16
307 Botryosphaeriaceae species included in mock 1 had a match with at least one clade
308 corresponding to closely related species in the same genus and including the target species
309 (Table 2). N. parvum was assigned to N. australe and only three species had an exact match with
310 ITS. Although the DNA of each species was set to equimolar concentration in the DNA mixture,
311 important variations in the number of reads between taxa were observed (e.g., 168 reads for
312 Lasidiplodia vitis versus 7,094 for Neofusicoccum sp. for the LSU marker (Table 2).
313 Mock 2 was composed of eight fungal species with two Botryosphaeriaceae species (Diplodia
314 intermedia and Botryosphaeriaceae dothidea). A total of 9 and 12 ASVs corresponding to 3 and 7
315 clades (with more than 4 reads) were identified after LSU and ITS sequencing respectively (Table
316 2). Five of the species present in this mock were identified by the ITS marker. Eutypa lata had no
317 match. A misidentification occurred at the species level for Aureobasidium pullulans and
318 Diplodia intermedia (Table 2). With LSU, most reads were assigned to a clade of Diplodia species
319 including the target species D. intermedia. Many less reads corresponded to a second clade
320 corresponding to Mycosphaerellaceae.
321 In view of these results, only the LSU marker was used to describe the Botryosphaeriaceae
322 community in the subsequent analyses, using the defined clades as taxonomic units. The ITS
323 marker was used to describe the whole fungal community, only taking into account the genus
324 level to avoid any misidentifications at the species level. Moreover, read counts as an indicator
325 of species abundance were not used for this study due to the inconsistency observed in the
326 mock analysis, and indicators of diversity using abundancy were avoided.
Page 16 of 53
Benoit Laurent
Phytobiomes Journal
17
327 Endophytes and Botryosphaeriaceae species identified from the metabarcoding sequencing
328 MiSeq sequencing of ITS and LSU amplicons from the 353 environmental samples and the 6
329 negative controls produced 24,358,756 and 14,657,024 reads respectively. From these,
330 12,435,073 ITS reads passed the quality filter and were grouped in 6,252 amplicon sequence
331 variants (ASVs, Table 3); similarly, 10,278,555 LSU reads passed the quality filter and were
332 grouped into 2,500 ASVs (Table 3). Some ASVs gave no hit after blast against the custom
333 database: 164 ASVs in 8,341 reads (0.07% of initial read number) for the ITS sequences (Table 3)
334 and 1,310 ASVs in 1,106,150 reads (10.8% of initial read number) for the LSU sequences (Table
335 3). Among ITS sequences, no assignment was found after blasting the nucleotide collection of
336 the NCBI for 122 out of 164 ASVs, possibly corresponding to chimeric sequences or unknown
337 sequences. For the 42 remaining ASVs (for 3,440 reads) that matched the NCBI nucleotide
338 collection, a total of 27 different sequences from the NCBI were identified corresponding to 11
339 different taxa: 7 plant taxa, 3 fungal taxa and 1 bacterial taxon. Sequences corresponding to
340 Vitis vinifera were the most frequent and abundant, with 1,608 reads in 52 different samples,
341 and were all found in grapevine twig samples. Regarding the LSU sequences giving no hit after
342 blast against the custom database, only 557 out of 1310 ASVs for 846,742 reads matched 183
343 different sequences and 36 taxa from the NCBI nucleotide collection. The 753 remaining ASVs
344 gave no hit on the NCBI nucleotide collection, probably corresponding to chimeric ASVs or ASVs
345 from uncharacterised species. The most frequent non-Botryosphaeriales fungal taxa amplified
346 with LSU primers belonged to Pleosporales (76.0% of samples), Mycosphaerellaceae (51.1%),
347 Ciborinia (48.6%) and Hysterographium (32.4%). Non-matching LSU sequences in the
Page 17 of 53
Benoit Laurent
Phytobiomes Journal
18
348 Botryosphaeriales order were discarded (Table 3). Finally, 37.2% of ITS sequences for 18.8% of
349 initial read number (Table 3), as well as 43.7% of LSU sequences for 49.7% of initial read number
350 (Table 3) were filtered out based on low blast bit-score. Details of ITS and LSU ASVs obtained
351 throughout the pipeline are available in Supplementary file 3.
352 The final dataset for the ITS amplicon sequencing is composed of 3,763 ASVs for 10,082,422
353 reads, grouped in 565 different taxonomic clades and corresponding to genus or upper
354 taxonomic levels. Of these, 75.2% of ASVs belonged to Ascomycota and 24.6% belonged to
355 Basidiomycota. Probable trophic mode could be determined for 394 out of the 565 clades,
356 revealing a majority of pathotroph or saprotroph fungi (77.7% of taxon with determined
357 probable trophic mode, Supplementary file 4). The most frequently identified genera were
358 Cladosporium (78.1% of samples), Alternaria (68.2%), Aureobasidium (58.0%), Devriesia (56.3%),
359 Mycosphaerella (54.5%), Vishniacozyma (54.3%) and Epicoccum (54.0%). The twenty most
360 frequently identified clades, as well as their frequencies according to host tissue, are shown in
361 Supplementary file 4.
362 The final dataset for the LSU amplicon sequencing is composed of 208 ASVs for 4,063,783 reads
363 (Table 3). The 208 ASVs matched 22 clades, each of them corresponding to a group of species
364 and generally in the same genus. Seven clades, encompassing 4 genera - Neofusicoccum,
365 Diplodia, Botryosphaeria, Dothiorella - were identified at a frequency greater than 5% of the
366 samples (Table 4). The most frequent clade corresponds to Neofusicoccum spp. (31.5% of the
367 samples, Table 4), Diplodia sp. (29.8%, corresponding species: D. mutila, D. scrobiculata, D.
368 seriata, D. conspersa), Botryosphaeria dothidea (22.73%), another clade of Diplodia spp.
Page 18 of 53
Benoit Laurent
Phytobiomes Journal
19
369 (19.03%, corresponding species: D. corticola, D. gallae, D. intermedia, D. sapinea, D. seriata),
370 and Diplodia mutila or pyri (11.65%). Fifteen rare clades in Botryosphaeriales were identified,
371 with frequencies ranging from 0.28% to 0.85% of the samples (Table 4).
372 Factors influencing fungal richness
373 Average clade richness per sample and per site was of 23.5 and 133.6 respectively for the total
374 wood fungal community, and 1.4 and 6.8 respectively for the Botryosphaeriaceae community
375 (Table 5). Average ASV richness per sample and per site was of 36.3 and 327.6 respectively for
376 all fungi and 1.9 and 14.0 respectively for the Botryosphaeriaceae community (Table 5). The
377 total number of ASVs was higher in grapevine samples (n=138, Table 3) compared to pine and
378 oak samples (n=54 and n=33 respectively, Table 3). ANOVA revealed a significant effect of host
379 on the clade richness for both the wood fungal community and the Botryosphaeriaceae (p-value
380 = < 2.2e-16 and p-value < 1.02e-05 respectively, Table 6). A significantly greater fungal clade
381 richness was observed in Quercus and Pinus twigs than in Vitis twigs (p-value < 1e-05 for both
382 comparisons, Figure 1b), and inversely, a greater richness of Botryopshaeriaceae was observed
383 within grapevine twigs compared to pine and oak twigs (p-value = 0.18 and p-value < 1e-04,
384 Figure 1a). No significant effect was detected for symptomatic status (presence/absence of
385 wood necrosis) nor vineyard appellation on the Botryosphaeriaceae clade richness. Additionally,
386 significant effects on the total fungal clade richness were observed for vineyard appellations (p-
387 value = 1.28e-04), symptoms (p-value = 1.85e-03), and for the interaction of host and
388 appellations. More precisely, a greater richness in asymptomatic samples than symptomatic
389 ones was observed, as well as a reduced richness in Graves and Vin des Sables compared to
Page 19 of 53
Benoit Laurent
Phytobiomes Journal
20
390 other regions. Similar patterns than those observed for clade richness were observed for ASV
391 richness (Table 6). However, a significant effect of the interaction of host and appellations was
392 demonstrated on the Botryosphaeriaceae ASV richness (p-value = 3.88e-13). The variance
393 explained by the models was of 12.9% for the Botryosphaeriaceae clade richness and 47.2% for
394 the fungal clade richness (Table 6); 31.6% for Botryosphaeriaceae ASV richness and 69.9% for
395 fungal ASV richness (Table 6).
