Molecular Variability of the Fusarium solani Species Complex Associated with Fusarium Wilt of Melon in Iran

Species of the Fusarium solani species complex (FSSC) are responsible for the Fusarium wilt disease of melon (Cucumis melo), a major disease of this crop in Iran. According to a recent taxonomic revision of Fusarium based primarily on multilocus phylogenetic analysis, Neocosmospora, a genus distinct from Fusarium sensu stricto, has been proposed to accommodate the FSSC. This study characterized 25 representative FSSC isolates from melon collected in 2009–2011 during a field survey carried out in five provinces of Iran. Pathogenicity assays showed the isolates were pathogenic on different varieties of melon and other cucurbits, including cucumber, watermelon, zucchini, pumpkin, and bottle gourd. Based on morphological characteristics and phylogenetic analysis of three genetic regions, including nrDNA internal transcribed spacer (ITS), 28S nrDNA large subunit (LSU) and translation elongation factor 1-alpha (tef1), Neocosmospora falciformis (syn. F. falciforme), N. keratoplastica (syn. F. keratoplasticum), N. pisi (syn. F. vanettenii), and Neocosmospora sp. were identified among the Iranian FSSC isolates. The N. falciformis isolates were the most numerous. This is the first report of N. pisi causing wilt and root rot disease in melon. Iranian FSSC isolates from different regions in the country shared the same multilocus haplotypes suggesting a long-distance dispersal of FSSC, probably through seeds.


Introduction
Melon (Cucumis melo L.) is one of the most economically important horticultural crops among the Cucurbitaceae and comprises diverse varieties, such as C. melo L. var. cantalopensis Naudin (cantaloupe) and C. melo L. var. indorus Naudin (long melon). The top 10 melon producing countries are China, Turkey, India, Kazakhstan, and Iran in Asia, Egypt in Africa, Spain in Europe, United States, Guatemala, and Mexico in America [1]. Iran, where melon cultivation probably dates back more than 5000 years [2], is the fifth producing country in the world. The oldest archaeological finds of melon crop are from China and Iran and are seeds dating back to 3000 B.C. [3]. Consequently, Iran was proposed as a putative center of domestication of melon [4,5]. Like other crops, melon is affected by several fungal and oomycete diseases that reduce yield and fruit quality [6]. Fusarium oxysporum f. sp. melonis [7][8][9][10], Phytophthora melonis [11], Monosporascus cannonballus [12], Paramyrotheium foliicola [13], Plectosphaerella cucumerina [14], and Neoscytalidium hyalinum [15] are among the most important fungal pathogens reported from melons in Iran.
The Fusarium solani species complex (FSSC) comprises filamentous fungi with a worldwide distribution, which causes disease in many economically important crops [16,17]. As a whole, FSSC has a wide host range of over 100 agricultural crops and typically causes foot and root rot in host plants [16][17][18]. The infected plants show wilting, stunting, chlorosis,

Morphological Characterization
In order to study the in vitro pigmentation and growth rates of isolates, singleconidium subcultures were grown on fresh PDA dishes and incubated under alternating dark and light with a 12-h photoperiod at 25 • C for two weeks. For microscopic examination, all strains were grown on carnation leaf-piece agar (CLA) [48,49], potassium chloride agar (KCL-Agar) [50], and spezieller nahrstoffarmer agar (SNA) [51] Petri dishes, which were incubated at 25 • C for 4 days (KCL-Agar dishes) and also for 14 days (CLA and SNA dishes) under alternating dark and light with a 12-h photoperiod. Colony and conidia characteristics including the colony growth rate, pigmentation, mean size of 30 randomly selected well-developed macroconidia, and number of septa were recorded. Finally, the descriptions by Summerbell and Schroers [33], Leslie and Summerell [52], Nalim et al. [28], and Short et al. [32] were used for morphological species determination.

