A Novel Seimatosporium and Other Sporocadaceae Species Associated with Grapevine Trunk Diseases in Cyprus

Besides well-known grapevine trunk disease (GTD)-related pathogens, there is an increased interest in wood-colonizing fungi that infect grapevines. During 2017–2018, a survey was conducted in Cyprus and wood samples were collected from vines exhibiting typical GTD symptoms. Based on morphological and multilocus phylogenetic analyses (ITS, LSU, bt2, tef1-a), four species in the Sporocadaceae family were described and typified; two in the genus of Seimatosporium: Seim. cyprium sp. nov. and Seim. vitis-viniferae and two in Sporocadus: Spo. kurdistanicus and Spo. rosigena. The teleomorph of Seim. cyprium sp. nov. was also described. Pathogenicity trials with representative isolates of each species were performed on woody stems of two-year-old potted grapevines for 12 months under field conditions. All isolates were pathogenic, causing dark brown to black vascular discoloration, extending upward and downward from the inoculation point. Sporocadus isolates were significantly more aggressive than Seimatosporium with lesion lengths ranging from 9.24 to 6.90 and 4.13 to 4.00 cm, respectively. Successful re-isolations were also evident for all species and isolates. Seim. cyprium sp. nov. is a newly described species, while Spo. kurdistanicus and Spo. rosigena are reported for the first time in Europe on Vitis vinifera, suggesting the potential role of Sporocadaceae in the GTDs complex.


Introduction
Grapevine trunk diseases (GTDs) form an aggregate of fungal diseases that are currently considered the most destructive biotic factor of grapevines globally [1][2][3][4]. Multifaceted adverse effects due to GTDs include reduced longevity and profitable vineyard lifespan [5], cumulative yield losses, increased costs due to applied management practices, and premature replanting of severely affected vineyards [6,7]. Although long considered a deteriorating factor of viticulture, GTDs' incidence and subsequent consequences have spiked along the last three decades, mainly due to the circulation of potentially contaminated planting material, an industry shift towards cultivation systems that render vines more prone to wood infections, as well as the lack of effective plant protection products [6,[8][9][10].
To date, over 140 fungal species (predominately ascomycetous) belonging to 35 genera have been reported to be associated with GTDs worldwide, while numerous grapevine microbiome studies enrich our knowledge on species involved with GTDs; however, the degree of involvement of many GTD-related species remains elusive [6,[11][12][13].

Morphological Description
The color of Sporocadaceae colonies ranged from white to grayish and from light brown to light sepia on PDA and MEA, while on OA, they were from grayish white to light sienna or brown. Characters used to distinguish the species included the relative cell

Morphological Description
The color of Sporocadaceae colonies ranged from white to grayish and from light brown to light sepia on PDA and MEA, while on OA, they were from grayish white to light sienna or brown. Characters used to distinguish the species included the relative cell lengths of conidia, conidial septation, and the appendage morphology. Based on microscopic observations, all isolates produced septate conidia, which matched the descriptions for Sporocadaceae-like asexual forms. However, Sporocadaceae isolates were separated into two different groups. The first consisted of isolates (n = 6) showing mostly 3-septate, obovoid, thick-walled, pale brown to brown conidia that lacked appendages at both the basal and apical ends. These features matched those described earlier for the genus Sporocadus [34]. The second group (n = 10) included isolates with fusoid, pale brown, mostly 3-up to 6septate conidia, bearing basal and apical (sometimes absent appendages, which resembled those of the genus Seimatosporium [34]. There were significant differences in conidial length (p [F (1,209) > 1058] < 0.0001) and width (p [F (1,209) > 1208] < 0.0001) between Sporocadus species, with Spo. kurdistanicus being on average longer and wider as compared to Spo. rosigena. Similarly, Seim. vitis-viniferae produced, on average, smaller conidia to Seim. cyprium, with significant differences in their conidial length (p [F( 1, 121) > 57.41] < 0.0001) and width (p [F (1,121) > 88.15] < 0.0001). Homothallic crosses were performed to induce sexual reproductive structures in obtained Sporocadaceae isolates. Seim. cyprium was the only species that produced abundant perithecia immersed on grapevine wood segments bearing eight uniseriate, single-septate, ascospores per ascus.