396 Diversity of the Botryosphaeriaceae communities across host plants.
397 Out of the twenty-one clades identified with the LSU marker targeting Botryosphaeriaceae,
398 eight were found in the three different hosts, including the seven major clades identified at a
399 frequency greater than 5% of the samples (Table 4), in addition to one rare clade (belonging to
400 Neofusicoccum sp., Figure 2). Twelve clades were host specific: four to grapevine, four to oak
401 and four to pine. Plant-specific clades always corresponded to rare ones (identified in 1 to 3
402 samples); for example, Melanops sp. and Dothiorella dulcispinae or Do. oblonga in oak twigs,
403 Pseudofusicoccum ardesiacum or P. kimberleyense in grapevine twigs, and Tiarosporella
404 africana/tritici in pine twigs (Figure 2b). Another rare clade was shared between pine and oak
405 (Aplosporella/Bagnisiella sp.). In contrast to LSU clade distribution, a large majority of the 208
406 ASV (n=194, i.e., 94%) were host specific, with more than 60% (n=128) exclusively found in
407 grapevine samples, 20% (n=42) in pine samples and 12% (n=24) in oak samples (Figure 2). Seven
408 ASV were found in all three hosts, three others were shared between grapevine and pine, and
409 two between pine and oak. All shared ASV, as well as the majority of plant-specific ASV (89% for
Page 20 of 53
Benoit Laurent
Phytobiomes Journal
21
410 grapevine samples, 58% for oak samples and 79% for pine samples) belonged to the seven
411 major LSU clades (Figure 2).
412 Effect of host and geographic location on Botryosphaeriaceae frequency
413 Quantitative differences of occurrence between hosts were investigated for the seven major
414 Botryosphaeriaceae LSU clades, revealing a significant host effect for four out of the seven
415 clades: Diplodia sp. in clade 2, Neofusicoccum sp. in clade 1, and Botryosphaeria dothidea and
416 Diplodia sp. in clade 4. More precisely, significant higher frequencies were observed in
417 grapevine samples for Diplodia sp. in clade 2 (58.2% on average per site) compared to oak
418 samples (17.1%, p-value < 1e-6) and pine samples (18.3%, p-value = 1.01e-5). Similar patterns
419 were observed for Neofusicoccum sp. in clade 1 (an average of 50.6% of grapevine samples per
420 site compared to 21.7% and 26.9% for oak and pine respectively, significant with oak only) and
421 for Botryosphaeria dothidea (an average of 34.5% of grapevine samples per site compared to
422 19.2% and 14.2% for oak and pine respectively). The reverse was observed for the Diplodia sp. in
423 clade 4, with higher frequencies in oak and pine twigs (23.2% and 26.5% respectively) compared
424 to grapevine twigs (6.5%, significant with oak only). Overall, Diplodia sp. in clade 2 (96% of the
425 sites), Neofusicoccum sp. in clade 1 (93% of the sites) and B. dothidea (93% of the sites) were
426 the most frequent Botryosphaeriaceae. Diplodia sp. in clade 4 was found in 89% of the forest
427 patches where pine trees were sampled and 77% of the forest patches where oak trees were
428 sampled (Figure 3b). In contrast to the host effect, no geographic effect could be shown in the
429 frequency of the major Botryosphaeriaceae clades and similar frequencies were obtained in the
430 different appellations (Figure 4).
Page 21 of 53
Benoit Laurent
Phytobiomes Journal
22
431 Discussion
432 Botryosphaeriaceae fungi have been increasingly recognised as causal agents of diseases in
433 many different hosts and environments (Slippers et al. 2017). The wide host range of at least
434 some species in this group has been emphasised (De Wet et al. 2008; Phillips et al. 2013;
435 Slippers et al. 2017). Several species also show very wide geographic distributions (Phillips et al.
436 2013), although the natural dispersal of Botryosphaeriaceae fungi is generally assumed to
437 mainly occur via rain-splashing of conidia - thus at rather short distances - at least for species
438 with no sexual form (Slippers and Wingfield 2007; Úrbez-Torres et al. 2010; Baskarathevan et al.
439 2013; Silva et al. 2019). However, the frequent movement of these pathogens as a result of
440 trade of plant products at both global and regional scales has been emphasised (Bihon et al.
441 2011; Sakalidis et al. 2013; Slippers et al. 2017). The endophytic behaviour of these fungi, i.e.
442 their ability to colonize plant tissues without causing symptoms, is probably an important factor
443 explaining their unintentional transport in contaminated commodities and, more generally, the
444 fact that their biogeographic distribution is still not well known.
445 Our study aimed to give a better understanding of the factors shaping the Botryosphaeriaceae
446 communities at the landscape level. A striking result of our study is that the diversity patterns of
447 Botryosphaeriaceae fungi were much more structured by the ecological compartment (forest vs.
448 vineyard) than by geographic factors in South western France, at a scale of several tens of
449 kilometers. Indeed, contrary to expectations based on total fungal diversity, the richness of
450 Botryosphaeriaceae fungi was higher in grapevine twigs than in forest trees and the patterns of
451 richness and frequency of the major clades were more similar between Pinus and Quercus than
Page 22 of 53
Benoit Laurent
Phytobiomes Journal
23
452 between Vitis and Quercus (both Angiosperms). This result does not follow the phylogenetic
453 signal hypothesis, stating that the likelihood that a pathogen can infect two plant species
454 decreases continuously with the phylogenetic distance between the plants (Gilbert and Webb
455 2007).
456 However, a main limitation of the metabarcoding approach was the difficulty in obtaining a
457 resolution at species level, despite some improvements having been made to the method.
458 Finally, our study provides a first insight into Botryosphaeriaceae diversity in pines and oaks in
459 France, showing their widespread occurrence as endophytic fungi in twigs. We discuss these
460 main findings below.
461 A critical overview of the metabarcoding approach targeting Botryosphaeriaceae
462 Generally, metabarcoding analyses take advantage of the “universality” of the method to
463 describe the most exhaustive snapshots of large microbial communities and to study how these
464 communities vary in space and time. Additionally, recent examples suggest a new and increasing
465 use of the metabarcoding approach to obtain occurrence data on targeted groups or species
466 (Zinger et al. 2019). For example, by mining datasets constructed from the universal fungal ITS
467 barcode sequencing, Da Lio et al. (2018) were able to identify several Colletotrichum species
468 associated with the walnut anthracnose in France, and Bérubé et al. (2018) were able to detect
469 the presence of Diplodia corticola in Canada by studying air samples. It is also possible to
470 directly target the group of interest using dedicated primer pairs, giving access to more sensitive
471 and reliable detection, as recently done by Cobo-Díaz et al. (2019) when investigating a
472 Fusarium community in soil and plant samples.
Page 23 of 53
Benoit Laurent
Phytobiomes Journal
24
473 Our goal was to develop such primer pairs to target the Botryosphaeriaceae family. In 2017,
474 Yang et al. suggested that multi-locus phylogeny using the LSU and RPB2 regions was effective
475 at delineating taxa of Botryosphaeriaceae at the generic level, and RPB2 in addition to other
476 markers like ITS were effective for species delimitation. In view of the difficulty in obtaining
477 primer pairs specific to the Botryosphaeriaceae family for the RPB2 and ITS loci, only the LSU
478 locus could be targeted. This marker was used as a complement to the fungal ITS marker in our
479 study, advantageously giving valuable information about the whole fungal community. In a first
480 attempt, the ITS dataset was constructed after assembly of both sequence pairs produced by
481 Illumina sequencer (R1 and R2, data not shown) revealing very few Botryosphaeriaceae. We
482 observed that the ITS1F forward primer showed mismatches with some Botryosphaeriaceae
483 species, as previously shown with other fungal taxa (Bellemain et al. 2010; Op De Beeck et al.
484 2014). The second attempt at amplicon construction - inspired by Pauvert et al. (2019) and
485 mentioned herein - used only the R1 read pair produced by Illumina sequencing and bypassed
486 potential misassembly of the ITS sequence. This strategy drastically improved the performance
487 of the analysis for the detection of wood fungi, including Botryosphaeriaceae. Our experience
488 confirms that despite the ITS region being considered as the universal barcode for fungi (Schoch
489 et al. 2012), its use can be problematic when dealing with some groups (Op De Beeck et al.
490 2014); moreover, the choice of bioinformatic pipeline will greatly impact the overall analysis
491 (Pauvert et al. 2019).