Pathogenicity Tests
Overall, 25 isolates were evaluated for their pathogenicity on their host of origin. Moreover, four of these isolates, Iv-Km50, Iv2r30, Far-317, and FS-Spa, were assessed for their pathogenicity on 11 different cucurbit varieties and species, including different varieties of melon, i.e., Garmak-Ahmadabadi (Cucumis melo var. reticulatus), Shahd-e-Shiraz (C. melo var. cantalopensis), Kharboz-e-Mashhadi (C. melo var. indorous), Snake melon (C. melo var. flexusus), Til-Mashhad (C. melo L.), and Semsoori (C. melo L.), as well as five other species of cucurbits, cucumber (C. sativus L.), watermelon (Citrullus lanatus L.), zucchini (Cucurbita pepo L.), pumpkin (C. moschata L.), and bottle gourd (Lagenaria siceraria L.). Inoculum was prepared by adding five discs (4 mm diameter) from 7-day-old Fusarium colonies grown on PDA to flasks containing sterile wheat seeds as described by Sabahi et al. [13]. Sterile agar plugs were added to wheat seeds to inoculate control plants. Flasks were incubated at room temperature for 15 days to ensure complete colonization of the grains. Inoculation of 15-day-old cucurbit seedlings was performed using a procedure described previously [13,53]. Inoculated plants were kept under greenhouse conditions (25 • C and 65% relative humidity) for symptoms' development until 30 days post-inoculation (dpi). To determine the virulence and pathogenicity of FSSC isolates, the seedlings were examined 30 dpi; they were cut at cotyledon level, and the symptom severity (SS) was scored according to a scale from 0 to 5 as described by Nagao et al. [54]. Nine plants of each host were inoculated per fungal isolate, and the same number of plants inoculated with water was used as a control. SS measurements were converted from original ordinal scale to ratio scale and normalized (from 0 to 1) using the following formula: SS = [(0 × n 0 ) + (1 × n 1 ) + (2 × n 2 ) + (3 × n 3 ) + (4 × n 4 ) + (5 × n 5 )]/N × M i where n 0 , n 1 , n 2 , n 3 , n 4 , and n 5 are the number of symptomatic plants per each scale level (from 0 to 5), N is total number of plants examined, and M i is the highest score scale. SS data were analyzed using analysis of variance (ANOVA), and differences between the strains were analyzed using the GLM procedure of the SPSS software (SPSS, version 16). Tukey's test was used for pairwise mean comparisons.
After symptom scoring, re-isolation was performed from symptomatic seedlings using PDA. The fungal isolates were identified based on morphological characteristics and sequencing of three genetic regions (ITS, LSU, and tef1) to fulfill Koch's postulates. The experiments were repeated once with similar results.
Newly obtained sequences were blasted against databases available at BLAST [58] on NCBI-GenBank database. Three loci in a set of worldwide FSSC strains were retrieved from the GenBank database and included in the phylogenetic analysis (Table S1). Sequences were aligned using the CLUSTAL W program, and concatenated following the alphabetic order of the genes, ending in a sequence of 1778 base pairs: nucleotides 1 to 667 for ITS, 668 to 1111 for LSU, and 1112 to 1778 for tef1.
Phylogenetic trees were constructed using the concatenated sequences of three loci via maximum likelihood with MEGA 6.06 [59]. The best model of evolution was determined using the Modeltest option from MEGA 6.06, and the phylogenetic tree was constructed with bootstrapping (1000 replications). Fusarium staphyleae (NRRL 22316) was used as an outgroup, and the final tree was drawn using infix pdf editor [60].

Genetic Diversity and Haplotype Network Analysis
Nucleotide diversity, number of haplotypes, haplotype frequency, haplotype diversity, number of segregating sites, number of mutations, percentage of polymorphism site, and the minimum number of recombination events were estimated using DnaSP v. 5.10 software for the sequences of each gene as well as concatenated sequences of FSSC isolates. The class I neutrality test (Tajima's D, Fu and Li'D * , and Fu and Li's F * statics) were also calculated for detecting departure from the mutation-drift equilibrium [61]. NeighborNet networks were constructed, and the pairwise homoplasy index (PHI-Test) was estimated for each gene region and the combined data set by SplitsTree v. 4.18.2 [62]. To see the phylogeographic relationship between the FSSC isolates, a haplotype network was generated for the concatenated sequences data set using TCS (Tata Consultancy Services) algorithm [63], which is implemented in Population Analysis with Reticulate Trees (PopART v. 1.7) [64]. The geographic origin of FSSC isolates investigated in this study and the FSSC strains retrieved from the GenBank database (Table 1) were allocated into the haplotype network as described by Leigh and Bryant [64]. A total of 41 Fusarium-like isolates were recovered from symptomatic melo of two varieties, C. melo L. var. cantalopensis Naudin (cantaloupe) and C. melo L. dorous Naudin (long melon). All isolates were morphologically identified as F. sol Leslie and Summerell [51]. Among the FSSC isolates examined in this study, 15 were from Khorasan and Semnan provinces, respectively, seven and five from F Yazd provinces, respectively, and four from Isfahan province.