Effect of Temperature on Mycelial Growth
All Sporocadaceae isolates were able to grow at all tested temperatures and an analyses of variance (ANOVA) indicated no significant differences (p < 0.05) of the mycelial growth among the experiments, thus the data were combined. The relationship between mycelial growth at different temperatures was best described by a cubic response model (y = aT 3 + bT 2 + cT + d). All regression coefficients were significantly different (p < 0.01), and the coefficients of determination (R 2 ) ranged from 0.89 to 0.99 (Table 1). Based on the adjusted models derived per isolate, the optimum temperatures of mycelial growth on PDA ranged from 20.3 to 24.1 • C with significant differences detected among them (Table 1). More specifically, optimum temperature ≥ 23 • C included only Seim. cyprium isolates. Isolates with an optimum temperature between 22 and 21 • C were Seim. vitis-vineferae and Spo. rosigena isolates, as well as the Spo. kurdistanicus isolate P158. The only species that had an optimum temperature < 21 • C was the Spo. kurdistanicus L181. The maximum mycelial growth rates also differed significantly (mean = 4.4; median = 4.4). More specifically, all isolates had a maximum growth rate > 4 mm/day, ranging from 4.1 to 4.8, except L181 with < 4 mm/day (Table 1). Furthermore, maximum growth rates on MEA cultures ranged from 2.6 to 5 mm/day (mean = 4.3; median = 4.9), with isolates L181 and the Seim. cyprium L111 bearing the highest and lowest values, respectively, and optimum growth temperatures 20.4-24.4 • C (data not shown). The highest average growth rate for Sporocadaceae isolates was evident on OA ranging from 4.4 to 5.9 mm/day (mean = 5.3; median = 5.5), with P60 and L34 being the slowest-and fastest-growing isolates, respectively, and optimum growth temperatures from 21.4 to 24.3 • C (data not shown).
Morphological comparisons, coupled with phylogenetic analyses of the ITS, LSU, bt2, and tef1-a dataset, identified four distinct and strongly supported lineages, one of which has no apparent species name. Thus, we propose the following new species name to properly circumscribe and typify this unique taxon.
Notes: Based on multi-locus phylogenetic analyses Seim. cyprium formed a distinct, highly supported clade that is closely located to Seim. marivanicum (ML/MP = 100/99), Seim. vitifusiforme (ML/MP = 100/99), and Seim. luteosporum (ML/MP = 100/100). Seim. cyprium differs from Seim. marivanicum in ITS (one substitution), tub2 (two substitutions), and tef1-a (one insertion or deletion), while their LSU data are identical, however the species can be distinguished in the tef1-a phylogram; with a total of four PWD. Despite their high sequence similarities, both species have distinct morphological differences in terms of conidial dimensions and septation, with a mean length/width ratio of 4.1 vs. 5(-6), septa number of 3 vs. 3(-6), and media cell length of 5.2 ± 0.58 vs. 8 ± 0.7 µm for Seim. cyprium and Seim. marivanicum, respectively. In addition, the apical appendages of Seim. cyprium are distinctly longer compared to the ex-type of Seim. marivanicum (Table 2). Similarly, phylogenetic analyses and estimates of evolutionary divergence between concatenated sequences used herein, support strong species differentiation (ML/MP = 99/100) of Seim. pistaciae (CBS 138865) and Seim. rosae (CBS 139823), given that they exhibit three PWD and discrete morphological characteristics (Supplementary Table S1; Figure 1) [34]. Seim. luteosporum has significant sequence data differences in ITS (two substitutions and two insertions or deletions), LSU (four substitutions), tub2 (eleven substitutions and one deletion or insertion), and tef1-a (twelve substitutions and seven deletions or insertions) from Seim. cyprium and they could be clearly separated in all the single locus phylogenies, Plants 2022, 11, 2733 8 of 23 except ITS. However, Seim. cyprium cannot be clearly differentiated from Seim. luteosporum by conidial characteristics ( Table 2). The LSU sequences of Seim. cyprium and Seim. vitifusiforme are identical, but they differ in ITS (one substitution), tub2 (one substitution) and tef1-a (five substitutions) and can be clearly distinguished only in the tef1-a phylogram. The conidial characteristics of both species are also discrete ( Table 2). Unlike Seim. cyprium, no teleomorph has been currently reported for Seim. luteosporum, Seim. marivanicum, and Seim. vitifusiforme.
Notes: Pronounced disparities in conidial size and characteristics can distinguish Seim. vitis and Seim. vitis-viniferae. More specifically, conidia of Seim. vitis ex-type specimen (MFLUCC 14-0051) [31] are reported to be larger than the conidia of the relevant Seim. vitis-viniferae specimen (CBS 123004; Table 2) [34]. Furthermore, the same authors report that Seim. vitis bears 3-septate conidia with basal appendages, while Seim. vitis-viniferae may produce 3-6-septate and appendaged conidia basally or at both ends ( Table 2). In the present study, conidial characteristics of obtained isolates matched with those attributed to Seim. vitis-viniferae. However, the abovementioned distinct morphological differences are not supported by phylogenetic analyses of the available sequence data. The only available