492 In general, the designed LSU primers showed good specificity for amplifying the
493 Botryosphaeriaceae community, but the taxonomical resolution of the amplified barcode
Page 24 of 53
Benoit Laurent
Phytobiomes Journal
25
494 showed only slight improvement over ITS for some groups of species, especially within Diplodia;
495 this confirms the difficulty in identifying Botryosphaeriaceae with only one molecular marker,
496 due to their close phylogenetic proximity (Slippers et al. 2013; Yang et al. 2017). Some authors
497 have proposed methods to develop new molecular markers with high species resolution using
498 other genes than those traditionally used in phylogeny (e.g., Feau et al. 2011), but their use in
499 metabarcoding is currently hampered by the lack of DNA sequences in databases. Finally, our
500 assignations were most frequently at a so-called clade-level, corresponding to several potential
501 species. As a result, diversity was probably underestimated and host-taxa associations could not
502 be fully described. Overall, the metabarcoding strategy developed herein is a powerful tool for
503 giving a preliminary insight into Botryosphaeriaceae diversity in internal twig tissues, although
504 more focused diagnostic tools will be needed to confirm the taxonomy of the species of
505 interest.
506 A rich fungal community in twigs, including Botryosphaeriaceae species among other known
507 pathogens.
508 Knowledge about the microbiome of the plant internal tissue -the endosphere - is more limited
509 than for the microbiome of the rhizosphere and phyllosphere, especially for trees (Terhonen et
510 al. 2019). Nonetheless, it seems that Ascomycota species are predominant within plant tissue as
511 observed in our study, unlike in soil environments, where Basidiomycota seem to be more
512 abundant (Kemler et al. 2013; Bruez et al. 2014; Kovalchuk et al. 2018; Terhonen et al. 2019).
513 Several taxa frequently identified in the wood tissue in this study, such as Cladosporium sp.,
514 Alternaria sp., Aureobasidium sp., Mycosphaerella sp. or Epicoccum sp., are well known
Page 25 of 53
Benoit Laurent
Phytobiomes Journal
26
515 ubiquitous fungi of the phyllosphere (e.g., Fort et al. 2016) and of the endosphere (Bensch et al.
516 2012; Kemler et al. 2013; Bruez et al. 2014; Pinto and Gomes 2016). Most of these frequent taxa
517 have species members known to be plant pathogens, but this does not necessarily mean that
518 they are pathogenic in the sampled plants (Ganley, Brunsfeld and Newcombe 2004; Caruso,
519 Rillig and Garlaschelli 2012); therefore, determining a clear ecological role for these taxa is not
520 yet possible (Baldrian 2016). More interestingly, the large representation of these ubiquitous
521 fungal taxa can hamper the detection of less abundant taxa, which may explain why
522 Botryosphaeriaceae were not as frequently identified using the ITS marker as with the LSU
523 marker.
524 The state of the Botryosphaeriaceae diversity in vineyards and adjacent forests in South West
525 of France
526 Using the LSU marker, it was possible to identify Botryosphaeriaceae in all sampled sites and to
527 confirm their high frequency in French vineyards as previously reported (Bruez et al. 2014). In
528 French vineyards, seven species of Botryosphaeriaceae have been described in association with
529 “black dead arm” disease in the past: B. dothidea, D. intermedia, D. mutila, D. seriata, Do.
530 viticola, N. parvum et L. viticola (Nivault 2017). All of these species are included in clades
531 detected in our study, except Lasiodiplodia viticola. The high frequency of Diplodia clade 2 and
532 of Neofusicoccum clade 1 in our study is likely related to that of D. mutila, D. seriata, and N.
533 parvum, often considered to be the most frequent and aggressive Botryosphaeriaceae species
534 on grapevine (Bellée et al. 2016; Nivault 2017). Furthermore, we identified Neofusicoccum
535 australe, a species originally thought to be native to the southern hemisphere and not
Page 26 of 53
Benoit Laurent
Phytobiomes Journal
27
536 previously reported in France, but recently reported on a variety of hosts - including on
537 grapevine - worldwide, including in Europe, such as in Italy (Linaldeddu et al. 2010; Pitt et al.
538 2010; Besoain et al. 2013; Baskarathevan et al. 2017). Nonetheless, this taxonomic assignment
539 will need to be confirmed using other genetic markers because N. australe could be confounded
540 with genetically close species, like N. luteum which is broadly distributed in Europe (Alves et al.
541 2013, Barradas et al. 2013).
542 In French forests, a main source of information about the Botryosphaeriaceae diversity is the
543 Forest Health Service (“DSF”) database, which includes all disease reports, with suspected
544 causative agents, made by Forest Health technicians. In this database (queried on 2019/13/11
545 for the period from 2007 to 2019), 87.6% of the 756 reports related to Botryosphaeriaceae fungi
546 and validated by molecular methods referred to D. sapinea on different pine species, confirming
547 the serious impact of this pathogen on pines in France and elsewhere (Fabre et al. 2011; Luchi et
548 al. 2014; Brodde et al. 2018; Kaya et al. 2019; Zlatković et al. 2019). However, the frequency of
549 reports is relatively low for maritime pine compared to black pines (Pinus laricio and P. nigra), as
550 also observed by Fabre et al (2011). In oaks, a few reports from the DSF identified D. corticola/B.
551 corticola (n=8), D. mutila/B. stevensii (n=6), B. dothidea (n=1), D. seriata (n=1), Do. iberica (n=2)
552 and N. parvum (n=1) and were generally associated with observations of cankers or dieback.
553 Linaldeddu et al. (2017) also reported D. corticola as a cause of canker and dieback on Quercus
554 ilex, Q. petraea, and Q. suber in Corsica. Hence, the high frequency of diverse
555 Botryosphaeriaceae species on maritime pine and oak reported during this study was rather
556 unexpected. Our finding can be explained by the high sensitivity of the metabarcoding strategy
Page 27 of 53
Benoit Laurent
Phytobiomes Journal
28
557 compared to standard isolation. This may be due to the proximity of vineyards, but most
558 importantly, it may result from the fact that asymptomatic tissues were also taken into account,
559 in contrast to the French Forest Health Service reports that aim to identify the causal agent(s) of
560 damaged trees.
561 Moreover, no association between symptoms and Botryosphaeriaceae diversity was revealed
562 during this study. This is not really surprising since other agents of necrosis may be involved in
563 the observed symptoms, and Botryosphaeriaceae fungi are well known endophytes; i.e., they
564 can develop within plant tissues without causing symptoms (Slippers and Wingfield 2007; Jami
565 et al. 2013; Bruez et al. 2014). Our results therefore support the importance of
566 Botryosphaeriaceae species as prevalent endophytes in pines and oaks as previously suggested
567 for other hosts and regions (Slippers and Wingfield 2007; Jami et al. 2013). Most of the species
568 reported in the DSF database and elsewhere (Linaldeddu et al. 2017; Fabre et al. 2011;
569 Decourcelle et al. 2015) are included in the most frequent clades identified in our study:
570 Diplodia clade 4: (Diplodia corticola, D. gallae, D. intermedia, D. sapinea and/or D. seriata),
571 Diplodia clade 2 (Diplodia mutila, D. scrobiculata, D. seriata and/or D. conspersa),
572 Neofusicoccum sp. and B. dothidea. It can be hypothesized that the high frequency of Diplodia
573 clade 4 is associated with D. sapinea (Decourcelle et al. 2015) in pine twigs, and with D. corticola
574 in oak twigs (Linaldeddu et al. 2017). On the other hand, the prevalence of Neofusicoccum
575 species in forests is more intriguing. Recently, N. parvum was isolated at high frequency from
576 declining oak stands in Italy, as well as from the insects present in the stands (Panzavolta et al.
577 2017), and N. australe was reported to be pathogenic to oaks in Portugal (Barradas et al. 2013).
Page 28 of 53
Benoit Laurent
Phytobiomes Journal
29
578 Severe damage by Neofusicoccum species have not yet been reported in French forests, and our
579 study did not correlate Neofusicoccum occurrence with symptoms either. However, like the
580 other Botryosphaeriaceae species identified during this analysis, Neofusicoccum species may
581 become serious threats if the conditions become more favourable for these opportunistic
582 pathogens especially favoured by water stress (Slippers and Wingfield 2017); they should
583 therefore be monitored more closely.
