What is Fusarium forma specialis?

ABSTRACT

The forma specialis concept has been a cornerstone in Fusarium phytopathology for 85 years, classifying pathogens based on host specificity. However, its validity as a natural and practical framework has been increasingly questioned. In this study, we critically re-evaluate the forma specialis concept through an extensive survey of Fusarium wilt diseases across 37 crop hosts from 23 provinces of China. Through multi-locus phylogenetic analyses, morphological assessments, and pathogenicity tests on 659 strains isolated from 171 diseased samples, we identified 46 Fusarium species, including seven newly described taxa, and uncovered extensive cross-host pathogenicity, with up to 21 species associated with the same wilt disease. In addition, 57% of samples exhibited multiple species co-infections, revealing significant inconsistencies with the forma specialis framework. These findings challenge the long-standing paradigm of host specificity in Fusarium pathogens and advocate for a shift of perspective to a “pathobiome” framework, where disease dynamics are largely driven by community-level interactions rather than single-pathogen relationships. Applying these conceptual advances to Fusarium wilt research could fundamentally transform our comprehension of host-pathogen relationships and facilitate the development of more ecologically sustainable disease management approaches.

GRAPHICAL ABSTRACT

1. Introduction

Fusarium wilt diseases pose a severe threat to a wide variety of plants, including economically important crops such as banana, cotton, cucumber, melon, strawberry, and tomato (Husaini et al. 2018; Fones et al. 2020; Henry et al. 2020a). These diseases also affect ornamental crops, such as cyclamen, gerbera, and orchids, as well as weeds and parasitic plants, including broomrape and witchweed (Leslie and Summerell 2006). The economic ramifications of these diseases are considerable, manifesting in diminished crop yields and quality, particularly in commercially significant crops (Leslie and Summerell 2006; Gordon 2017; Savary et al. 2019). Prominent examples of Fusarium wilt diseases include Panama disease (Gutiérrez-Sánchez et al. 2023), tomato wilt (Srinivas et al. 2019), strawberry wilt (Mirmajlessi et al. 2015), watermelon wilt (Martyn 2014), and bitter gourd wilt (Zang et al. 2021). These diseases are characterised by their persistence and rapid spread, often resulting in recurrent outbreaks that cause long-term agricultural damage (Gordon 2017). For instance, Panama disease has devastated the global banana industry, with historical outbreaks leading to the abandonment of numerous plantations and significant economic consequences for the affected regions (Koenig et al. 1997; Fourie et al. 2009). Similarly, Fusarium wilt in watermelon results in stunted plant growth, reduced fruit quality, lower sugar content, and significant market value loss (Martyn 2014).

The primary causal agent of these wilt diseases is F. oxysporum (Raza et al. 2017; Savary et al. 2019), a pathogen ranked fifth among the top ten scientifically and economically important plant pathogens (Dean et al. 2012). Accurate identification of F. oxysporum is critical for effective disease management. However, the taxonomic classification of this species has undergone substantial revisions, driven by evolving classification criteria that incorporate both morphological and molecular data (Schlechtendal 1824; Seaver and Wollenweber 1913; Wollenweber 1935; Leslie and Summerell 2006; Lombard et al. 2019; Crous et al. 2021). The foundational classification framework in plant pathology was proposed by Snyder and Hansen (1940), who categorised F. oxysporum based on shared morphological characteristics and applied the concept of forma specialis (f.sp.) to distinguish strains by their host-specific pathogenicity. This pragmatic framework revolutionised Fusarium research, enabling efficient identification and communication of Fusarium pathogens (Gordon 1965). As of January 2025, over 30,000 articles on F. oxysporum are indexed in the Web of Science, with approximately 130 formae speciales of F. oxysporum recorded in the Index Fungorum and MycoBank databases.

Despite its historical significance and widespread adoption, the concept of Fusarium forma specialis has faced growing challenges in recent years. Recent advances have revealed polyphyletic relationships among several formae speciales. For example, strains of F. oxysporum f.sp. niveum are present in two distinct phylogenetic lineages, with representatives of other formae speciales intermixed between these branches (Zhang and Ma 2017). Furthermore, the host specificity that underpins the concept of forma specialis has proven to be unreliable in practice (Gerlagh and Blok 1988). As summarised by Edel-Hermann and Lecomte (2019), nearly half of the formae speciales described in earlier studies exhibit a broader host range and occasional cross-pathogenicity. This ambiguity undermines quarantine measures, as many quarantine-listed formae speciales lack clear genetic markers for accurate identification, resulting in regulatory inefficiencies (Qin et al. 2024). For example, F. oxysporum f.sp. cubense, listed on China’s quarantine list, encompasses multiple phylogenetic clades (Maryani et al. 2019a). Additionally, the proposal of the “pathobiome” concept has challenged the traditional one-pathogen-one-host paradigm in Fusarium diseases of cereals. Recent studies have highlighted the role of co-infections by multiple species in Fusarium diseases of cereals, suggesting that community-level interactions may be more significant than the dynamics of a single pathogen (Han et al. 2023).