sequences of the Seim. vitis ex-type (ITS and LSU) are identical to the Seim. vitis-viniferae ex-type and the results are not different when other available datasets of ITS and LSU sequences are compared. Minor differences were evident when the tub2 sequence data obtained in this study (n = 8) and all others currently available in the NCBI database (n = 23) of Seim. vitis and Seim. vitis-viniferae were analyzed. The dissimilarities found (two substitutions and three insertions or deletions) are not consistently represented among the species, suggesting that the tub2 locus is not informative for species separation. Regarding the tef1-a locus, eight substitutions and two insertions or deletions are found within the same set of sequence data. A single, consistent nucleotide substitution in position 257 (T for Seim. vitis and to C for Seim. vitis-viniferae isolates) was detected, while all other nucleotide differences appear to be variably distributed between the species (Supplementary Figure S5). In conclusion, the available sequence data (ITS, LSU, bt2, and tef1-a) cannot strongly support differentiation of these two species.
The colony of Spo. kurdistanicus isolate L181 (CBS 149022) on PDA growing, slightly raised, white with smooth, wooly, smooth margin, reaching 44 mm in diameter after 14 days at 25 • C in darkness. Colony on MEA growing, slightly raised, wooly, white with wooly, smooth margin, reaching 44 mm in diameter after 14 days at 25 • C in darkness. Colony on OA, fast-growing, slightly raised, greyish-white with wooly, smooth margin, reaching 77.75 mm in diameter after 14 days at 25 • C in darkness. The colony of Spo. kurdistanicus isolate L181 (CBS 149022) on PDA growing, slightly raised, white with smooth, wooly, smooth margin, reaching 44 mm in diameter after 14 days at 25 °C in darkness. Colony on MEA growing, slightly raised, wooly, white with wooly, smooth margin, reaching 44 mm in diameter after 14 days at 25 °C in darkness. Colony on OA, fast-growing, slightly raised, greyish-white with wooly, smooth margin, reaching 77.75 mm in diameter after 14 days at 25 °C in darkness.
Ecology and host characteristics: Isolates L158 (=CBS 149023), L164, and L181 (= CBS 149022) belonging to the present species were recovered from vines of the indigenous wine grape cv. Xynisteri. Isolates L105 and L181 were co-isolated with Ph. chlamydospora, while L164 was co-isolated with P. minimum.
Distribution: Cyprus and Iran on V. vinifera. Notes: Multigene, as well as single locus phylogeny of LSU, bt2, and tef1-a cluster extype and Spo. kurdistanicus isolates were collected in this study in a highly supported clade. Spore dimensions of isolates collected in this study are in the same range to the extype isolate from Iran [26], sharing similar characteristics (Table 2).  Ecology and host characteristics: Isolates L158 (=CBS 149023), L164, and L181 (= CBS 149022) belonging to the present species were recovered from vines of the indigenous wine grape cv. Xynisteri. Isolates L105 and L181 were co-isolated with Ph. chlamydospora, while L164 was co-isolated with P. minimum.
Distribution: Cyprus and Iran on V. vinifera. Notes: Multigene, as well as single locus phylogeny of LSU, bt2, and tef1-a cluster ex-type and Spo. kurdistanicus isolates were collected in this study in a highly supported clade. Spore dimensions of isolates collected in this study are in the same range to the ex-type isolate from Iran [26], sharing similar characteristics ( Table 2).
Ecology and host characteristics: Isolates L105, L106 (=CBS 149021), and L240 (=CBS 149020) belonging to the present species were recovered from vines of the indigenous wine grape cvs. Xynisteri and Giannoudi. Isolate L105 was co-isolated with Ph. chlamydospora.
Notes: Multigene, as well as single locus phylogeny of LSU, bt2, and tef1-a cluster extype and Spo. rosigena isolates were collected in this study in a highly supported clade. The spore dimensions and all other conidial characteristics of the obtained isolates are in the same range to the ex-type (MFLU 15-0782). Colony of Spo. rosigena isolate L106 (CBS 149021) on PDA fast-growing, flat, dark mouse gray to light sepia with smooth margin, reaching 58.25 mm in diameter after 14 days at 25 • C in darkness, conidiomata black, acervular, gregarious or confluent, erumpent. Colony on MEA fast-growing, slightly raised, wooly, grayish-white with smooth dark sienna margin, reaching in 62.25 mm in diameter after 14 days at 25 • C in darkness, conidiomata black, gregarious or confluent. Colony on OA, fast-growing, slightly raised, wooly, white with smooth olivaceous buff margin, reaching 78.75 mm diameter after 14 days at 25 • C in darkness.
Ecology and host characteristics: Isolates L105, L106 (=CBS 149021), and L240 (=CBS 149020) belonging to the present species were recovered from vines of the indigenous wine grape cvs. Xynisteri and Giannoudi. Isolate L105 was co-isolated with Ph. chlamydospora.
Notes: Multigene, as well as single locus phylogeny of LSU, bt2, and tef1-a cluster ex-type and Spo. rosigena isolates were collected in this study in a highly supported clade. The spore dimensions and all other conidial characteristics of the obtained isolates are in the same range to the ex-type (MFLU 15-0782).