584 Do Botryosphaeriaceae spread from vineyard to forest and vice versa?
585 The importance of fluxes across wild and cultivated compartments in agricultural landscapes for
586 the epidemiology and evolution of plant diseases has only recently been emphasised and
587 studied (Burdon and Thrall 2008; Papaïx et al. 2015). Like forest trees, grapevines are perennial
588 plants which facilitate fungal exchanges over time. For example, it was shown that the
589 emergence of a virulent isolate of Venturia inaequalis in European domestic apple orchards had
590 its origin in a non-agricultural host (Lemaire et al. 2016) and that gene flow was high between
591 the agricultural and non-agricultural compartments (Leroy et al. 2016). The generalist behaviour
592 of some Botryosphaeriaceae species further increases their chance of finding host plants in
593 different compartments. For example, common pathogenic Botryosphaeriaceae species were
594 frequently isolated from apple and pear orchards and nearby vineyards in South Africa (Cloete
595 2010), as well as from marula trees and adjacent mango trees (Mehl et al., 2016). For the latter
596 case, gene flow between N. parvum populations colonising both trees was demonstrated (Mehl
597 et al., 2016). In this study, we report the existence of common Botryosphaeriaceae species
598 between vineyard and forest, such as B. dothidea and N. australe. The identification of other
Page 29 of 53
Benoit Laurent
Phytobiomes Journal
30
599 common Botryosphaeriaceae species was hampered by the inability to better separate the
600 taxonomy into more robust clades. Nonetheless, common ASVs, which are more likely to
601 represent species and even populations within species (Callahan et al. 2016), were identified in
602 all three host tissues; this suggest that vineyards and forest frequently share
603 Botryosphaeriaceae species. Therefore, the risk of one pathogen spillover from one
604 compartment to another under new selective pressures probably exists. More specific
605 population genetic studies will help to discover the potential gene flow occurring between the
606 populations of Botryosphaeriaceae which inhabit forests and vineyards in Nouvelle-Aquitaine.
607 Fungal richness and host associations are better explained by the nature of the agro-
608 ecosystem than by phylogeny, including for Botryosphaeriaceae
609 The different indicators of richness and species frequencies used in our study consistently
610 suggested that the Botryosphaeriaceae communities share similar patterns between oaks and
611 pines as opposed to grapevine, despite there being a closer phylogenetic relationship between
612 oaks and grapevines (Kumar et al. 2017). In 2016, Fort et al. showed that host species was the
613 main factor explaining the composition and seasonal changes of foliar fungal communities in
614 grapevines and adjacent oaks, hornbeams and chestnuts. This host effect was suggested to
615 include microclimatic conditions and agricultural practices that differed between grapevine and
616 the adjacent forest patches. Consistently, our results may indicate that features of the
617 compartment or agro-ecosystem, such as management practices that largely differ between the
618 vineyards and the adjacent forests, are more likely to structure the fungal community than the
619 host itself. The influence of human activity on Botryosphaeriaceae species composition was
Page 30 of 53
Benoit Laurent
Phytobiomes Journal
31
620 previously reported by Pavlic-Zupanc et al 2014. In this study, N. parvum isolates was
621 predominantly found on Syzygium cordatum in human-disturbed environment and was absent
622 from natural stands, whereas other Neofusicoccum species predominate in these areas. The
623 high occurrence of Botryosphaeriaceae in vineyards has already been highlighted (Jayawardena
624 et al. 2018) and could be explained by some cultural practices over grapevine lifetime, such as
625 pruning, which probably favour the entry and the development of such opportunist pathogens
626 over unmanaged or less managed forest trees (Urbez-Torres 2011)..Pruning practices were also
627 suggested as an important factor explaining frequency of Botryosphaeriaceae and related
628 cankers in various orchard crops such as avocado (McDonald and Eskalen 2011), mango
629 (Sakalidis et al. 2011) or almond trees and other nut crops (Moral et al. 2019).
630 The observed lower richness of endophytic fungi in grapevines compared to forest trees could
631 also be linked to management practices (Compant et al. 2019). Whether increased
632 Botryosphaeriaceae richness is a cause or a consequence of an alteration of the plant
633 microbiome through management practices requires further investigation. Nonetheless,
634 considering the importance of microbial equilibrium for plant health (Wallenstein 2017;
635 Hartman et al. 2018; Sharma et al. 2018, Compant et al. 2019), we believe that adapting
636 management practices in order to maintain or enhance microbial diversity will be key to the
637 development of durable and resilient systems for agriculture and forestry.
638 Acknowledgements:
Page 31 of 53
Benoit Laurent
Phytobiomes Journal
32
639 The authors would like to acknowledge the other partners of the BOdEGA project and other
640 colleagues for occasional valuable and helpful exchanges: Lucia Guérin-Dubrana, Sylvie Bastien,
641 Adrien Rusch and Marie-France Corio-Costet from the INRA research unit SAVE (UMR 1065);
642 Cécile Robin, Cyril Dutech, Martine Martin-Clotté, Arthur Demené, Julie Faivre-D’arcier and
643 Charlotte Mouden from the INRA research unit BioGeCo (UMR1202). Sample preparation and
644 genome sequencing as described in the present publication was performed using the Genomic
645 and Transcriptomic Facility of Bordeaux (https://pgtb.cgfb.u-bordeaux.fr/en). Disease
646 observations and reports from French forests were kindly provided by the Département Santé
647 des Forêts.
648 Literature cited
649 Alves, A., Barradas, C., Phillips, A. J. L., and Correia, A. 2013. Diversity of Botryosphaeriaceae species
650 associated with conifers in Portugal. Eur. J. Plant Pathol. 135:791–804
651 Baldrian, P. 2016. Forest microbiome: diversity, complexity and dynamics. Rev. 41:fuw040
652 Barradas, C., Correia, A., and Alves, A. 2013. First Report of Neofusicoccum australe and N. luteum
653 associated with canker and dieback of Quercus robur in Portugal. Plant Dis. 97:560–560
654 Barton, K. (2018). MuMIn: Multi-Model Inference. R Package Version 1.42.1. Available online at:
655 https://CRAN.R-project.org/package=MuMIn
656 Baskarathevan, J., Jaspers, M. V., Jones, E. E., and Ridgway, H. J. 2013. Development of isolate-specific
657 markers for Neofusicoccum parvum and N. luteum and their use to study rainwater splash dispersal
658 in the vineyard. Plant Pathol. 62:501–509
659 Baskarathevan, J., Jones, E. E., Jaspers, M. V., and Ridgway, H. J. 2017. High genetic and virulence
660 diversity detected in Neofusicoccum luteum and N. australe populations in New Zealand vineyards.
Page 32 of 53
Benoit Laurent
Phytobiomes Journal
33
661 Plant Pathol. 66:268–276
662 Bates, D., Mächler, M., Bolker, B., and Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4.
663 Journal of Statistical Software 67.
664 Op De Beeck, M., Lievens, B., Busschaert, P., Declerck, S., Vangronsveld, J., and Colpaert, J. V. 2014.
665 Comparison and validation of some its primer pairs useful for fungal metabarcoding studies. PLoS
666 One. 9:e97629
667 Begoude Boyogueno, A. D., Slippers, B., Perez, G., Wingfield, M. J., and Roux, J. 2012. High gene flow and
668 outcrossing within populations of two cryptic fungal pathogens on a native and non-native host in
669 Cameroon. Fungal Biol. 116:343–353
670 Bellée, A., Comont, G., Nivault, A., Abou-Mansour, E., Coppin, C., Dufour, M. C., and Corio-Costet, M. F.
671 2016. Life traits of four Botryosphaeriaceae species and molecular responses of different grapevine
672 cultivars or hybrids. Plant Pathol.
673 Bellemain, E., Carlsen, T., Brochmann, C., Coissac, E., Taberlet, P., and Kauserud, H. 2010. ITS as an
674 environmental DNA barcode for fungi: an in silico approach reveals potential PCR biases. BMC
675 Microbiol. 10:189
676 Bensch, K., Braun, U., Groenewald, J. Z., and Crous, P. W. 2012. The genus Cladosporium. Stud. Mycol.
677 72:1–401
678 Bérubé, J. A., Gagné, P. N., Ponchart, J. P., Tremblay, É. D., and Bilodeau, G. J. 2018. Detection of Diplodia
679 corticola spores in Ontario and Québec based on High Throughput Sequencing (HTS) methods. Can.
680 J. Plant Pathol. 40:378–386
681 Besoain, X., Torres, C., Díaz, G. A., and Latorre, B. A. 2013. First report of Neofusicoccum australe
682 associated with Botryosphaeria canker of grapevine in Chile. Plant Dis. 97:143–143
683 Bihon, W., Burgess, T., Slippers, B., Wingfield, M. J., and Wingfield, B. D. 2011. Distribution of Diplodia
Page 33 of 53
Benoit Laurent
Phytobiomes Journal
34
684 pinea and its genotypic diversity within asymptomatic Pinus patula trees. Australas. Plant Pathol.