To address these challenges, we systematically collected 171 representative wilt disease samples from 23 provinces across China, encompassing 37 crop hosts from major agricultural regions. Through a comprehensive analysis integrating multi-locus phylogenetics, morphology, and pathogenicity data, we aim to critically reassess the validity of the Fusarium forma specialis concept and propose an updated framework that more accurately reflects the complex realities of Fusarium wilt diseases.

2. Materials and methods

2.1. Sample collection

Fusarium wilt samples were collected across China through collaboration between our team members and local farmers (Figure 1), who adhered to a standardised protocol outlined in our Citizen Science Initiative titled “Collecting Fusarium wilt samples”, shared via social media platforms. The samples represented a broad spectrum of host plants (Figure 1; Table S1), including species from Brassicaceae, Cucurbitaceae, Fabaceae, Solanaceae, and additional families. Geographically, a total of 171 samples exhibiting disease symptoms were gathered from 23 provinces, autonomous regions, and municipalities (Table S1). Collected samples were air dried and deposited in the Fungarium at the Institute of Microbiology, Chinese Academy of Sciences, Beijing, China (HMAS).

Figure 1.

Symptoms of Fusarium wilt from various plants. (a) Mature melon (Cucumis melo) plant exhibiting wilting and death. (b) Celery (Apium graveolens) leaves turning green to yellow. (c) Cucumber (Cucumis sativus) basal stem exhibits constriction. (d) Vascular browning observed in the stem of eggplant (Solanum melongena). (e) Discoloration of the pseudostem in banana (Musa nana). (f) Mature gourd (Lagenaria siceraria) plants showing severe wilting, eventually leading to death. (g) External wilting symptoms on yam (Dioscorea opposita) leaves. (h) External wilting symptoms on leaves of Canavalia gladiata. (i) Mature tomato (Solanum lycopersicum) plants showing wilting and death. (j) Withering of Zingiber officinale leaves and stems. (k) Yellowing and withering of Trichosanthes kirilowii leaves. (l) Withering of pepper (Capsicum annuum) leaves. (m) Kohlrabi (Brassica campestris) leaves showing yellow discoloration, with inhibited bulb formation. (n) Rubus idaeus plants exhibiting wilting and death. (o) Strawberry (Fragaria ananassa) leaves turning green to yellow. (p) Luffa (Luffa aegyptiaca) leaves displaying wilting. (q) Severe wilting observed in mature bitter gourd (Momordica charantia) plants. (r–y) Sample collection by scientific volunteers from various regions in China.

2.2. Strains isolation, preservation, and selection

A total of 750 fungal strains were isolated from infected crops exhibiting symptom, using tissue isolation, direct hyphal isolation, or single spore isolation methods as described by Han et al. (2023). Among these, 659 isolates were identified as members of the genus Fusarium based on colony morphology and translation elongation factor 1-alpha (tef1) sequences data (Figure S1; Table S1). The process for selecting representative strains followed three main steps. First, for strains exhibiting 100% tef1 sequence identity from the same symptomatic tissue of a single sample collected at one location, only one representative strain was chosen for further study. Second, genus-level phylogenetic analyses were performed using tef1 sequences, and the selected strains were grouped into six Fusarium species complexes. For each of these complexes, tef1 and RNA polymerase second largest subunit (rpb2) phylogeny were re-analysed, and within each subclade lacking sequence variation, only one isolate from the same sample was retained. Finally, multi-locus datasets were employed to amplify additional genes for detailed phylogenetic analyses of each species complex. These included rpb2, tef1, and beta-tubulin (tub2) for the F. fujikuroi species complex (FFSC); calmodulin (CaM), rpb2, and tef1 for the F. incarnatum-equiseti species complex (FIESC); CaM, rpb2, tef1, and tub2 for the F. oxysporum species complex (FOSC); histone (H3), RNA polymerase largest subunit (rpb1), rpb2, and tef1 for the F. sambucinum species complex (FSAMSC); rpb2 and tef1 for both the F. nisikadoi species complex (FNSC) and F. tricinctum species complex (FTSC). After applying these criteria, 343 representative strains were selected. Type specimens of newly described species were deposited in HMAS, while ex-type living cultures were stored at the China General Microbiological Culture Collection Centre (CGMCC). All isolates analyzed by Shiling Han (HSL) in this study were also preserved in Lei Cai’s personal culture collection (LC) at the Institute of Microbiology, Chinese Academy of Sciences, Beijing, China. Taxonomic novelties, descriptions, and nomenclature were registered in Fungal Names.

2.3. DNA extraction and amplification

Genomic DNA was extracted from freshly harvested fungal mycelia cultured on potato dextrose agar (PDA; Difco^TM, Becton, Dickinson and Company, Sparks, MD, USA) using the thermolysis method (Zhang et al. 2010). The extracted DNA samples were preserved at −20 °C until use in polymerase chain reaction (PCR) amplifications. PCR was carried out in 25 μL reaction mixtures containing 12.5 μL of 2× Taq PCR Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China), 1 μL of each primer (10 μmol/L), 2 μL of undiluted genomic DNA, and sterile distilled deionised water to adjust the volume. Six loci were amplified and sequenced: CaM (O’Donnell et al. 2000), H3 (Roux et al. 2001), rpb1 (O’Donnell et al. 2010), rpb2 (Liu et al. 1999; Reeb et al. 2004), tef1 (O’Donnell et al. 1998), and tub2 (O’Donnell and Cigelnik 1997). Details of primer pairs and PCR amplification conditions are provided in Table S2. PCR amplicons were visualised on 1% agarose gels using electrophoresis. DNA sequencing was performed by SinoGenoMax Company and Tianyi Huiyuan Company, both based in Beijing, China.