Pathogenicity
After a 12-month incubation period, all Seimatosporium and Sporocadus isolates evaluated were pathogenic to 2-year-old cv. Xynisteri potted grapevines. Dark brown to black vascular discoloration developed on the wood tissue below the bark, extending upward and downward from the point of inoculation, with the mean lengths shown in Figure 7, while foliar symptoms were absent in all treatments. Although all fungal species caused lesions, the virulence varied among the genera and species, with all Sporocadus isolates being more aggressive than Seimatosporium (p < 0.0001; Figure 8). More specifically, there was no significant difference in wood discoloration among the Sporocadus isolates, ranging from 9.24 ± 1.40 to 6.9 ± 0.85 cm (0.103 = p > 0.999). Similarly, the aggressiveness of Seim. vitifusiforme and Seim. vitis-viniferae isolates were not significantly different (p > 0.999), causing lesion lengths from 4.00 ± 0.83 to 4.13 ± 0.71 cm (Figure 8). No symptoms were evident in the non-inoculated (negative control) plants; thus, they were excluded from the statistical analyses. Successful re-isolations were made only from the inoculated vines and the recovery percentages ranged from 35-67% for Seim. vitis-viniferae, 40% for Seim. cyprium, 25-32% for Spo. rosigena, and 28-44% for Spo. kurdistanicus. Retrieved isolates were confirmed with those used in grapevine inoculations based on morphology (culture and conidial characteristics). Furthermore, no fungal isolates were obtained from the negative-control plants.
After a 12-month incubation period, all Seimatosporium and Sporocadus isolates eval-uated were pathogenic to 2-year-old cv. Xynisteri potted grapevines. Dark brown to black vascular discoloration developed on the wood tissue below the bark, extending upward and downward from the point of inoculation, with the mean lengths shown in Figure 7, while foliar symptoms were absent in all treatments. Although all fungal species caused lesions, the virulence varied among the genera and species, with all Sporocadus isolates being more aggressive than Seimatosporium (p < 0.0001; Figure 8). More specifically, there was no significant difference in wood discoloration among the Sporocadus isolates, ranging from 9.24 ± 1.40 to 6.9 ± 0.85 cm (0.103 = p > 0.999). Similarly, the aggressiveness of Seim. vitifusiforme and Seim. vitis-viniferae isolates were not significantly different (p > 0.999), causing lesion lengths from 4.00 ± 0.83 to 4.13 ± 0.71 cm (Figure 8). No symptoms were evident in the non-inoculated (negative control) plants; thus, they were excluded from the statistical analyses. Successful re-isolations were made only from the inoculated vines and the recovery percentages ranged from 35-67% for Seim. vitis-viniferae, 40% for Seim. cyprium, 25-32% for Spo. rosigena, and 28-44% for Spo. kurdistanicus. Retrieved isolates were confirmed with those used in grapevine inoculations based on morphology (culture and conidial characteristics). Furthermore, no fungal isolates were obtained from the negative-control plants.