685 40:540–548
686 Brodde, L., Adamson, K., Camarero, J. J., Castaño, C., Drenkhan, R., Lehtijärvi, A., Luchi, N., Migliorini, D.,
687 Sánchez-Miranda, Á., Stenlid, J., Özdağ, Ş., and Oliva, J. 2018. Diplodia tip blight on its way to the
688 north: drivers of disease emergence in Northern Europe. Front. Plant Sci. 9:1818
689 Bruez, E., Lecomte, P., Grosman, J., Doublet, B., Bertsch, C., Fontaine, F., Ugaglia, A., Teissedre, P.-L.,
690 Costa, J.-P. DA, Guerin-Dubrana, L., and Rey, P. 2012. Overview of grapevine trunk diseases in
691 France in the 2000s. Phytopathol. Mediterr. 52:262–275
692 Bruez, E., Vallance, J., Gerbore, J., Lecomte, P., Da Costa, J.-P., Guerin-Dubrana, L., and Rey, P. 2014.
693 Analyses of the temporal dynamics of fungal communities colonizing the healthy wood tissues of
694 esca leaf-symptomatic and asymptomatic vines. PLoS One. 9:e95928
695 Burdon, J. J., and Thrall, P. H. 2008. Pathogen evolution across the agro-ecological interface: implications
696 for disease management. Evol. Appl. 1:57–65
697 Burgess, T. I., Sakalidis, M. L., and Hardy, G. E. S. J. 2006. Gene flow of the canker pathogen
698 Botryosphaeria australis between Eucalyptus globulus plantations and native eucalypt forests in
699 Western Australia. Austral Ecol. 31:559–566
700 Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J. A., and Holmes, S. P. 2016. DADA2:
701 High-resolution sample inference from Illumina amplicon data. Nat. Methods. 13:581–3
702 Camacho, C., Coulouris, G., Avagyan, V., Ma, N., Papadopoulos, J., Bealer, K., and Madden, T. L. 2009.
703 BLAST+: architecture and applications. BMC Bioinformatics. 10:421
704 Caruso, T., Rillig, M. C., and Garlaschelli, D. 2012. On the application of network theory to arbuscular
705 mycorrhizal fungi-plant interactions: the importance of basic assumptions. New Phytol. 194:891–
706 894
Page 34 of 53
Benoit Laurent
Phytobiomes Journal
35
707 Chethana, K. W. T., Li, X., Zhang, W., Hyde, K. D., and Yan, J. 2016. Trail of decryption of molecular
708 research on Botryosphaeriaceae in woody plants. Phytopathol. Mediterr. 55:147–171
709 Cloete, M. 2010. Pome fruit trees as alternative hosts of grapevine trunk disease pathogens.
710 Cobo-Díaz, J. F., Baroncelli, R., Le Floch, G., and Picot, A. 2019. A novel metabarcoding approach to
711 investigate Fusarium species composition in soil and plant samples. FEMS Microbiol. Ecol. 95
712 Comont, G., Mayet, V., and Corio-Costet, M. F. 2016. First report of Lasiodiplodia viticola ,
713 Spencermartinsia viticola and Diplodia intermedia associated with Vitis vinifera grapevine decline in
714 French vineyards. Plant Dis. 100:2328–2328
715 Compant, S., Samad, A., Faist, H., and Sessitsch, A. 2019. A review on the plant microbiome: Ecology,
716 functions, and emerging trends in microbial application. J. Adv. Res. 19:29–37
717 Crespo, M., Moral, J., Michailides, T. J., and Trouillas, F. P. 2018. First report of black rot on apple fruit
718 caused by Diplodia seriata in California. Plant Dis. 102:824
719 Decourcelle, T., Piou, D., and Desprez-Loustau, M.-L. 2015. Detection of Diplodia sapinea in Corsican pine
720 seeds. Plant Pathol. 64:442–449
721 Desprez-Loustau, M.-L., Marçais, B., Nageleisen, L.-M., Piou, D., and Vannini, A. 2006. Interactive effects
722 of drought and pathogens in forest trees. Ann. For. Sci. 63:597–612
723 Díaz, G. A., Latorre, B. A., Ferrada, E., Gutiérrez, M., Bravo, F., and Lolas, M. 2018. First report of Diplodia
724 mutila causing branch dieback of english walnut cv. chandler in the maule region, Chile. Plant Dis.
725 102:1451
726 Díaz, G. A., Latorre, B. A., Ferrada, E., and Lolas, M. 2019. Identification and characterization of Diplodia
727 mutila, D. seriata, Phacidiopycnis washingtonensis and Phacidium lacerum obtained from apple
728 (Malus x domestica) fruit rot in Maule Region, Chile. Eur. J. Plant Pathol. 153:211–225
729 Fabre, B., Piou, D., Desprez-Loustau, M.-L., and Marçais, B. 2011. Can the emergence of pine Diplodia
Page 35 of 53
Benoit Laurent
Phytobiomes Journal
36
730 shoot blight in France be explained by changes in pathogen pressure linked to climate change?
731 Glob. Chang. Biol. 17:3218–3227
732 Feau, N., Decourcelle, T., Husson, C., Desprez-Loustau, M.-L., and Dutech, C. 2011. Finding single copy
733 genes out of sequenced genomes for multilocus phylogenetics in non-model fungi. PLoS One.
734 6:e18803
735 Fontaine, F., Gramaje, D., Armengol, J., Smart, R., Nagy, Z. A., Borgo, M., Rego, C., and Marie-France
736 Corio-Costet. 2016. Grapevine Trunk Diseases. A review. ©OIV Publ. 1st Editio
737 Fort, T., Robin, C., Capdevielle, X., Delière, L., and Vacher, C. 2016. Foliar fungal communities strongly
738 differ between habitat patches in a landscape mosaic. PeerJ. 4:e2656
739 Ganley, R. J., Brunsfeld, S. J., and Newcombe, G. 2004. A community of unknown, endophytic fungi in
740 western white pine. Proc. Natl. Acad. Sci. U. S. A. 101:10107–12
741 Gardes, M., and Bruns, T. D. 1993. ITS primers with enhanced specificity for basidiomycetes - application
742 to the identification of mycorrhizae and rusts. Mol. Ecol. 2:113–118
743 Gilbert, G. S., and Webb, C. O. 2007. Phylogenetic signal in plant pathogen–host range. Proc. Natl. Acad.
744 Sci. 104:4979–4983
745 Hartman, K., van der Heijden, M. G. A., Wittwer, R. A., Banerjee, S., Walser, J.-C., and Schlaeppi, K. 2018.
746 Cropping practices manipulate abundance patterns of root and soil microbiome members paving
747 the way to smart farming. Microbiome. 6:14
748 Hothorn, T., Bretz, F., Westfall, P., and Heiberger, R. M. (2008).multcomp: simultaneous inference in
749 general parametric models. URLhttp://CRAN.R-project.org. R package version 1.0-0.
750 Jami, F., Slippers, B., Wingfield, M. J., and Gryzenhout, M. 2013. Greater Botryosphaeriaceae diversity in
751 healthy than associated diseased Acacia karroo tree tissues. Australas. Plant Pathol. 42:421–430
752 Jayawardena, R. S., Purahong, W., Zhang, W., Wubet, T., Li, X., Liu, M., Zhao, W., Hyde, K. D., Liu, J., and
Page 36 of 53
Benoit Laurent
Phytobiomes Journal
37
753 Yan, J. 2018. Biodiversity of fungi on Vitis vinifera L. revealed by traditional and high-resolution
754 culture-independent approaches. Fungal Divers. 90:1–84
755 Kaliterna, J., Milicevic, T., Ivic, D., Bencic, D., and Mesic, A. 2012. First report of Diplodia seriata as causal
756 agent of olive dieback in Croatia. Plant Dis. 96:290–290
757 Kaya, A. G. A., Yeltekin, Ş., Lehtijarvi, T. D., Lehtijarvi, A., and Woodward, S. 2019. Severity of Diplodia
758 shoot blight (caused by Diplodia sapinea) was greatest on Pinus sylvestris and Pinus nigra in a