2.4. Phylogenetic analyses

Consensus sequences were generated using MEGA v. 7 (Kumar et al. 2016), and multiple sequence alignments for each genetic locus were performed with MAFFT v. 7 (Katoh and Standley 2013). Manual corrections were applied to misalignments where needed. Phylogenetic reconstructions were conducted on both individual loci and concatenated datasets using Maximum-Likelihood (ML) and Bayesian Inference (BI) approaches via the CIPRES Science Gateway platform (https://www.phylo.org/; Miller et al. 2010; Stamatakis 2014).

ML analyses were performed with RAxML-HPC v. 8.2.12 (Stamatakis 2014), using 1,000 bootstrap replicates under the GTRGAMMA + I substitution model. BI analyses were conducted with MrBayes v. 3.2.7a (Ronquist and Huelsenbeck 2003), based on the optimal nucleotide substitution models selected using MrModelTest v. 2.4 (Nylander et al. 2008) for each locus. Bayesian Markov Chain Monte Carlo (MCMC) sampling involved four parallel chains run for 10 million generations, with trees sampled every 1,000 generations. The first 25% of trees were discarded as burn-in, and the remaining trees were used to compute a 50% majority-rule consensus tree with posterior probability (PP) estimates. Clades were considered robustly supported if they exhibited RAxML bootstrap values ≥ 70% and Bayesian PP values ≥ 0.9. The resulting phylogenies were visualised using FigTree v. 1.4.2 (http://tree.bio.ed.ac.uk/software/figtree). All sequence data and alignments generated from the 659 Fusarium strains in this study have been deposited in GenBank (Table S1) and TreeBASE (32181), respectively.

Phylogenetic relationships among distinct Fusarium species complexes were carried out using multi-locus sequence datasets, following previous studies (Lombard et al. 2019; Han et al. 2023). Details regarding locus combinations, outgroup selection, number of aligned characters, and the best-fit models are summarised in Table S3. Specifically, combined datasets of rpb2, tef1, and tub2 were used for analysing the FFSC, with F. nirenbergiae CBS 744.97 serving as the outgroup. Phylogenetic analyses of the FIESC employed CaM, rpb2, and tef1 datasets, rooted with F. concolor NRRL 13459. For the FNSC, combined rpb2 and tef1 datasets were used, rooted with F. concolor NRRL 13994. The FOSC were analyzed with datasets comprising CaM, rpb2, tef1, and tub2, rooted with F. udum CBS 177.31. Phylogenetic analyses of the FSAMSC were performed by using the combined H3, rpb1, rpb2, and tef1 datasets, rooted with F. nelsonii NRRL 13338. Finally, the FTSC was examined using rpb2 and tef1 datasets, rooted with F. concolor NRRL 13459.

2.5. Morphological observation

Fungal isolates were studied based on their macroscopic and microscopic characteristics following established protocols (Crous et al. 2021). Petri plates were incubated at 25 °C for seven days. Agar plugs (approximately 5 × 5 mm) were excised from the colony margins grown on synthetic nutrient-poor agar (Nirenberg 1976) and transferred to specific media for morphological characterisation. PDA and oatmeal agar (OA) were used for observing morphological traits. Colony characteristics, including growth rate, pigmentation and odour, were assessed after seven days of incubation in the dark on PDA and OA (Crous et al. 2021). Color descriptions followed the colour charts of Kornerup and Wanscher (1978).

Detailed morphological comparisons were conducted using carnation leaf agar (CLA) as described by Fisher (1982). After seven days of incubation at 25 °C under a 12 h near-ultraviolet light/dark cycle (Leslie and Summerell 2006; Crous et al. 2021), key microscopic features, including sporodochia, conidiophores, phialides, conidia (both sporodochial and aerial), and chlamydospores, were examined. Observations were made using water-mounted specimens under a Nikon 80i compound microscope equipped with differential interference contrast (DIC) optics, and a Nikon SMZ1500 dissecting microscope. For each species, measurements were taken from 30 phialides and chlamydospores, and 50 conidia, to calculate mean, standard deviation, and range (minimum to maximum).

2.6. Diversity analyses

All statistical analyses and diversity plots were generated using R software (v. 4.1.0) (R Core Team 2021). The integration of evolutionary tree with the host information was achieved using the ggtree and treeio packages (Yu et al. 2017, 2018; Wang et al. 2020; Yu 2020), while the diversity patterns of Fusarium species across different host plants were visualised using the ggplot2 package (Wickham 2016).