Discussion
This is the first study to explore the diversity and pathogenicity of Sporocadaceae species isolated from grapevines with GTD-related symptoms in the Mediterranean country of Cyprus. Four species of Sporocadaceae were identified by incorporating reference type and non-type sequences in multigene phylogenetic analyses. Seim. cyprium is a newly de- Figure 8. Mean lesion lengths caused 12 months post-inoculation by Sporocadaceae isolates on 2-yearold vines (cv. Xynisteri) during pathogenicity assays under field conditions. Each bar represents an individual tested isolate (n = 10), and vertical error bars indicate the corresponding standard deviation. Asterisks (****) and ns indicate the statistically significant (p < 0.0001) and non-significant differences (p < 0.05), respectively, following the analysis of variance and Tukey's mean separation test procedures.

Discussion
This is the first study to explore the diversity and pathogenicity of Sporocadaceae species isolated from grapevines with GTD-related symptoms in the Mediterranean country of Cyprus. Four species of Sporocadaceae were identified by incorporating reference type and non-type sequences in multigene phylogenetic analyses. Seim. cyprium is a newly described and typified species, whereas two species in the genus of Sporocadus: Spo. kurdistanicus and Spo. rosigena constitute new reports from Europe on grapevines. Furthermore, Seim. vitis-viniferae is a new report from Cyprus.
Conidial morphology has been extensively used to dissect coelomyetous genera [35,48]; however, phylogenetic relationships at the species level have been refined in combination with sequence data analyses [31][32][33]50]. Nevertheless, until recently, the delineation of phylogenetic lineages in Sporocadaceae was still ambiguous to a certain extent, especially due to the lack of ex-type-sequence data in public repositories and designated epitypes, as well as the deficit of more-informative loci, such as partial protein-coding regions [26,34,38].
Phylogenetic analysis based on ITS and LSU sequences grouped the collected pestalotioid isolates in the genera Seimatosporium and Sporocadus. However, tef1-a phylogeny, as well as concatenated DNA sequence datasets (ITS, LSU, bt2, and tef1-a) combined with conidial phenology were sufficient for the species identification of obtained isolates as Seim. cyrpium, Seim. vitis-viniferae, Spo. kurdistanicus, and Spo. rosigena.
A combined four-loci phylogenetic analysis showed that Seim. cyprium formed an independent, fully supported clade that was phylogenetically distinct from Seim. marivanicum (CBS 143781), Seim. luteosporum (CBS142599) and Seim. vitifusiforme (CBS 142600). Seim. cyprium can be also distinguished from the phylogenetically most closely related species Seim. marivanicum by distinct conidial morphology (conidial dimensions, number of septa, morphology, and dimensions of conidial appendages), thus it is identified as a new species. Sexual fruiting structures of Seim. cyprium were also produced and described in in vitro homothallic pairings of collected isolates.
All available ITS LSU, tub2 and tef1-a sequence data in the NCBI database, as well as those obtained presently, do not support species differentiation of Seim. vitis and Seim. vitis-viniferae, suggesting that their phylogenetic delineation is erratic. However, based on the conidial morphology of the ex-type cultures of Seim. vitis (MFLUCC 14-0051) and Seim. vitis-viniferae (CBS 123004), the two species can be clearly distinguished, suggesting that the most found Sporocadacea species in Cyprus is considered as Seim. vitis-viniferae. Previous studies [23,38,51] have described Seim. vitis in Iran and the USA, although the conidial descriptions provided do not match that of the ex-type specimen of the species, being closer to Seim. vitis-viniferae [31,34]. Furthermore, another report from Hungary has tentatively identified Seim. vitis based solely on ITS data [52], while a single study from Italy [53] reports Seim. vitis (dimensions are not provided) associated with grapevine states that the conidial morphology of the collected isolates matched the ex-type provided. Due to the abovementioned ambiguities, we recommend that the bt2 and tef1-a sequence data of the Seim. vitis ex-type should be retrieved and analyzed, while its conidial description should be also reassessed.
Two non-appendaged Sporocadaceae species in the genus of Sporocadus associated with GTDs were also recorded in our study. Spo. kurdistanicus and Spo. rosigena isolates from Cyprus formed distinct and fully supported clades (ML/MP = 100/100) that clustered with respective ex-type cultures CBS 143778 and CBS 182.50, while they also shared similar conidial characteristics.
Seimatosporium and Sporocadus consist predominately of endophytes and saprobes of woody plants [17,33,34,43], although some species are pathogens of different plant hosts, such as eucalypt, blackberry, and grapevine [13,54,55]. More specifically, out of the nine Seimatosporium species previously reported from diseased or dead grapevine wood worldwide, only five (Seim. botan, Seim. luteosporum, Seim. vitifusiforme, Seim. vitis, Seim. vitis-viniferae) have been evaluated in wood inoculations of intact grapevines to confirm pathogenicity and Koch's postulates. Seim. botan was isolated in Chile from mature vines bearing symptoms resembling those of Botryosphaeria dieback and was reported to be pathogenic on detached green shoots and rooted plants [56]. Similarly, Seim. luteosporum and Seim. vitifusiforme were first reported by Lawrence et al. (2018) in California, USA, from mature grapevines exhibiting typical dieback-type trunk disease symptoms, but due to a lack of re-isolation from inoculated vines, their pathogenic status remains unclear [38]. However, Grinbergs et al. (2021), confirmed the pathogenicity of Seim. vitifusiforme in rooted Petit Shyrah plants in Chile [37]. Seim. vitis has currently been the most reported species of the genus, found to be associated with GTD symptoms in California, USA, Hungary, Iran, and Italy, causing lesions and wood discoloration in pathogenicity assays [23,38,53,54]. The pathogenicity of Seim. vitis-viniferae was recently confirmed in Italy causing lesions of up to 24.70 cm on 1-year-old canes in an open field experiment with an 8-month incubation period [13]. Currently, there is a lack of information for Seim. hysterioides, Seim. lonicerae, Seim. marivanicum, and Seim. parasiticum. Similarly, little information is available on the pathogenicity of Sporocadus species associated with V. vinifera. Spo. kurdistanicus was recently isolated from grapevines exhibiting trunk disease symptoms in the Kurdistan province of Iran; however, it was found to be non-pathogenic on > 10-year-old vines. Furthermore, Spo. rosigena was also detected in association with GTDs in New Zealand [57], Spo. rhododendri has been isolated from V. vinifera canes in Australia [27], while Spo. lichenicola has been reported to cause blackberry cane dieback in Serbia and Oregon, USA [56,58]; however, their pathogenicity on grapevines has not been assessed yet. Herein, we present the ability of four Sporocadaceae species to cause vascular necrosis on grapevines, thus, contributing to the expanding knowledge on the agriculturally important group of GTD-related pathogens. In pathogenicity assays, isolates of all species were able to infect and colonize, and were able to produce brown to black vascular discoloration. Both Sporocadus species (Spo. kurdistanicus and Spo. rosigena) were significantly more aggressive compared to the Seimatosporium (Seim. cyprium and Seim. vitisviniferae), while there was no difference in aggressiveness among isolates of the same genus. High recoveries of all tested isolates from inoculated vines were also evident, confirming Koch's postulates.
GTDs are a disease complex resulting from interactions between taxonomically unrelated fungi colonizing grapevine wood [1,6,38]. Although limited research has been conducted on endophytes in economically important crops, particularly on their roles as latent pathogens [59], it is evident that GTD pathogens are routinely isolated from apparently healthy vines and found to express relatively long latency times in disease development [6,60,61], suggesting that some of them are likely commensal endophytes and/or latent saprobes that may act as opportunistic pathogens triggered by stressful edaphoclimatic factors or colonization density shifts of the residing microbiome [10,21,[62][63][64]. In the present study, Seimatosporium and Sporocadus isolates were retrieved from wood cankers either solely or in combination with other known trunk pathogens, such as Ph. chlamydospora, P. minimum, and B. dothidea, while they have also been recovered from asymptomatic vine tissues [11,12,43,65]. Recent studies, that investigated the effect of dual inoculations between Sporocadaceae from grapevines and known GTD pathogens, report synergistic, as well as antagonistic interactions, confirming their involvement in the GTD complex [38,65]. Since mixed fungal infections are commonly found in vineyards, it is critical to clarify the complex microbial networks and elucidate pathogenesis and symptom development [66]. Understanding GTDs' complexity will help us to prolong the sustainability and profitability of grape production, via cultural (late or double pruning) or other plant protection practices (application of plant protection products on pruning woods and the use of healthy planting material) [5,6,38,66].