Based on the geographical origins, host, and morphological characteristics lates representing the overall diversity of the original set of isolates were selecte depth investigation (Table 1). An isolate of FSSC from Spain (FS-Spa) was include study as a reference isolate. According to the sequencing data of three genetic regi LSU, and tef1) and morphological characteristics, four phylogenetic species wer fied among the 25 Iranian isolates (Tables 1 and 2, Figure S1). Of these, F. falcifor Neocosmospora falciformis) (18 isolates), F. keratoplasticum (syn. N. keratoplastica) isolate), and F. vanettenii (syn. N. pisi) (five isolates) are known species. One isolate reference isolate from Spain were of an undescribed phylogenetic species of F sensu lato, tentatively named FSSC 5 or Neocosmospora sp. A total of 41 Fusarium-like isolates were recovered from symptomatic melon plants of two varieties, C. melo L. var. cantalopensis Naudin (cantaloupe) and C. melo L. var. inodorous Naudin (long melon). All isolates were morphologically identified as F. solani sensu Leslie and Summerell [51]. Among the FSSC isolates examined in this study, 15 and 10 were from Khorasan and Semnan provinces, respectively, seven and five from Fars and Yazd provinces, respectively, and four from Isfahan province.
Based on the geographical origins, host, and morphological characteristics, 25 isolates representing the overall diversity of the original set of isolates were selected for in depth investigation (Table 1). An isolate of FSSC from Spain (FS-Spa) was included in this study as a reference isolate. According to the sequencing data of three genetic regions (ITS, LSU, and tef1) and morphological characteristics, four phylogenetic species were identified among the 25 Iranian isolates (Tables 1 and 2, Figure S1). Of these, F. falciforme (syn. Neocosmospora falciformis) (18 isolates), F. keratoplasticum (syn. N. keratoplastica) (a single isolate), and F. vanettenii (syn. N. pisi) (five isolates) are known species. One isolate and the reference isolate from Spain were of an undescribed phylogenetic species of Fusarium sensu lato, tentatively named FSSC 5 or Neocosmospora sp.
Significant differences between distinct phylogenetic species were noticed in some morphometric and cultural characteristics, such as the shape of macroconidia and the growth rate on PDA (Table 2). In particular, the shape index (length to width ratio) of macroconidia of F. falciforme isolates ranged from 7.9 to 8.3, with a mean ± SD of 8.1 ± 0.12, while the same index for macroconidia of F. venattenii isolates ranged from 10.2 to 10.6, with a mean ± SD of 10.4 ± 0.23. The value of shape index of macroconidia of the only F. keratoplasticum isolate was 7.9, while the corresponding values for the two FSSC 5 isolates (Tay-r2-r, from Iran, and FS-Spa, from Spain) were 7.9 and 7.0, respectively. The growth rate on PDA of F. falciforme isolates ranged from 7.0 to 9.0 mm/day, with a mean ± SD of 8.1 ± 0.75, while the growth rate of F. venattenii isolates ranged from 5.0 to 6.0 mm/day, with a mean ± SD of 5.6 ± 0.42 mm/day. The growth rate of the F. keratoplasticum isolate was 8.5 mm/day, while the corresponding values for the two FSSC 5 isolates (Tay-r2-r, from Iran, and FS-Spa, from Spain) were 8.5 and 8.0 mm/day, respectively. Other morphological characteristics overlapped among different phylogenetic species or were not enough discriminant, as they showed a great intraspecific variability.

Pathogenicity and Host Range of FSSC Strains
Under greenhouse conditions, all 25 Iranian isolates and the reference isolate from Spain evaluated in this study were shown to be pathogenic on the host plant from which they were isolated; symptoms of wilting and crown-and root-rot were observed in artificially inoculated plants ( Figure S1). In the host range assays, performed with four strains, each representing a distinct phylospecies, all melon varieties were severely affected by Iv2-r-30 and Iv-km-50 isolates, with SS values around 1, while the other cucurbit crops were less severely affected with the least severe symptoms being observed on zucchini plants with an SS value of 0.7 (Figure 2a).
The symptoms induced by Far-317 and FS-Spa isolates on six varieties of melons were less severe than those induced by Iv2-r-30 and Iv-km-50 isolates. Conversely, the degree of aggressiveness of FS-Spa strain on cucumber, watermelon, zucchini, pumpkin, and bottle gourd plants was higher than the other FSSC isolates evaluated in this study. Among all four isolates tested, Far-317 was the least aggressive on cucurbits (Figure 2b).