759 plantation containing five pine species. Phytopathol. Mediterr. 58:249–259
760 Kemler, M., Garnas, J., Wingfield, M. J., Gryzenhout, M., Pillay, K.-A., and Slippers, B. 2013. Ion Torrent
761 PGM as tool for fungal community analysis: a case study of endophytes in Eucalyptus grandis
762 reveals high taxonomic diversity J.A. Gilbert, ed. PLoS One. 8:e81718
763 Kovalchuk, A., Mukrimin, M., Zeng, Z., Raffaello, T., Liu, M., Kasanen, R., Sun, H., and Asiegbu, F. O. 2018.
764 Mycobiome analysis of asymptomatic and symptomatic Norway spruce trees naturally infected by
765 the conifer pathogens Heterobasidion spp. Environ. Microbiol. Rep. 10:532–541
766 Kumar, S., Stecher, G., Suleski, M., and Hedges, S. B. 2017. TimeTree: A Resource for Timelines,
767 Timetrees, and Divergence Times. Mol. Biol. Evol. 34:1812–1819
768 Larignon, P., Fulchic, R., Cere, L., and Dubos, B. 2001. Observation on Black Dead Arm in French
769 vineyards. Phytopathol. Mediterr. 40:S336–S342
770 Lemaire, C., De Gracia, M., Leroy, T., Michalecka, M., Lindhard-Pedersen, H., Guerin, F., Gladieux, P., and
771 Le Cam, B. 2016. Emergence of new virulent populations of apple scab from nonagricultural disease
772 reservoirs. New Phytol. 209:1220–1229
773 Leroy, T., Caffier, V., Celton, J.-M., Anger, N., Durel, C.-E., Lemaire, C., and Le Cam, B. 2016. When
774 virulence originates from nonagricultural hosts: evolutionary and epidemiological consequences of
775 introgressions following secondary contacts in Venturia inaequalis. New Phytol. 210:1443–1452
Page 37 of 53
Benoit Laurent
Phytobiomes Journal
38
776 Li, G. Q., Liu, F. F., Li, J. Q., Liu, Q. L., and Chen, S. F. 2018. Botryosphaeriaceae from Eucalyptus
777 plantations and adjacent plants in China. Persoonia - Mol. Phylogeny Evol. Fungi. 40:63–95
778 Linaldeddu, B. T., Maddau, L., and Franceschini, A. 2017. First report of Diplodia corticola causing canker
779 and dieback of Quercus ilex , Q. petraea , and Q. suber in Corsica (France). Plant Dis. 101:256–256
780 Da Lio, D., Cobo-Díaz, J. F., Masson, C., Chalopin, M., Kebe, D., Giraud, M., Verhaeghe, A., Nodet, P.,
781 Sarrocco, S., Le Floch, G., and Baroncelli, R. 2018. Combined metabarcoding and multi-locus
782 approach for genetic characterization of colletotrichum species associated with common walnut
783 (Juglans regia) anthracnose in France. Sci. Rep. 8:10765
784 Luchi, N., Oliveira Longa, C. M., Danti, R., Capretti, P., and Maresi, G. 2014. Diplodia sapinea : the main
785 fungal species involved in the colonization of pine shoots in Italy. For. Pathol. 44:372–381
786 Marsberg, A., Kemler, M., Jami, F., Nagel, J. H., Postma-Smidt, A., Naidoo, S., Wingfield, M. J., Crous, P.
787 W., Spatafora, J. W., Hesse, C. N., Robbertse, B., and Slippers, B. 2016. Botryosphaeria dothidea : a
788 latent pathogen of global importance to woody plant health. Mol. Plant Pathol.
789 McDonald, V., and Eskalen, A. 2011. Botryosphaeriaceae species associated with avocado branch cankers
790 in california. Plant Dis. 95:1465–1473
791 Mehl, J. W. M., and Roux, J. 2017. Overlap of latent pathogens in the Botryosphaeriaceae on a native and
792 agricultural host. Fungal Biol. 121:405–419
793 Moral, J., Morgan, D., and Michailides, T. J. 2019. Management of Botryosphaeria canker and blight
794 diseases of temperate zone nut crops. Crop Prot. 126:104927
795 Nakagawa, S., and Schielzeth, H. (2013). A general and simple method for obtaining R2 from generalized
796 linear mixed-effects models. Methods Ecol. Evol. 4, 133–142.
797 Nguyen, N. H., Song, Z., Bates, S. T., Branco, S., Tedersoo, L., Menke, J., Schilling, J. S., and Kennedy, P. G.
798 2016. FUNGuild: An open annotation tool for parsing fungal community datasets by ecological
Page 38 of 53
Benoit Laurent
Phytobiomes Journal
39
799 guild. Fungal Ecol. 20:241–248
800 Nivault, A. 2017. Diversité et traits d’histoire de vie des Botryosphaeriaceae et évaluation du potentiel de
801 défense de différents cultivars de Vitis. Thesis Manuscript 2017BORD0811.
802 http://www.theses.fr/2017BORD0811/document
803 Paez, C. A., and Smith, J. A. 2018. First report of Diplodia sapinea and Diplodia scrobiculata causing an
804 outbreak of tip blight on slash pine in florida. Plant Dis. 102:1657
805 Panzavolta, T., Panichi, A., Bracalini, M., Croci, F., Ginetti, B., Ragazzi, A., Tiberi, R., and Moricca, S. 2017.
806 Dispersal and propagule pressure of Botryosphaeriaceae species in a declining oak stand is affected
807 by insect vectors. Forests. 8:228
808 Papaïx, J., Burdon, J. J., Zhan, J., and Thrall, P. H. 2015. Crop pathogen emergence and evolution in agro-
809 ecological landscapes. Evol. Appl. 8:385–402
810 Pauvert, C., Buée, M., Laval, V., Edel-Hermann, V., Fauchery, L., Gautier, A., Lesur, I., Vallance, J., and
811 Vacher, C. 2019. Bioinformatics matters: The accuracy of plant and soil fungal community data is
812 highly dependent on the metabarcoding pipeline. Fungal Ecol. 41:23–33
813 Pavlic-Zupanc, D., Wingfield, M. J., Boissin, E. & Slippers, B. The distribution of genetic diversity in the
814 Neofusicoccum parvum/N. ribis complex suggests structure correlated with level of disturbance.
815 Fungal Ecol. 13, 93–102 (2015).
816 Phillips, A. J. L. J. L., Alves, A., Abdollahzadeh, J., Slippers, B., Wingfield, M. J. J., Groenewald, J. Z. Z., and
817 Crous, P. W. W. 2013. The Botryosphaeriaceae: genera and species known from culture. Stud.
818 Mycol. 76:51–167
819 Pinto, C., and Gomes, A. C. 2016. Vitis vinifera microbiome: from basic research to technological
820 development. BioControl. 61:243–256
821 R Core Development Team (2013). R: A Language and Environment for Statistical Computing. Vienna: R
Page 39 of 53
Benoit Laurent
Phytobiomes Journal
40
822 Foundation for Statistical Computing. Available online at: http://www.R-project.org/
823 Pitt, W. M., Huang, R., Steel, C. C., and Savocchia, S. 2010. Identification, distribution and current
824 taxonomy of Botryosphaeriaceae species associated with grapevine decline in New South Wales
825 and South Australia. Aust. J. Grape Wine Res. 16:258–271
826 Sakalidis, M. L., Ray, J. D., Lanoiselet, V., Hardy, G. E. S., and Burgess, T. I. 2011. Pathogenic
827 Botryosphaeriaceae associated with Mangifera indica in the Kimberley region of Western Australia.
828 Eur. J. Plant Pathol. 130:379–391
829 Sakalidis, M. L., Slippers, B., Wingfield, B. D., Hardy, G. E. S. J., and Burgess, T. I. 2013. The challenge of
830 understanding the origin, pathways and extent of fungal invasions: global populations of the
831 Neofusicoccum parvum-N. ribis species complex J. Austin, ed. Divers. Distrib. 19:873–883
832 Schoch, C. L., Seifert, K. A., Huhndorf, S., Robert, V., Spouge, J. L., Levesque, C. A., Chen, W., Fungal
833 Barcoding Consortium, F. B., Fungal Barcoding Consortium, and Schindel, D. 2012. Nuclear
834 ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi.