Sampling and Fungal Isolation
During 2017-2018, 10 vineyards in the provinces of Limassol and Paphos, Cyprus were surveyed and sampled for trunk diseases (Supplementary Table S2). Wood samples were collected from 3 vines per vineyard of the indigenous wine grape cvs. Giannoudi, Mavro, Promara, and Xynisteri exhibiting different typical GTD symptoms, including cankers, dead cordons and spurs, and tiger-stripe foliar symptoms (Figure 9). For fungal isolation wood, segments (1-2 cm thick) were cut off, debarked, washed in distilled water (dH 2 O), and fragmented in 4-5 pieces (5 mm thick). Accordingly, discolored wood pieces were disinfected in 95% ethyl alcohol for 1 min, rinsed with sterile dH 2 O, dried off in a laminar flow cabinet, and plated on potato dextrose Agar (PDA; HiMedia) amended with streptomycin sulfate (50 µg/mL). The plates were then incubated at 25 • C in darkness for up to 1-2 weeks and inspected daily to prevent the loss of slow-growing colonies from fastgrowers. Hyphal tips of selected isolates were excised and transferred to fresh PDA plates at 25 • C to establish pure cultures that were maintained under the same conditions, and kept in an 40% aqueous glycerol solution in -80 • C. Based on colony characteristics, 16 Sporocadaceae isolates and isolates belonging to Botryosphaeriaceae, Ph. chlamydospora, and Phaeoacremonium, respectively, were used for morphological and molecular characterization [34,67,68].

DNA Isolation, PCR, and Sequencing
To avoid direct contact of the mycelium to the culture medium and its subsequent interference in the DNA extraction process, fungal cultures were grown on sterile cellophane discs (Sigma-Aldrich, St. Louis, MI, USA), which were placed on the top of the culture medium surface. Developed mycelia (approximately 14-day-old), were scraped off with a sterile spatula, lyophilized for 48 h, and homogenized with plastic micro-pestles into powder in the presence of liquid nitrogen. Total DNA was extracted according to Cary et al. (2009), following nanodrop quantification spectrophotometry [69]. The final DNA concentration of each isolate was adjusted to 20 ng/mL and stored at −20 °C for further use. The initial identification was based on sequences of ribosomal DNA fragments that included the ITS region. Furthermore, partial regions of the LSU, tef1-a, and tub2 genes were amplified and sequenced for all Sporocadaceae isolates to elucidate their phylogenetic status. The sequence data of the tub2 and tef1-a loci were used for species identification of the Botryosphaeriaceae and Phaeoacremonium isolates, while partial actin (act) gene data were also used for the latter group of isolates. PCR reactions were of 30-μL and performed using the KAPA Taq PCR kit (Sigma Aldrich, catalog no. BK1002) in a C1000TM Thermal Cycler (Bio-Rad). The PCR mixture contained 6 μL of KAPA Taq Buffer