. Fungi 2023, 9, x FOR PEER REVIEW 8 of 19 Significant differences between distinct phylogenetic species were noticed in some morphometric and cultural characteristics, such as the shape of macroconidia and the growth rate on PDA (Table 2). In particular, the shape index (length to width ratio) of macroconidia of F. falciforme isolates ranged from 7.9 to 8.3, with a mean ± SD of 8.1 ± 0.12, while the same index for macroconidia of F. venattenii isolates ranged from 10.2 to 10.6, with a mean ± SD of 10.4 ± 0.23. The value of shape index of macroconidia of the only F. keratoplasticum isolate was 7.9, while the corresponding values for the two FSSC 5 isolates (Tay-r2-r, from Iran, and FS-Spa, from Spain) were 7.9 and 7.0, respectively. The growth rate on PDA of F. falciforme isolates ranged from 7.0 to 9.0 mm/day, with a mean ± SD of 8.1 ± 0.75, while the growth rate of F. venattenii isolates ranged from 5.0 to 6.0 mm/day, with a mean ± SD of 5.6 ± 0.42 mm/day. The growth rate of the F. keratoplasticum isolate was 8.5 mm/day, while the corresponding values for the two FSSC 5 isolates (Tay-r2-r, from Iran, and FS-Spa, from Spain) were 8.5 and 8.0 mm/day, respectively. Other morphological characteristics overlapped among different phylogenetic species or were not enough discriminant, as they showed a great intraspecific variability.

Pathogenicity and Host Range of FSSC Strains
Under greenhouse conditions, all 25 Iranian isolates and the reference isolate from Spain evaluated in this study were shown to be pathogenic on the host plant from which they were isolated; symptoms of wilting and crown-and root-rot were observed in artificially inoculated plants ( Figure S1). In the host range assays, performed with four strains, each representing a distinct phylospecies, all melon varieties were severely affected by Iv2-r-30 and Iv-km-50 isolates, with SS values around 1, while the other cucurbit crops were less severely affected with the least severe symptoms being observed on zucchini plants with an SS value of 0.7 (Figure 2a).
(a) The symptoms induced by Far-317 and FS-Spa isolates on six varieties of melons were less severe than those induced by Iv2-r-30 and Iv-km-50 isolates. Conversely, the degree of aggressiveness of FS-Spa strain on cucumber, watermelon, zucchini, pumpkin, and bottle gourd plants was higher than the other FSSC isolates evaluated in this study. Among all four isolates tested, Far-317 was the least aggressive on cucurbits (Figure 2b).
The inoculated fungi were re-isolated from symptomatic, artificially infected plants on PDA medium and their identity was confirmed by their morphological characteristics on SNA, CLA, and KCL-Agar media, as well as the sequencing of three genetic regions (ITS, LSU, and tef1). The negative (mock-inoculated) control plants remained healthy and did not develop any symptoms. The same results were observed in a replication of the pathogenicity and host range assays.

Phylogenetic Analyses
Sequence analysis of three genetic regions (ITS, LSU, and tef1) was conducted on all 25 selected Iranian FSSC isolates and the reference isolate from Spain. The nucleotide sequences of FSSC isolates were deposited in GenBank database (see Table 1).
Based on BLASTn searches using the sequence of tef1 gene on the NCBI GenBank database, the 26 strains were identified as either F. falciforme, F. keratoplasticum, F. vanettenii, or an undescribed species belonging to the FSSC. Overall, 17 of the 25 isolates from Iran clustered with the F. falciforme group with high bootstrap support (86%). Fusarium falciforme strains are divided into three subclusters, two of which, subcluster I and III, were found in Iran (Figure 3). The isolates of subcluster I showed two and three nucleotide differences in the sequences of ITS and tef1 gene regions, respectively, compared to the The inoculated fungi were re-isolated from symptomatic, artificially infected plants on PDA medium and their identity was confirmed by their morphological characteristics on SNA, CLA, and KCL-Agar media, as well as the sequencing of three genetic regions (ITS, LSU, and tef1). The negative (mock-inoculated) control plants remained healthy and did not develop any symptoms. The same results were observed in a replication of the pathogenicity and host range assays.

Phylogenetic Analyses
Sequence analysis of three genetic regions (ITS, LSU, and tef1) was conducted on all 25 selected Iranian FSSC isolates and the reference isolate from Spain. The nucleotide sequences of FSSC isolates were deposited in GenBank database (see Table 1).