835 Proc. Natl. Acad. Sci. U. S. A. 109:6241–6
836 Sharma, M., Kansal, R., and Singh, D. 2018. Endophytic microorganisms: their role in plant growth and
837 crop improvement. New and Future Developments in Microbial Biotechnology and Bioengineering
838 2018, Pages 391-413
839 Silva, F. J. A., Santos, K. M. dos, Rêgo, T. J. dos S., Armengol, J., Rossi, V., Michereff, S. J., and Gonzalez-
840 Dominguez, E. 2019. Temporal conidial dispersal pattern of Botryosphaeriaceae species in table-
841 grape vineyards in Northeastern Brazil. Phytopathol. Mediterr. 57:547–556
842 Slippers, B., Boissin, E., Phillips, A. J. L., Groenewald, J. Z., Lombard, L., Wingfield, M. J., Postma, A.,
843 Burgess, T., and Crous, P. W. 2013. Phylogenetic lineages in the Botryosphaeriales: a systematic and
844 evolutionary framework. Stud. Mycol. 76:31–49
Page 40 of 53
Benoit Laurent
Phytobiomes Journal
41
845 Slippers, B., Crous, P. W., Jami, F., Groenewald, J. Z., and Wingfield, M. J. 2017. Diversity in the
846 Botryosphaeriales: Looking back, looking forward. Fungal Biol. 121:307–321
847 Slippers, B., and Wingfield, M. J. 2007. Botryosphaeriaceae as endophytes and latent pathogens of
848 woody plants: diversity, ecology and impact. Fungal Biol. Rev. 21:90–106
849 Stanosz, G. R., Blodgett, J. T., Smith, D. R., and Kruger, E. L. 2002. Water stress and Sphaeropsis sapinea
850 as a latent pathogen of red pine seedlings. New Phytol. 149:531–538
851 Stanosz, G. R., Smith, D. R., and Leisso, R. 2007. Diplodia shoot blight and asymptomatic persistence of
852 Diplodia pinea on or in stems of jack pine nursery seedlings. For. Pathol. 37:145–154
853 Sturrock, R. N., Frankel, S. J., Brown, A. V., Hennon, P. E., Kliejunas, J. T., Lewis, K. J., Worrall, J. J., and
854 Woods, A. J. 2011. Climate change and forest diseases. Plant Pathol. 60:133–149
855 Terhonen, E., Blumenstein, K., Kovalchuk, A., and Asiegbu, F. O. 2019. Forest tree microbiomes and
856 associated fungal endophytes: functional roles and impact on forest health. Forests. 10:42
857 Urbez-Torres, J. R. 2011. The status of Botryosphaeriaceae species infecting grapevines. Phytopathol.
858 Mediterr. 50:5–45
859 Úrbez-Torres, J. R., Boulé, J., and O’Gorman, D. T. 2016. First report of Diplodia seriata and D. mutila
860 causing apple dieback in British Columbia. Plant Dis. 100:1243
861 Úrbez-Torres, J. R., Bruez, E., Hurtado, J., and Gubler, W. D. 2010. Effect of temperature on conidial
862 germination of Botryosphaeriaceae species infecting grapevines. Plant Dis. 94:1476–1484
863 Úrbez-Torres, J. R., Leavitt, G. M., Voegel, T. M., and Gubler, W. D. 2006. Identification and distribution of
864 Botryosphaeria spp. associated with grapevine cankers in california. Plant Dis. 90:1490–1503
865 Úrbez-Torres, J. R., Peduto, F., Vossen, P. M., Krueger, W. H., and Gubler, W. D. 2013. Olive twig and
866 branch dieback: etiology, incidence, and distribution in California. Plant Dis. 97:231–244
867 Wallenstein, M. D. 2017. Managing and manipulating the rhizosphere microbiome for plant health: A
Page 41 of 53
Benoit Laurent
Phytobiomes Journal
42
868 systems approach. Rhizosphere. 3:230–232
869 De Wet, J., Slippers, B., Preisig, O., Wingfield, B. D., and Wingfield, M. J. 2008. Phylogeny of the
870 Botryosphaeriaceae reveals patterns of host association. Mol. Phylogenet. Evol. 46:116–126
871 White, T. J., Bruns, T., Lee, S., and Taylor, J. 1990. Amplification and direct sequencing of fungal
872 ribosomal rna genes for phylogenetics. Pages 315–322 in: PCR Protocols, Elsevier.
873 Yang, T., Groenewald, J. Z., Cheewangkoon, R., Jami, F., Abdollahzadeh, J., Lombard, L., and Crous, P. W.
874 2017. Families, genera, and species of Botryosphaeriales. Fungal Biol. 121:322–346
875 Zhou, Y. 2017. Botryosphaeria rosaceae sp. nov. and B. ramosa, new botryosphaeriaceous taxa from
876 China. Mycosphere. 8:162–171
877 Zinger, L., Bonin, A., Alsos, I. G., Bálint, M., Bik, H., Boyer, F., Chariton, A. A., Creer, S., Coissac, E., Deagle,
878 B. E., De Barba, M., Dickie, I. A., Dumbrell, A. J., Ficetola, G. F., Fierer, N., Fumagalli, L., Gilbert, M. T.
879 P., Jarman, S., Jumpponen, A., Kauserud, H., Orlando, L., Pansu, J., Pawlowski, J., Tedersoo, L.,
880 Thomsen, P. F., Willerslev, E., and Taberlet, P. 2019. DNA metabarcoding—Need for robust
881 experimental designs to draw sound ecological conclusions. Mol. Ecol. 28:1857–1862
882 Zlatković, M., Ke?a, N., Wingfield, M. J., Jami, F., and Slippers, B. 2017. New and unexpected host
883 associations for Diplodia sapinea in the Western Balkans. 47:e12328
884 Zlatković, M., Wingfield, M. J., Jami, F., and Slippers, B. 2019. Genetic uniformity characterizes the
885 invasive spread of Neofusicoccum parvum and Diplodia sapinea in the Western Balkans. For. Pathol.
886 49:e12491
887 Zlatković, M., Wingfield, M. J., Jami, F., and Slippers, B. 2018. Host specificity of co-infecting
888 Botryosphaeriaceae on ornamental and forest trees in the Western Balkans. For. Pathol. 48:e12410
889
890
Page 42 of 53
Benoit Laurent
Phytobiomes Journal
43
891 Tables
892 Table 1: Summary of the site and the sample numbers according to the appellations, the hosts
893 and the presence or not of symptoms.
Number of samples
(Asymptomatic / Symptomatic)
Appellations
Site
number
Pinus
Quercus
Vitis
Total
Entre deux mers
3
2/0
9/4
4/8
27 (15/12)
Graves
5
16/5
17/5
6/16
65 (41/24)
Libournais
5
16/1
19/9
7/14
66 (42/24)
Médoc
4
12/1
11/8
7/11
50 (30/20)
Pessac-Léognan
5
17/3
11/4
9/14
58 (37/21)
Tursan
3
9/0
15/4
2/16
46 (26/20)
Vin des sables
3
15/1
5/4
5/10
40 (25/15)
Total
28
87/11
87/38
40/89
352 (214/138)
894
Page 43 of 53
Benoit Laurent
Phytobiomes Journal
44
896 Table 2: Taxa identified from the mock sequencing. First mock community was composed of
897 Botryosphaeriaceae species (D. corticola, D. mutila, D. sapinea, D. seriata, L. vitis, N. parvum)
898 and the second Mock was composed of endophytes species including two Botryosphaeriaceae
899 (D. intermedia, B. dothidea, E. nigrum, E. lata, Alternaria sp., A. pullulans, Cladosporium sp.,
900 Phoma sp.) Black indicates exact match and grey indicates genera match with wrong or
901 imprecise species assignation.
902
Botryosphaeriaceae mock
Endophytes mock
Database sequence(s)
Read
count
Taxonomy
Diplodia corticola
Diplodia mutila
Diplodia sapinea
Diplodia seriata
Lasiodiplodia vitis
Neofusicoccum parvum
Diplodia intermedia
Botryosphaeria dothidea
Epicoccum nigrum
Eutypa lata
Alternaria sp.
Aureobasidium pullulans
Cladosporium sp.
Phoma sp.
Bot_LSU_TaxID_61
7094
Neofusicoccum australe
Bot_LSU_TaxID_27
3629
Diplodia sp.
Bot_LSU_TaxID_1
2591
Diplodia mutila, D. pyri
Bot_LSU_TaxID_15_48
2213
Diplodia sp., Dothoriella sp.
Bot_LSU_TaxID_39_42
1249
Diplodia sp. a
Bot_LSU_TaxID_59
947
Neofusicoccum sp
Bot_LSU_TaxID_55
168
Lasiodiplodia vitis
Bot_LSU_TaxID_39_42
17184
Diplodia sp.a
Bot_LSU_TaxID_2
2359
Botryosphaeria dothidea
LSU sequencing
HE820797
108
Mycosphaerellaceae sp.
Bot_ITS_TaxID_99
6212
Dichomera versiformis
Bot_ITS_TaxID_83
5272
Diplodia sapinea, D. seriata
Bot_ITS_TaxID_100
5265
Diplodia rosulata, D.
medicaginis
Bot_ITS_TaxID_9
2460
D. mutila, D. stevensii
MH716405
148
Lasiodiplodia
pseudotheobromae
KX784262
11284
Cladosporium sp.
KT895930, KX516021,
MG818910, GU721359,
EU479964
7700
Alternaria sp.