DNA Isolation, PCR, and Sequencing
To avoid direct contact of the mycelium to the culture medium and its subsequent interference in the DNA extraction process, fungal cultures were grown on sterile cellophane discs (Sigma-Aldrich, St. Louis, MI, USA), which were placed on the top of the culture medium surface. Developed mycelia (approximately 14-day-old), were scraped off with a sterile spatula, lyophilized for 48 h, and homogenized with plastic micro-pestles into powder in the presence of liquid nitrogen. Total DNA was extracted according to Cary et al. (2009), following nanodrop quantification spectrophotometry [69]. The final DNA concentration of each isolate was adjusted to 20 ng/mL and stored at −20 • C for further use. The initial identification was based on sequences of ribosomal DNA fragments that included the ITS region. Furthermore, partial regions of the LSU, tef1-a, and tub2 genes were amplified and sequenced for all Sporocadaceae isolates to elucidate their phylogenetic status. The sequence data of the tub2 and tef1-a loci were used for species identification of the Botryosphaeriaceae and Phaeoacremonium isolates, while partial actin (act) gene data were also used for the latter group of isolates. PCR reactions were of 30-µL and performed using the KAPA Taq PCR kit (Sigma Aldrich, catalog no. BK1002) in a C1000TM Thermal Cycler (Bio-Rad). The PCR mixture contained 6 µL of KAPA Taq Buffer B (10×), 3 µL Mg 2+ , 0.6 µL of the dNTPs mixture (2.5 mmol/mL for each nucleotide), 1.5 µL of each primer (10 µM), 0.3 µL of Taq DNA polymerase (5 U/µL), 3 µL of DNA template (20 ng/µL), and 14.1 µL of sterile ddH 2 O. The primers used were ITS1 and ITS4 [70] for ITS, LROR [71] and Un-Lo28S1220 [72] for LSU, EF1-728F and EF1-986R [73] for tef1-a, T1 [74], Bt2b [75] for tub2, and ACT-512F and ACT-783R for act [73]. ITS amplification was performed using the following program: initial denaturation at 94 • C for 5 min, followed by 40 cycles of denaturation at 94 • C for 30 s, annealing at 57 • C for 30 s and extension at 72 • C for 1 min and a final extension step at 72 • C for 3 min. For LSU amplification, the n conditions were as follows: initial denaturation for 3 min at 95 • C; 35 cycles of denaturation for 30 s at 95 • C, primer annealing for 30 s at 57 • C, and extension for 1 min at 72 • C; and a final extension for 10 min at 72 • C. Cycling conditions for the rest of the loci were as described for LSU, but with at an annealing temperature of 58 • C for act and 60 • C for tef1-a and tub2, respectively. PCR amplicons were resolved on 1.5% agarose gels in Tris-acetate-EDTA buffer with a SYBR Safe DNA gel stain (Invitrogen, Carlsbad, CA, USA) and visualized under UV light. After confirmation by agarose gel electrophoresis, the PCR products were sequenced in both directions using the same primer pairs used for amplification by the Macrogen Europe B.V. (Amsterdam, The Netherlands).

Molecular Identifications and Phylogenetic Analysis
The isolates morphologically resembling typical trunk pathogens retrieved herein, such as Ph. chlamydospora, and the species belonging to Botryosphaeriaceae and Phaeoacremonium, were identified to the species level via BLASTn searches of the National Center for Biotechnology Information (NCBI), using respective sequences retrieved per taxon. Furthermore, after a literature review and NCBI BLASTn searches, ITS, LSU, tub2, and tef1-a sequences of 37 Sporocadaceae taxa (38 isolates) were retrieved and included in our phylogenetic analyses, along with the 16 Sporocadaceaea isolates obtained in this study ( Table 3).
The consensus sequences of the four individual loci (ITS, LSU, tub2, and tef1-a) were aligned with MAFFT v. 7.490 [76] using default parameters, manually adjusted using Unipro Ugene v. 43.0 [77], and individual gene sequences were concatenated using SeqKit [78]. Single locus and concatenated alignments were then subjected to a maximum-likelihood (ML) analysis using IQtree version 2.0.3 [79]. The most appropriate substitution model was chosen based on the Bayesian Information Criterion (BIC) using the ModelFinder algorithm as implemented in IQtree version 2.0.3 [80]. Moreover, branch support was obtained using the bootstrap approximation option of IQtree [81], performing 1000 bootstrap replicates. Further support to the phylogenetic inference was provided by a maximum-parsimony (MP) analysis using MPBoot [82]. Similarly, to ML, the MP analysis was performed including 1000 bootstrap replicates. Beltrania rhombica (CBS 123.58) was selected as the outgroup taxon for both the ML and MP analyses ( Table 3). The number of pairwise differences (PWD) between taxa on the concatenated sequence data was obtained using the distance estimation option implemented in MEGA X (Supplementary Table S1) [83]. New generated sequences were deposited in the NCBI GenBank database (   MH554636 MH554395 a GenBank accession numbers for the sequences of four loci: ribosomal DNA (rDNA) internal transcribed spacer region (ITS), rDNA large subunit (LSU), β-tubulin (tub2), and translation elongation factor 1-a (tef1-a) that were generated in this study (in bold) or from others. b Sequences from isolates in our collection are highlighted in bold. c CBS: Culture collection of the Westerdijk Fungal Biodiversity Institute, Utrecht, The Netherlands; CPC: Culture collection of Pedro Crous, housed at the Westerdijk Institute; MFLUCC: Mae Fah Luang University Culture Collection; NBRC: Biological Resource Center, National Institute of Technology and Evaluation, Chiba, Japan. d Status of the isolates = ET: ex-epitype; T: ex-type strain. Accession numbers in bold indicate isolates collected and characterized in the present study.