Based on BLASTn searches using the sequence of tef1 gene on the NCBI GenBank database, the 26 strains were identified as either F. falciforme, F. keratoplasticum, F. vanettenii, or an undescribed species belonging to the FSSC. Overall, 17 of the 25 isolates from Iran clustered with the F. falciforme group with high bootstrap support (86%). Fusarium falciforme strains are divided into three subclusters, two of which, subcluster I and III, were found in Iran (Figure 3). The isolates of subcluster I showed two and three nucleotide differences in the sequences of ITS and tef1 gene regions, respectively, compared to the sequence of the subcluster III isolates. However, no nucleotide differences in the sequences of the LSU region between isolates of subclasters I and III were observed. The phylogenetic tree further showed a strongly-supported relationship (96% bootstrap support) between F. vanettenii (NRRL 22820 and NRRL 22278) obtained from GenBank and six isolates from melon plants in Iran. In addition, the tree showed that one isolate from Iran (Tay-r2-r) and the reference isolate from Spain (FS-Spa) belong to the undescribed species FSSC 5 (Neocosmospora sp.) with high bootstrap value (95%), while another Iranian isolate (Iv-km-50) was placed in F. keratoplasticum with strong phylogenetic affinity (100% bootstrap support) (Figure 3). Similar results were obtained when the sequences of gene tef1 were analyzed separately ( Figure S3), while results of the phylogenetic analysis of the ITS and LSU regions were not consistent ( Figures S2 and S4). Based on ITS maximum likelihood phylogeny, all F. falciforme isolates were placed in one group, while F. solani f. sp. robiniae and F. petroliphilum isolates from GenBank clustered together. As well, F. vanettenii and F. solani f. sp. mori isolates retrieved from GenBank clustered in one group, and Iranian isolates clustered in a separate group ( Figure S2). In the LSU-based maximum likelihood phylogeny, all F. falciforme isolates clustered together, while F. vanettenii isolates clustered with the F. solani f. sp. mori and F. solani f. sp. robiniae isolates retrieved from GenBank ( Figure S4). phylogenetic tree further showed a strongly-supported relationship (96% bootstrap support) between F. vanettenii (NRRL 22820 and NRRL 22278) obtained from GenBank and six isolates from melon plants in Iran. In addition, the tree showed that one isolate from Iran (Tay-r2-r) and the reference isolate from Spain (FS-Spa) belong to the undescribed species FSSC 5 (Neocosmospora sp.) with high bootstrap value (95%), while another Iranian isolate (Iv-km-50) was placed in F. keratoplasticum with strong phylogenetic affinity (100% bootstrap support) (Figure 3). Similar results were obtained when the sequences of gene tef1 were analyzed separately ( Figure S3), while results of the phylogenetic analysis of the ITS and LSU regions were not consistent ( Figures S2 and S4). Based on ITS maximum likelihood phylogeny, all F. falciforme isolates were placed in one group, while F. solani f. sp. robiniae and F. petroliphilum isolates from GenBank clustered together. As well, F. vanettenii and F. solani f. sp. mori isolates retrieved from GenBank clustered in one group, and Iranian isolates clustered in a separate group ( Figure S2). In the LSU-based maximum likelihood phylogeny, all F. falciforme isolates clustered together, while F. vanettenii isolates clustered with the F. solani f. sp. mori and F. solani f. sp. robiniae isolates retrieved from GenBank ( Figure S4).

Genetic Diversity
The FSSC isolates recovered from melons in Iran carried different allelic forms and sequence types (STs) and corresponded to 10 multilocus haplotypes (MHs) based on the concatenated sequences of the three gene regions. Six, one, two, and one MHs belonged to F. falciforme, F. keratoplasticum, F. vanettenii, and FSSC 5 isolates, respectively (Table 1; Figure 4). As for the individual gene regions, 4, 10, and 5 STs were detected for the ITS, tef1 and LSU, respectively, in the Iranian isolates while 8, 13, and 4 STs were identified for these gene regions in the FSSC isolates from other countries (Table 1; Figure S5). Sequence variation statistics and diversity parameters were estimated by DnaSP v. 5.10 software for three individual gene regions, as well as concatenated sequences in all evaluated FSSC isolates in this study (Table 3). Similarly, the diversity parameters were calculated among the F. falciforme isolates recovered in Iran and compared with F. falciforme isolates from other countries ( Table 3). As for the F. falciforme isolates from Iran, there were 1, 6, and 2 STs in the sequences of ITS, tef1, and LSU gene regions, respectively, while 1, 1, and 1 STs were found among the F. keratoplasticum isolates, 1, 2, and 1 STs among the F. vanettenii isolates, and 1, 1, and 1 STs among the isolates of the undescribed species FSSC 5 for the sequences of ITS, Tef-1α, and LSU gene regions, respectively (Table 1; Figure S5). Based on the concatenated sequences of three gene regions, the haplotype frequency (HF) and haplotype diversity (HD) indices were 0.352, and 0.721, respectively, for F. falciforme isolates from Iran and 0.857, and 0.952, respectively, for F. falciforme isolates from other countries, indicating higher genetic diversity among isolates from other countries (Table 3).
quencing data of three gene regions (ITS, LSU, and tef1). Scale bar indicates number of substitutio per site. Bootstrap values higher than 70 are shown. Fusrium staphyleae (NRRL 22316) was used the outgroup. Red dots indicate Iranian isolates recovered in this study.