KT693730
796
Aureobasidium namibiae
Bot_ITS_TaxID_83
311
Diplodia sapinea, D. seriata
KC843295_refs
70
Phoma foeniculina
FJ426996_refs
66
Epicoccum nigrum
ITS sequencing
Bot_ITS_TaxID_8
27
Botryosphaeria dothidea
903 a Diplodia corticola, D. gallae, D. intermedia, D. sapinea, D. seriata
Page 44 of 53
Benoit Laurent
Phytobiomes Journal
45
904 Table 3: Data processing for ITS sequencing and LSU sequencing.
ITS sequencing
LSU sequencing
Step 1
Step 2
Step 3
Step 4
Step 5
Step 1
Step 2
Step 3
Step 4
Step 5
Step 6
Total
number of
reads after
sequencing
Filtrated
reads /
clustered
in ASVs
Reads/ASV
s with no
Blast
results
Reads/ASVs
with poor
Blast results
Reads/ASV
s kept for
analysis
Total
number of
reads after
sequencing
Filtrated
reads /
clustered in
ASVs
Reads/ASVs
with no
Blast results
Reads/ASVs
with no
Botryosphae
riaceae
match
Reads/ASVs
with poor
Blast results
Reads/ASVs
kept for
analysis
Pine twigs
(n=98)
6,577,463
3,267,399 /
2,710
5,273 / 127
1,068,410 /
1,181
2,193,716 /
1402
3,978,758
2,635,225 /
1,229
423,992 /
724
1,754,402 /
405
23,451 / 46
433,380 / 54
Oak twigs
(n=125)
6,659,029
3,423,599 /
3,055
1,455 / 31
952,336 /
1,141
2,469,808 /
1,883
3,686,005
2,411,420 /
764
441,111 /
339
825,492 /
226
834,379 /
166
310,438 / 33
Grapevine
twigs
(n=129)
11,113,431
5,740,793 /
1,617
1,613 / 6
320,282 /
308
5,418,898 /
1,303
6,989,772
5,231,766 /
715
241,047 /
281
1,347,357 /
197
323,473 / 99
3,319,889 /
138
Negative
control
(n=6)
8,833
1,641 / 24
0 / 0
16 / 5
1,625 / 19
2,489
144 / 21
25 / 12
33 / 4
10 / 1
76 / 8
Total
24,358,756
12,435,073
/ 6,252
8,341 / 164
2,341,044 /
2,324
10,082,422
/ 3,763
14,657,024
10,278,555 /
2,500
1,106,175 /
1,310
3,927,284 /
828
1,181,313 /
265
4,063,783 /
208
Page 45 of 53
Benoit Laurent
Phytobiomes Journal
46
906 Table 4: List of the seven most frequently identified Botryosphaeriaceae clades
Database sequence(s)
Clade_ID
Potential assignations within the clade
Frequency in
samples (%)
Bot_LSU_TaxID_59
Clade 1
Neofusicoccum algeriense, N. andinum,
N. arbuti, N. batangarum, N.
cryptoaustrale, N. kwambonambiense,
N. luteum, N. macroclavatum, N.
mangiferae, N. nonquaesitum, N.
occulatum, N. parvum, N. pistaciae, N.
ribis, N. umdonicola, N. ursorum, N.
viticlavatum
31.53
Bot_LSU_TaxID_27
Clade 2
Diplodia mutila, D. scrobiculata, D.
seriata, D. conspersa
29.83
Bot_LSU_TaxID_2
Clade 3
Botryosphaeria dothidea
22.73
Bot_LSU_TaxID_39,
Bot_LSU_TaxID_42
Clade 4
Diplodia corticola, D. gallae, D.
intermedia, D. sapinea, D. seriata
19.03
Bot_LSU_TaxID_1
Clade 5
Diplodia mutila, D. pyri
11.65
Bot_LSU_TaxID_61
Clade 6
Neofusicoccum australe
7.10
Bot_LSU_TaxID_15,
Bot_LSU_TaxID_48
Clade 7
Diplodia coryli, D. juglandis, D.
medicaginis, D. seriata, D.
spegazziniana; Dothiorella citricola, Do
iberica, Do mangifericola, Do omnivora,
Do parva, Do plurivora, Do rosulata, Do
sarmentorum, Do sempervirentis, Do
vidmadera, Do viticola, Do westralis, Do
rosulata
5.97
907
908
Page 46 of 53
Benoit Laurent
Phytobiomes Journal
47
909 Table 5: Mean clade richness per sample and per site and mean ASV richness per sample and per
910 site. “n.sa” indicates number of samples and “n.si” the number of sites, +/- XX gives the standard
911 deviation.
Clade richness
ASVs richness
Total
Vitis
vinifera
Quercus
sp.
Pinus sp.
Total
Vitis
vinifera
Quercus sp.
Pinus sp.
n.sa =
352 n.si =
27
n.sa = 129
n.si = 27
n.sa = 125
n.si = 26
n.sa = 95
n.si = 18
n.sa = 352
n.si = 27
n.sa = 129
n.si = 27
n.sa = 125
n.si = 26
n.sa = 95
n.si = 18
Endophytes
23.5
(+/- 12.5)
20.0
(+/- 8.9)
26.5
(+/- 12.1)
24.2
(+/- 15.6)
36.3
(+/- 22.5)
30.4
(+/- 14.5)
39.5
(+/- 21.3)
39.9
(+/- 30.0)
Per
Samples
Botryo
1.4 (+/-
1.0)
1.8
(+/- 1.0)
1.0
(+/- 0.9)
1.1
(+/- 0.9)
1.9
(+/- 2.3)
2.9
(+/- 2.7)
1.1
(+/- 1.2)
1.6
(+/- 2.5)
Endophytes
133.6
(+/- 30.4)
52.2
(+/- 14.1)
80.5
(+/- 20.3)
74.9
(+/- 26.5)
327.6
(+/- 116.1)
104.4
(+/- 35.8)
155.5
(+/- 48.7)
165.0
(+/- 81.8)
Per Sites
Botryo
6.8
(+/- 1.7)
4.8
(+/- 1.0)
3.8
(+/- 1.9)
4.3
(+/- 1.5)
14.0
(+/- 6.9)
9.9
(+/- 5.4)
4.3
(+/- 2.5)
6.6
(+/-4.6)
912
913
Page 47 of 53
Benoit Laurent
Phytobiomes Journal
48
914 Table 6: p-value of ANOVA from the model testing: effect of the host (genera); presence or
915 absence of symptoms in the woods; appellation; and interaction of host and symptoms on clade
916 and ASV richness. Significant p-values (<0.01) are indicated in bold. The estimate effects, using
917 the marginal coefficient of determination (R²m) or the conditional coefficient of determination
918 (R²c), are indicated at the bottom.
Factors / Response
Botryo clade
richness
Endophyte
clade richness
Botryo ASV
richness
Endophyte
ASV richness
Host
1.02E-05
< 2.2E-16
< 2.2E-16
5.02E-16
Symptoms
0.297
1.85E-03
0.4634
8.64E-15
Appelations
0.991
1.28E-04
0.2285
2.52E-05
Host x Symptoms
0.708
0.1318158
0.379
0.05632
Host x Appellations
0.2853
< 2.2E-16
3.88E-13
< 2.2E-16
R²m
0.129
0.472
0.316
0.616
R²c
0.129
0.522
0.363
0.699
919
920
Page 48 of 53
Benoit Laurent
Phytobiomes Journal
49
921 Figure captions
922 Figure 1: Average clade diversity according to the host sampled for the Botryosphaeriaceae
923 community and for the endophyte fungal community. Bars represent standard-errors.
924 Figure 2: Distribution of ASVs and clade according to hosts.
925 Figure 3: Histogram of clade frequency. a. Average clade frequency per site. b. Total clades
926 frequency across sites. * means significant host effect and letters (i and j) indicate significant
927 difference between hosts.
928 Figure 4: Aquitaine map introducing the different appellations sampled and the histogram
929 showing the clades frequencies in these appellations.
Page 49 of 53
Average clade diversity according to the host sampled for the Botryosphaeriaceae community and for the
endophyte fungal community. Bars represent standard-errors.
338x190mm (96 x 96 DPI)
Page 50 of 53
Distribution of ASVs and clade according to hosts.
338x190mm (96 x 96 DPI)
Page 51 of 53
Histogram of clade frequency. a. Average clade frequency per site. b. Total clades frequency across sites. *
means significant host effect and letters (i and j) indicate significant difference between hosts.
338x190mm (96 x 96 DPI)
Page 52 of 53
Aquitaine map introducing the different appellations sampled and the histogram showing the clades
frequencies in these appellations.
338x190mm (96 x 96 DPI)
Page 53 of 53