2.7. Pathogenicity tests

Pathogenicity assays were carried out to assess whether Fusarium species isolated from the vascular bundles tissue of wilt plants were pathogenic. Given that Solanum lycopersicum (tomato) and Cucumis sativus (cucumber) are among the most widely cultivated crops in the families Solanaceae and Cucurbitaceae in China (http://www.stats.gov.cn/), and are significantly affected by wilt disease nationwide, Fusarium isolates from the vascular bundles tissues of these two crops were selected as representative examples for pathogenicity evaluation.

Representative isolates were inoculated onto seedlings of the Solanum lycopersicum cv. Zhongza 9 and Cucumis sativus cv. Zhongnong 6 following the methods of van Dam et al. (2016) and El Komy et al. (2021) with slight modifications. Specifically, hypha from 7-day-old colonies grown on SNA were transferred to 100 mL of carboxymethylcellulose sodium liquid medium, and cultured for seven days at 25 °C with shaking at 200 r/min. The conidial concentration (CC) of the resulting Fusarium suspension was adjusted to 5 × 10⁶ conidia/mL using a haemocytometer. Tomato and cucumber seeds were germinated in a moist chamber for 48 h and then sown in sterilised soil under controlled conditions (12 h light/12 h dark at 25 °C).

When seedlings reached the two-true-leaf stage (tomato) or two-flat-cotyledon stage (cucumber), roots were immersed in the prepared conidial suspension for 4 h at 25 °C, then transplanted back into soil (25 °C, 12 h light/12 h dark). Sterile water was used as a negative control. For each fungal strain tested, 10 plant replicates were included. Pathogenicity was assessed 10 days post-inoculation, and Koch’s postulates were confirmed by re-isolating the fungus from symptomatic plants. Disease severity was visually evaluated using a 0 to 4 scale as described by Li et al. (2020): 0: no visible symptoms; 1: 1–2 cotyledons yellowing or dropping; 2: 1–2 true leaves yellowing, or the plant exhibiting yellow-green discoloration and slight wilting; 3: whole plant showing significant wilting or severe yellowing, stunted growth, or slight dwarfing; 4: entire plant severely wilted or dead.

The disease severity index (DSI) was computed following the method of Chiang et al. (2017), using the formula: DSI (%) = [sum (class frequency × score of rating class)/(total number of plants × maximal disease severity score)] × 100%.

3. Results

3.1. Species diversity of Fusarium wilt pathogens

To elucidate the species diversity and taxonomic status associated with Fusarium wilt diseases, a citizen science initiative was launched in September 2023 to systematically survey crop wilt diseases across diverse regions of China (Figure 1). Due to the widespread and severe nature of this disease, with support from scientific volunteers nationwide, we collected 171 representative wilt samples from 37 crops across 23 provinces, municipalities, and autonomous regions in China in a limited time period (Table S1). Preliminary processing of the samples isolated 659 Fusarium strains. Phylogenetic analyses based on the tef1 locus revealed that these isolates were distributed across six Fusarium species complexes (Figure S1; Table S1).

Using the latest taxonomic system, detailed phylogenetic analyses were performed for each species complex with specific datasets. Combined with morphological characters, these isolates were identified as follows: six species in the FFSC (Figure 2), 22 species in the FIESC (Figure 3), one species in the FNSC (Figure 4), 10 species in the FOSC (Figure 5), six species in the FSAMSC (Figure 6), and one species in the FTSC (Figure 7). Notably, only 214 strains (32.5%) were classified within the FOSC, while 336 strains (51%) belonged to the FIESC, with the remaining strains distributed among the FFSC, FNSC, FSAMSC, and FTSC. In total, the 659 isolates, which, following traditional identification protocol, can be classified in 37 formae speciales of F. oxysporum, were herein identified as 46 distinct species (Table S1).

Figure 2.

Phylogeny inferred based on the combined rpb2-tef1-tub2 gene regions of the Fusarium fujikuroi species complex (FFSC). Fusarium nirenbergiae (CBS 744.97) was used as an outgroup. Strains isolated in this study were indicated in red. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type and ex-epitype strains were indicated in bold with T and ET, respectively.

Figure 3.

(Continued).

Figure 4.

Phylogeny inferred based on the combined rpb2-tef1 gene regions of the Fusarium nisikadoi species complex (FNSC). Fusarium concolor (NRRL 13994) was used as an outgroup. Strains isolated in this study were indicated in red. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type, strains were indicated in bold with T.

Figure 5.

(Continued).

Figure 6.

Phylogeny inferred based on the combined H3-rpb1-rpb2-tef1 gene regions of the Fusarium sambucinum species complex (FSAMSC). Fusarium nelsonii (NRRL 13338) was used as an outgroup. Strains isolated in this study were indicated in red. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type and ex-epitype strains were indicated in bold with T and ET, respectively.

Figure 7.

Phylogeny inferred based on the combined rpb2-tef1 gene regions of the Fusarium tricinctum species complex (FTSC). Fusarium concolor (NRRL 13459) was used as an outgroup. Strains isolated in this study were indicated in red. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type, ex-epitype, and ex-neotype strains were indicated in bold with T, ET, and NT, respectively.