Genetic Diversity
The FSSC isolates recovered from melons in Iran carried different allelic forms a sequence types (STs) and corresponded to 10 multilocus haplotypes (MHs) based on t concatenated sequences of the three gene regions. Six, one, two, and one MHs belong to F. falciforme, F. keratoplasticum, F. vanettenii, and FSSC 5 isolates, respectively (Table Figure 4). As for the individual gene regions, 4, 10, and 5 STs were detected for the IT tef1 and LSU, respectively, in the Iranian isolates while 8, 13, and 4 STs were identified these gene regions in the FSSC isolates from other countries (Table 1; Figure S5). Sequen variation statistics and diversity parameters were estimated by DnaSP v. 5.10 software three individual gene regions, as well as concatenated sequences in all evaluated FS isolates in this study (Table 3). Similarly, the diversity parameters were calculated amo the F. falciforme isolates recovered in Iran and compared with F. falciforme isolates fro other countries ( Table 3). As for the F. falciforme isolates from Iran, there were 1, 6, and STs in the sequences of ITS, tef1, and LSU gene regions, respectively, while 1, 1, and 1 S were found among the F. keratoplasticum isolates, 1, 2, and 1 STs among the F. vanette isolates, and 1, 1, and 1 STs among the isolates of the undescribed species FSSC 5 for t sequences of ITS, Tef-1α, and LSU gene regions, respectively (Table 1; Figure S5). Based the concatenated sequences of three gene regions, the haplotype frequency (HF) and ha lotype diversity (HD) indices were 0.352, and 0.721, respectively, for F. falciforme isola from Iran and 0.857, and 0.952, respectively, for F. falciforme isolates from other countri indicating higher genetic diversity among isolates from other countries (Table 3).  (Table 1), while the numbers on the right of the slash indica the number of isolates in a given MH.  (Table 1), while the numbers on the right of the slash indicates the number of isolates in a given MH. Most isolates from different regions of Iran were placed in separate MHs compared to isolates from other countries. Only one isolate, Iv-km-50 (F. keratoplasticum), recovered in Khorasan (Iran), shared the same MH as isolate NRRL 32780 from USA ( Figure 4). Some MHs were represented by several Iranian isolates. For instance, isolates Se-r-19, Iv-k-21, Yazd-m-23, Tj-90, Tj-3, Kht-r-f1, Kno-2, and Kho-r2-b of F. falciforme belonged to the same MH. Conversely, several MHs were unique and represented by only a single Iranian isolate. This was the case of isolate Tk-rs-1, Khaf-400, Ka-s-82, and Ga-r-30 of F. falciforme, and isolate Tay-r2-r of FSSC 5 ( Figure 4; Table 1).
A phylogenetic network was constructed by the NeighborNet method using the concatenated sequences of three gene regions. While the minor reticulations observed in the NeighborNet network indicate possible recombination events within the Iranian FSSC population, the PHI-Test did not find statistically significant evidence for recombination either for each gene region or in concatenated gene regions. The results of PHI-Test were in general accordance with those of the DnaSP results and did not show statistically significant evidence for recombination. Population neutrality indices (i.e., Fu and Li' D*, and Fu and Li's F*) were significantly negative for the LSU gene region of F. falciforme isolates (Table 3), indicating a recent selective sweep and/or population expansion after a recent bottleneck. However, considering the nonsignificant results obtained in sequences of ITS, and tef1 gene regions, further investigation by sequences of additional gene regions is needed to confirm these observations.