Morphological Description
Sporocadaceae isolates were morphologically described based on the colony, microscopic and stereoscopic characteristics of cultures grown on PDA, malt extract agar (MEA; Sigma-Aldrich), and oatmeal agar (OA; Sigma-Aldrich) at 25 • C for 2-4 weeks. Furthermore, 1-year-old lignified canes from the indigenous cv. Xynisteri were cut into 5 cm long segments and autoclaved twice (121 • C for 25 min, with a 24 h interval between autoclaves). The autoclaved segments were then placed in Petri dishes and immersed halfway in WA. Accordingly, three mycelial plugs from actively growing cultures were placed among the wood pieces, the cultures were incubated at 5 and 10 • C with a 12 h photoperiod and fruiting bodies formation was monitored weekly for up to 8 weeks [35]. Morphological observations of all reproductive structures were determined at appropriate magnifications using an Olympus BZX16 dissecting microscope and a Zeiss AX10 compound microscope, both equipped with color digital cameras Olympus ColorView I and Zeiss AxionCam MRc 5, respectively, and the minimum, maximum, mean, and standard deviation were calculated. The conidial length was measured from the base of the basal cell to the base of the apical appendage and the conidial width was measured to the widest point of the conidium. Furthermore, the conidial color, shape, and length of each conidial cell were also recorded. The colony morphology per culture medium was also described, while the colony colors were also rated per culture medium following Rayner's (1970) charts [84].

Effect of Temperature on Mycelial Growth
Eight isolates belonging to four different Sporocadaceae species detected herein were randomly selected for the estimation of mycelial growth cardinal temperatures. Mycelial plugs (4 mm in diameter) from the margins of actively grown cultures were transferred into Petri dishes on PDA, MEA, and OA and incubated in the darkness from 5 to 30 • C at 5 • C intervals. Two perpendicular measurements of the diameter were recorded after 14 days. Three replicate plates were used per isolate and the experiment was repeated once. The optimum temperatures for mycelial growth and the maximum daily growth rate for each isolate were calculated based on regression curves of the temperature versus daily radial growth.

Pathogenicity Tests
In May 2020, a field trial was set up to evaluate the pathogenicity of seven isolates, representative of the Seimatosporium and Sporocadus species identified by phenotypic and phylogenetic analyses. Two-year-old vines of the cv. Xynisteri were grown in 10 L pots filled with potting mix (Miskaar; Lambrou Agro Ltd., Limassol, Cyprus) and amended with a slow-release fertilizer (Itapollina 12-5-15 SK; Lambrou Agro Ltd., Limassol, Cyprus). Lignified canes were aseptically wounded by drilling between the third and the fourth internode from their base. Subsequently, a 4 mm mycelial plug from 14-day-old cultures grown on PDA was placed in the wound that was then sealed with Vaseline (Unilever PMT Ltd., Nicosia, Cyprus) and wrapped with Parafilm (Sigma Aldrich) to prevent inoculum desiccation. Negative controls were inoculated with sterile PDA plugs. Five replicates were used per isolate, with an equal number of controls in a completely randomized design and the experiment was repeated once. All plants were placed under shade netting in open-field conditions and were drip-irrigated according to the weather 1-2 times per week for 30 min (0.5 L/h). Throughout the pathogenicity trial, the inoculated stems were routinely inspected for foliar symptoms, and 12 months after inoculation, were excised and transferred to the lab for further analyses. The newly developed green shoots and the bark were removed, and the length of the wood discoloration was measured in both directions from the inoculation point. The woody canes were washed in a sodium hypochlorite solution (5%) for 2 min and then rinsed twice with dH 2 O. Accordingly, 10 small, discolored wood pieces from each stem were disinfected in 95% ethanol for 1 min, and after drying, were placed in PDA dishes amended with streptomycin (50 mg/L) and chloramphenicol (50 mg/L). Plates were incubated at 25 • C in darkness for 1-2 weeks and the re-isolated colonies were identified morphologically.

Statistical Analyses
The effect of temperature on hyphal growth and wood discoloration length for each isolate were analyzed using an ANOVA to test for the normality and homogeneity of variances. Regression curves were fitted over different temperature treatments versus growth rate for each isolate. Data were analyzed using the Kruskal-Wallis test (non-parametric ANOVA), followed by Dunn's posthoc multiple comparison test of means (p = 0.05). To assess the differences in the extent of the vascular discoloration induced by the different fungal isolates, pathogenicity data were log10 transformed to satisfy the normality requirements of the ANOVA. Bartlett's test for homogeneity of variances was also performed for repeated experiments. Since variances were homogeneous (p < 0.05), the data were combined, and the means were compared with Tukey's test at 5% of significance. All the statistical analyses were performed using SPSS (v. 25; IBM Corporation, New York, NY, USA).