Further analyses revealed exceptionally high Fusarium species diversity across most hosts (Figure 8a). For instance, 21 species were isolated from 24 tomato wilt samples (Solanaceae); 16 species from 16 cucumber wilt samples (Cucurbitaceae); and 16 species from 10 Fragaria wilt samples (Rosaceae). Notably, only eight hosts were isolated as a single Fusarium species. In contrast, from the remaining 29 hosts, at least two Fusarium species were isolated, with an average of six Fusarium species per host (Figure 8a). These findings demonstrate that most formae speciales investigated in this study are assemblages of multiple species.

Figure 8.

Diversity and composition of species causing Fusarium wilt among different hosts. (a) The number of Fusarium wilt samples and the Fusarium species isolated from each host species. (b) Detailed composition of Fusarium species among hosts with more than ten diseased samples. The number of isolates (0–7) is represented by a gradient of color intensity (light to dark).

3.2. Upturning the Fusarium forma specialis concept

On the other hand, the results demonstrated that a same Fusarium species could be isolated from two or more different hosts afflicted with wilt diseases (Figure 8b). Among the six hosts with more than ten diseased samples, F. sulawesiense and F. nirenbergiae were consistently isolated from all six hosts, i.e., Capsicum annuum (Solanaceae), Cucumis sativus (Cucurbitaceae), Fragaria ananassa (Rosaceae), Musa nana (Musaceae), Solanum lycopersicum (Solanaceae), and Solanum melongena (Solanaceae). These findings refute the prevailing notion of host specificity, which forms the theoretical basis of the Fusarium forma specialis classification system.

In addition, the results of the phylogenetic analysis indicated that the Fusarium strains isolated from a particular host did not constitute independent and exclusive branches. Rather, these strains were found to be mixed with strains isolated from other hosts (Figure 9). For instance, the Fusarium strains isolated from tomato wilt in this study were mixed with strains isolated from cucumber, strawberry, and other hosts (Figure 9). Three isolates which have been recognised as F. oxysporum f.sp. fragariae in the previous study (Henry et al. 2021) were identified as two distinct species (F. cugenangense and F. fabacearum) in the present study (Figure S2). Meanwhile, two strains of F. oxysporum f.sp. fragariae were mixed with strains of F. oxysporum f.sp. apii and F. oxysporum f.sp. cubense in the F. cugenangense clade (Figure S2). These results provided strong evidence upturning previous concept that one forma specialis causes one Fusarium wilt disease.

Figure 9.

Disparity between the Fusarium formae speciales and phylogenetic and morphogical results. The phylogenetic analyses reveals that Fusarium isolates from the same host do not form independent or exclusive clades but are intermixed with strains isolated from other hosts. The outer segments of the phylogenetic tree show the typical conidial morphologies of Fusarium species complexes. Scale bars = 10 μm.

3.3. Pathobiome for Fusarium wilt

The co-occurrence of different Fusarium species in a single diseased sample is a prevalent phenomenon (Figure 10). According to the findings of this study, 13 out of 24 samples of tomato wilt, 10 out of 16 samples of cucumber wilt, and 7 out of 10 samples of strawberry wilt showed co-infection (Figure 10). In brief, among the 171 wilt samples examined, 98 samples (57%) were confirmed for co-infection.

Figure 10.

Summary of co-infections identified in this study. White circles present the typical symptoms of Fusarium wilt on typical hosts. The heptagon represents the number of co-infected samples and average Fusarium species isolated per co-infection. The outer segments of the heptagon show the total number of diseased samples and isolated Fusarium species per host. Hosts with fewer than 10 samples are grouped as “other 31 hosts”.

To verify the co-infection phenomena in Fusarium wilt, we selected Solanum lycopersicum (Solanaceae) and Cucumis sativus (Cucurbitaceae) to conduct the pathogenicity tests for species isolated from wilt diseased crops. The results showed that, most species isolated from Fusarium wilt diseased samples have the ability to cause disease of Solanum lycopersicum and Cucumis sativus (Figures S3, S4). For example, the results of the pathogenicity test showed that 15 Fusarium species can cause wilt disease in Cucumis sativus (Figure S3), and 21 species can cause wilt disease in Solanum lycopersicum (Figure S4). Additionally, the pathogenicity of species within the FOSC did not exhibit stronger pathogenic capabilities compared to other species complexes. For instance, F. nirenbergiae belonged to the FOSC were recognised as the pathogen of cucumber wilt disease, and its pathogenicity was weaker than that of FIESC species (Figure S3). Moreover, several species, such as F. commune, F. nirenbergiae, F. paraclavum, etc., did not exhibit pathogenic differences on tomato (Figure S4).

Further analyses also showed that different species isolated from one sample were all highly pathogenic. In particular, two species (F. verticillioides and F. nirenbergiae) were isolated from one tomato wilt diseased sample (Figure 11), both of which could also cause tomato wilt disease independently. Three species, i.e., F. graminearum, F. ipomoeae, and F. luffae, were isolated from one cucumber wilt sample (Figure 11), while none of which belong to the FOSC but all capable of infecting cucumber independently. In summary, our results of the pathogenicity test well demonstrated a pathobiome infection of Fusarium wilt.

Figure 11.