Discussion
This study discloses the genetic variability, potential host range, and geographic distribution of the FSSC population associated with Fusarium wilt of melon in Iran. Field surveys for three consecutive years (2009)(2010)(2011) showed the disease occurred across five major melon-producing provinces of the country. Both molecular and morphological data were used to identify species of the FSSC recovered from symptomatic melon plants. In an earlier phylogenetic study, it was shown that DNA sequences of the LSU, ITS and tef1 gene regions can resolve evolutionary relationships within the FSSC [28]. Therefore, in the present study the combined data set of these three loci was used to identify FSSC isolates from melon at the species level. FSSC strains have been previously divided into three distinct subgroups, termed clades 1, 2, and 3 [25][26][27]. All Iranian FSSC isolates from melons, as well as the FS-Spa reference isolate from Spain, were found to be members of clade 3. Phylogenetic analyses of the concatenated three gene regions revealed that Iranian isolates grouped into five lineages, three of which had been identified earlier as F. vanettenii, F. keratoplasticum, and the undescribed species FSSC 5. According to the taxonomic criterion proposed by Sandoval-Denis and Crous [36] and Sandoval et al. [37], assigning the whole FSSC to Neocosmospora as a genus distinct from Fusarium s.s., the three above mentioned species should be named N. pisi, N. keratoplastica, and Neocosmospora sp., respectively, while F. falciforme is a synonym of N. falciformis. Here, both nomenclatural criteria have been interchangeably adopted to bypass the dispute concerning the taxonomy of the genus Fusarium and to make it easier to compare results from studies of diverse authors. The majority of Iranian FSSC isolates from melon fell into F. falciforme (FSSC 3 + 4) (syn. N. falciformis) and grouped into two subclusters. Similar results were obtained when sequences of tef1 were used individually for phylogenetic analyses, while all Iranian F. falciforme isolates were grouped into only one cluster based on phylogenetic analysis of ITS and LSU sequences. In ITS and LSU-based phylogenetic analysis, the F. vanettenii isolates clustered together with the isolates of F. solani f. sp. mori and F. solani f. sp. robiniae, indicating these gene regions were not suitable to distinguish these lineages.
The F. falciforme and F. vanettenii phylogenetic species showed distinct cultural and morphometric characteristics, such as the growth rate on agar medium and the shape of macroconidia. Conversely, no significant difference in these morphological traits was noticed between the F. falcforme isolates and either the F. vanettenii or the FSSC 5 isolates.
All FSSC isolates evaluated in this study were pathogenic on melon plants. Although N. keratoplastica and N. falciformis have been reported previously to cause root rot of muskmelons in Spain [65,66], to our knowledge, this is the first report of F. vanettenii and FSSC 5 causing wilt and root rot of melons in Iran. Fusarium vanettenii (synonym: Neocosmospora pisi), formerly Fusarium solani f. sp. pisi, is a recently recognized species in the FSSC [38]. This fungus has been reported as a destructive pathogen of legumes in different parts of the world, including Canada, Czech Republic, India, Iran, New Zealand, Southern Scandinavia, United Kingdom, and USA [47,67]. Šišić et al. [68] demonstrated this pathogen has a broad host range and raised doubts about the definition of it as a forma specialis. Consistently with this hypothesis, F. vanettenii was recently reported as a causal agent of tomato root rot in India [69]. Results of the present study confirm the host range of this species also encompasses non-leguminous hosts. It is the first time N. pisi is reported as a pathogen of melon worldwide. Members of FSSC 5 have been mostly reported as clinical opportunistic pathogens associated with infectious diseases of humans and animals [22]. Moreover, this phylogenetic species is a soil inhabitant and has been reported to be associated to the dry rot of potato [31,70]. In the present study, the FSSC 5 isolates were proved to be pathogenic on melon plants and other cucurbits, further expanding the known, already broad, host range of this Fusarium lineage. The melon varieties tested in this study, all widely grown in Iran, were susceptible to FSSC isolates. In greenhouse assays, besides melon, other crops of the Cucurbitaceae family, including watermelon, cucumber, zucchini, pumpkin, and bottle gourd, showed wilting and root rot symptoms when inoculated with FSSC isolates recovered from melon plants with natural infections. Although there was a noticeable variability in virulence among the isolates, no differences were observed in their host range on cucurbits.
FSSC species are considered cosmopolitan pathogens, as they occur in all climatic regions [71,72]. However, they prefer tropical hot areas [73,74]. In all provinces surveyed in this study, melon crops were in plains with warm to hot springs and summers and were planted in spring (March to April), so climatic conditions were relatively uniform. Despite this, the incidence of the disease was higher in Khorasan and Semnan. Moreover, a certain geographical structure in the Iranian FSSC population associated to Fusarium wilt of melon seems to exist. However, these aspects need to be further investigated to be confirmed.
A total of 24 MHs were identified among the FSSC isolates analyzed in this study; nine of them were only found in Iranian isolates, and one was shared between Iranian isolate and isolates from other country. Haplotypes MH2 and MH22 were only found in the Khorasan province, the first producer of melon in Iran with more than 48% of the national production [75].
Overall, 12 MHs were found among the isolates of F. falciforme (syn. N. falciformis). According to Posada and Crandall [76], the most frequent haplotype is probably the oldest in a given population. MH4 was the most common and widely distributed haplotype among Iranian F. falciforme isolates and included eight strains from five different provinces. The wide geographic distribution of this haplotype suggests effective mechanisms of dispersion over long distances. Very probably, in melon crops fungi of the FSSC have been transmitted prevalently by seeds, as F. solani s.l. is known to be a seed-borne pathogen [77]. A patchy distribution of symptomatic plants in surveyed melon crops, which is typical of seed-borne pathogens, would reinforce this hypothesis.