Overview of field observations, disease symptoms, and inoculation tests for tomato and cucumber wilt diseases. (a) Field symptom of tomato wilt. (b) Vascular browning in the root of tomato. (c, d) Conidia of Fusarium verticillioides and F. nirenbergiae isolated from vascular bundle tissues as shown in (b), scale bars = 10 μm. (e) Blank control treated with sterile water of tomato. (f, g) Tomato seedlings inoculated with F. verticillioides and F. nirenbergiae, separately. (h) Field symptom of cucumber wilt disease. (i) Root browning in cucumber. (j–l) Conidia of F. graminearum, F. ipomoeae, and F. luffae isolated from vascular bundle tissue as shown in (i), scale bars = 10 μm. (m) Blank control treated with sterile water of cucumber. (n–p) Cucumber seedlings inoculated with F. graminearum, F. ipomoeae, and F. luffae, separately. Additional details of the inoculation tests are provided in Figures S3 and S4.

3.4. Taxonomy

Based on the phylogenetic relationships, morphological characteristics, habitat preferences, and geographical distributions, seven novel species (i.e., F. aflagelliforme S.L. Han & L. Cai, F. hordeum S.L. Han, M.M. Wang & L. Cai, F. liaoningense S.L. Han & L. Cai, F. neolongifundum S.L. Han & L. Cai, F. paraclavm S.L. Han & L. Cai, F. paraodoratissimum S.L. Han & L. Cai, F. tongrenense S.L. Han & L. Cai) are herein introduced and described. Meanwhile, based on extensive sampling and thorough sequence reconciliation, we propose the synonymisations of three currently recognised taxa (F. callistephi, F. radicigenum, and F. rhinolophicola) with previously described species, as discussed in detail below.

Fusarium incarnatum-equiseti species complex (FIESC)

Fusarium aflagelliforme S.L. Han & L. Cai, sp. nov. Figure 12

Figure 12.

Morphology of Fusarium aflagelliforme (CGMCC 3.28874, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Aerial conidiophores and phialides. (d) Aerial conidia. (e) Chlamydospores. Scale bars = 10 μm.

Fungal names: FN 572697.

Etymology: Referring to its resemblance to F. flagelliforme.

Typus: China, Inner Mongolia Autonomous Region, Tongliao City (119.68°E, 45.48°N), obtained from Cucumis sativus, 17 Sep. 2023, C.X. Li & X.R. Zhu (holotype HMAS 353949, ex-type living culture CGMCC 3.28874 = LC21447 = HSL38691NDB).

Description: Sporodochia not observed. Aerial conidiophores borne on aerial mycelium, straight or flexuous, erect or prostrate, smooth- and thin-walled, bearing terminal or lateral phialides. Aerial phialides mono- and polyphialides, subulate, subcylindrical to navicular, smooth- and thin-walled, periclinal thickening inconspicuous or absent, 8.3–28.4 × 2.3–3.1 μm. Aerial conidia falcate, the dorsal side curved than the ventral; with a blunt, conical to pointed apical cell and papillate basal cell, hyaline, smooth- and thin-walled, 1–5-septate, mainly 3-septate. 1-septate conidia: 13.2–21.9 × 2.7–4.4 μm (av. ± SD: 17.2 ± 2.9 × 3.4 ± 0.4 μm); 2-septate conidia: 16.7–27.4 × 2.5–4.3 μm (av. ± SD: 22.4 ± 2.6 × 3.4 ± 0.4 μm); 3-septate conidia: 19.4–36.7 × 2.9–4.7 μm (av. ± SD: 27.1 ± 4.2 × 3.8 ± 0.5 μm); 4-septate conidia: 25.4–43.3 × 3.7–5.4 μm (av. ± SD: 37.1 ± 3.8 × 4.4 ± 0.5 μm); 5-septate conidia: 36.2–53.1 × 3.5–5.8 μm (av. ± SD: 43.5 ± 4.8 × 4.8 ± 0.5 μm). Chlamydospores abundant, globose, subglobose to ovoid, subhyaline, smooth or rough-walled, terminal or intercalary, solitary, in pairs or forming long chains, 6.1–12.2 μm diam.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 83–85 mm diam. in 7 d; raised, felty to velvety, filamentous, with abundant aerial mycelia, colony margin entire; surface white (–A1); reverse milk white (1A2) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 79–80 mm diam. in 7 d; crateriform, circular, with abundant aerial mycelium, margin entire; surface white (–A1), reverse mimosa (2B8); odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. aflagelliforme were resolved as a strongly supported and genealogically exclusive lineage in the phylogenetic tree inferred from combined CaM, rpb2, and tef1 loci (Figure 3). Fusarium aflagelliforme is closely related to F. flagelliforme, but differs by 19 bp in the three loci datasets. Morphologically, F. aflagelliforme is distinguished from F. flagelliforme based on the morphologies of conidia (1–5-septate conidia in F. aflagelliforme vs. 3–6-septate conidia in F. flagelliforme), and the absence of sporodochia (presented in F. flagelliforme) (Xia et al. 2019).

Figure 3.

(Continued).

Fusarium compactum (Wollenw.) Raillo, Fungi of the Genus Fusarium: 180. 1950.