Interestingly, haplotype MH13 included two isolates of F. keratoplasticum (syn. N. keratoplastica) recovered from melon in Iran and sea turtles in the USA. This is not so surprising as several taxa of the FSSC have been associated with both clinical infections and plant diseases [41,65]. For instance, N. keratoplastica, a relevant pathogen of animals (including humans), has been recently reported as causal agent of wilt and root rot of muskmelon and watermelon crops [65]. Similarly, N. petroliphila (syn. F. petroliphilum), responsible for human keratitis, has been formerly known among plant pathologists as F. solani f. sp. cucurbitae race 1, a causal agent of fruit, stem, and root rot of cucurbits [65,78]. Neocosmospora falciformis (syn. F. falciforme), which is frequently responsible for clinical infections on humans, has been reported as a pathogen of several host plants, including species of the Cucurbitaceae family, confirming a wide host and ecological range of members of the FSSC [41,66].

Conclusions
This study provides preliminary information on the genetic variability of FSSC populations associated with Fusarium wilt of melon in Iran. Four phylogenetic species, including N. falciformis (syn. F. falciforme), N. pisi (syn. F. vanettenii), N. keratoplastica (syn. F. keratoplasticum), and Neocosmospora sp. (FSSC 5), and 10 diverse haplotypes have been identified in a set of isolates collected in major melon producing provinces of the country. Isolates of diverse genotypes differed in virulence, but all were pathogenic on the most common melon varieties grown in Iran and on a wide range of other cucurbits. The diversity of FSSC genotypes associated with Fusarium wilt of melon as well as their broad host range and ecological plasticity have implications for epidemiology, disease management strategies, breeding programs for disease resistance, and quarantine measures. The study of the genetic variability of FSSC populations would benefit from the use of additional markers. Moreover, targeting pathogenesis-related genes could provide a better insight into the biology and epidemiology of members of this complex. Focusing on taxa with a very broad host range that comprises both plant and animal pathogens might be helpful to understand the pathogenesis mechanisms of these fungi and in particular genetic determinants of both their polyphagia and ability to switch from a saprophytic to a parasitic lifestyle.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jof9040486/s1, Table S1: Details on FSSC isolates and GenBank accessions for their sequence data. Figure S1: Symptoms of wilt (a), crown (b,c), and root (d) rot observed in pathogenicity test of FSSC isolate Iv2-r-30 on Cucumis melo plants 30 days after inoculation. Colony morphology of FSSC isolates grown on PDA for two weeks under 12-h alternating dark and light at 25 • C, front (left) and back (right) side, macroconidia and microconidia of FSSC isolates: F. keratoplasticum (isolate Iv-km50) (e-g), F. falciforme (isolate Iv2-r-30) (h-j), F. vanettenii (isolate Far-317) (k-m), and FSSC 5 (isolate FS-Spa) (n-p). (Scale bars = 10 µm). Figure S2: Maximum likelihood phylogeny of FSSC isolates recovered from melons in Iran based on ITS sequencing data. Scale bar indicates number of substitutions per site. Bootstrap values higher than 70 are shown. Fusarium staphyleae (NRRL 22316) was used as the outgroup. Red dots indicate strains isolated in this study. Figure S3: Maximum likelihood phylogeny of FSSC isolates recovered from melons in Iran based on tef1sequencing data. Scale bar indicates number of substitutions per site. Bootstrap values higher than 70 are shown. Fusarium staphyleae (NRRL 22316) was used as the outgroup. Red dots indicate strains isolated in this study. Figure S4: Maximum likelihood phylogeny of FSSC isolates recovered from melons in Iran based on LSU sequencing data. Scale bar indicates number of substitutions per site. Bootstrap values higher than 70 are shown. Fusarium staphyleae (NRRL 22316) was used as the outgroup. Red dots indicate strains isolated in this study. Figure S5: TCS multilocus haplotype network generated using the POPArt program from the partial sequences of three gene regions (ITS, LSU, and tef1) in FSSC isolates recovered from melons in Iran. The size of the circles indicates the relative frequency of sequences of a given haplotype. Hatch marks along the branches indicate the number of mutations. Each color represents one of the five provinces where the FSSC isolates were recovered. The isolates of F. falciforme, F. keratoplasticum, F. vanettenii, and FSSC 5 are framed by a blue, black, red, and grey boxes, respectively. The isolates retrieved from GenBank and the reference isolate FS-Spa received from Spain are assigned to the category 'Abroad'. The numbers of the left of a slash indicates the MH number (Table 1), while the numbers of the right of the slash indicates the number of isolates in a given MH.