Basionym: Fusarium scirpi var. compactum Wollenw., Fusaria Autogr. Delin. 3: no. 924. 1930.

Synonyms: Fusarium radicigenum H. Zhang & Y.L. Jiang, Mycosphere 14(1): 2112. 2023.

Material examined: See Table S1.

Notes: O’Donnell et al. (2009) revealed the phylogenetic relationships of F. compactum, and Xia et al. (2019) proposed an epitype (CBS 186.31) for this species to stabilise the use of the name F. compactum. Zhang et al. (2023) described a new taxon F. radicigenum (CGMCC 3.25478) in the FIESC, but in their analyses they strangely excluded previously published F. compactum and F. jinanense. Here, our multi-locus phylogenetic analyses clearly showed that “F. radicigenum H. Zhang & Y.L. Jiang” (Zhang et al. 2023) clustered well within the F. compactum clade (Figure 3). Further comparisons showed that the type strains of F. compactum and “F. radicigenum” have almost identical sequences of tef1 (OR043909, GQ505648, differs 1/500 bp) and CaM (OR043754, GQ505560, differs 2/536 bp). Therefore, we treat F. radicigenum as a synonym of F. compactum.

Figure 3.

(Continued).

Fusarium hainanense M.M. Wang et al., Persoonia 43: 82. 2019.

Synonym: Fusarium rhinolophicola Karun., Tibpromma & X.F. Liu, Mycosphere 14 (1): 561. 2023.

Material examined: See Table S1.

Notes: Fusarium hainanense was described by Wang et al. (2019). Subsequently, Liu et al. (2023) introduced a closely related new species F. rhinolophicola. However, in this study we found that F. rhinolophicola published in Liu et al. (2023) did not form an exclusive clade but clustered well in line with the ex-type strain (CGMCC 3.19478 = LC11638) and other representative strains of F. hainanense (Figure 3). Further analyses showed that the type strains of these two taxa have identical sequences of CaM (OR022061, MK289657), almost identical sequences of rpb2 (OR025917, MK289735, differs 1/768 bp), and slight differences sequences of tef1 (OR026001, MK289581, differs 2/478 bp). Therefore, we consider F. rhinolophicola to be synonymous with F. hainanense.

Figure 3.

Phylogeny inferred based on the combined CaM-rpb2-tef1 gene regions of the Fusarium incarnatum-equiseti species complex (FIESC). Fusarium concolor (NRRL 13459) was used as an outgroup. Strains isolated in this study were indicated in red, while (likely) synonyms were highlighted in green. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type, ex-epitype, and ex-neotype strains were indicated in bold with T, ET, and NT, respectively.

Fusarium neolongifundum S.L. Han & L. Cai, sp. nov. Figure 13

Figure 13.

Morphology of Fusarium neolongifundum (CGMCC 3.28875, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Sporodochia. (d, e) Sporodochial conidiophores and phialides. (f) Sporodochial conidia. Scale bars = 10 μm.

Fungal names: FN 572696.

Etymology: Referring to its resemblance to F. longifundum.

Typus: China, Jiangxi Province, Jiujiang City (115.72°E, 29.71°N), obtained from Capsicum annuum, 1 Oct. 2023, J.L. Cai (holotype HMAS 353950, ex-type living culture CGMCC 3.28875 = LC21350 = HSL3776_3DB).

Description: Sporodochia milk white (1A2), formed frequently on carnation leaves or agar. Sporodochial conidiophores densely and irregularly branched, bearing apical whorls of 3–5 phialides. Sporodochial phialides subulate to subcylindrical, smooth, thin-walled, 6.3–11.8 × 2.4–4.3 μm. Sporodochial conidia falcate, curved dorsoventrally, tapering towards both ends, with an elongated or whip-like curved apical cell and a well-developed to elongate foot-shaped basal cell, hyaline, thin- and smooth-walled, 5-septate: 36.6–62.6 × 4.0–5.8 μm (av. ± SD: 52.2 ± 6.0 × 5.3 ± 0.4 μm). Aerial conidia and chlamydospores not observed.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 76–79 mm diam. in 7 d; raised, circular, with abundant aerial mycelium, colony margin entire; surface milk white (1A2); reverse flame yellow (4A8) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 69–71 mm diam. in 7 d; raised, filamentous, with abundant aerial mycelium, margin filamentous; surface white (–A1), reverse curry yellow (4C8) in the centre, cream (4A3) at the margin; odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. neolongifundum were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, and tef1 loci (Figure 3). Fusarium neolongifundum is closely related to F. longifundum, but differs by 27 bp in the three loci datasets. Morphologically, F. neolongifundum differs from F. longifundum in the conidia (5-septate conidia in F. neolongifundum vs. 3–6-septate conidia in F. longifundum), and the absence of chlamydospores (present in F. longifundum) (Xia et al. 2019).

Fusarium paraclavum S.L. Han & L. Cai, sp. nov. Figure 14

Figure 14.

Morphology of Fusarium paraclavum (CGMCC 3.28876, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Aerial conidiophores and phialides. (d) Aerial conidia. (e, f) Chlamydospores. Scale bars = 10 μm.

Fungal names: FN 572695.

Etymology: Referring to its resemblance to F. clavum.

Typus: China, Hebei Province, Chengde City (118.73°E, 41.09°N), obtained from Cucumis sativus, 14 Sep. 2023, S.Q. Ren & L.Y. Ren (holotype HMAS 353951, ex-type living culture CGMCC 3.28876 = LC21424 = HSL3855).

Description: Sporodochia not observed. Aerial conidiophores borne on aerial mycelium, straight or flexuous, erect or prostrate, smooth- and thin-walled, bearing terminal or lateral phialides. Aerial phialides mono- and polyphialides, subulate to subcylindrical, smooth- and thin-walled, periclinal thickening inconspicuous or absent, 2.7–14.7 × 2.1–2.8 μm. Aerial conidia falcate, the dorsal side curved than the ventral; with a blunt apical cell and papillate basal cell, hyaline, smooth- and thin-walled, 3–5-septate, mainly 3-septate. 3-septate conidia: 22.2–34.1 × 2.7–4.4 μm (av. ± SD: 29.3 ± 3.7 × 3.7 ± 0.5 μm); 4-septate conidia: 30.4–35.0 × 4.0–4.4 μm (av. ± SD: 32.0 ± 2.6 × 4.2 ± 0.2 μm); 5-septate conidia: 33.0–34.7 × 4.0–5.1 μm (av. ± SD: 34.0 ± 0.9 × 4.4 ± 0.6 μm). Chlamydospores abundant, globose, subglobose to ovoid, subhyaline, smooth or rough-walled, terminal or intercalary, solitary, in pairs or forming long chains, 6.6–16.9 μm diam.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 76–81 mm diam. in 7 d; flat, filamentous, aerial mycelia lightly scant, colony margin filamentous; surface pinkish white (10A2) in the centre, white (–A1) at the margin; reverse flame yellow (4A8) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 78–83 mm diam. in 7 d; flat, filamentous, margin filamentous; surface white (–A1), reverse curry yellow (4C8) in the centre, cream (4A3) at the margin; odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. paraclavum were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, and tef1 loci (Figure 3). Fusarium paraclavum is closely related to F. clavum, but differs by 20 bp in the three loci datasets. Morphologically, F. paraclavus differs from F. clavum in the aerial conidia (3–5-septate conidia in F. paraclavum vs. 1–3-septate conidia in F. clavum), and the absence of sporodochia (present in F. clavum) (Xia et al. 2019).

Fusarium tongrenense S.L. Han & L. Cai, sp. nov. Figure 15

Figure 15.

Morphology of Fusarium tongrenense (CGMCC 3.28877, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Sporodochia. (d, e) Sporodochial conidiophores and phialides. (f) Sporodochial conidia. Scale bars = 10 μm.

Fungal names: FN 572693.

Etymology: Named after the city, Tongren, where the holotype was collected.

Typus: China, Guizhou Province, Tongren City (108.07°E, 27.64°N), obtained from Nicotiana tabacum, 19 Sep. 2023, Y.F. Xu (holotype HMAS 353952, ex-type living culture CGMCC 3.28877 = LC21064 = HSL3387).

Description: Sporodochia amber yellow (4B6), formed frequently on carnation leaves, and infrequently on agar. Sporodochial conidiophores densely and irregularly branched, bearing apical whorls of 3–4 phialides. Sporodochial phialides subulate to subcylindrical, smooth, thin-walled, 8.6–15.4 × 2.8–4.4 μm. Sporodochial conidia falcate, slightly slender, straight to slightly curved dorsiventrally, with a pointed to hooked apical cell and a poorly developed foot-shaped basal cell, hyaline, thin- and smooth-walled, 1–6-septate, mainly 3–5-septate. 1-septate conidia: 12.5–20.3 × 2.5–3.3 μm (av. ± SD: 16.2 ± 3.6 × 2.9 ± 0.3 μm), 2-septate conidia: 13.6–17.2 × 2.7–3.0 μm (av. ± SD: 15.7 ± 1.5 × 2.9 ± 0.1 μm), 3-septate conidia: 16.1–34.1 × 2.9–4.9 μm (av. ± SD: 25.5 ± 4.6 × 3.8 ± 0.5 μm), 4-septate conidia: 27.8–44.4 × 3.2–4.9 μm (av. ± SD: 34.5 ± 3.6 × 4.2 ± 0.5 μm), 5-septate conidia: 34.0–48.6 × 3.6–5.4 μm (av. ± SD: 40.1 ± 2.9 × 4.5 ± 0.5 μm), 6-septate conidia: 47.5–49.1 × 5.0–5.8 μm (av. ± SD: 48.4 ± 0.7 × 5.5 ± 0.4 μm). Aerial conidia and chlamydospores not observed.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 65–66 mm diam. in 7 d; raised, rhizoid, with abundant aerial mycelium, colony margin erose; surface flame yellow (12A4) in the centre, white (–A1) at the margin, reverse greyish ruby (12D5) in the centre to white (–A1) at the margin; odour absent. On OA in the dark reaching 69–71 mm diam. in 7 d; raised, circular, with abundant aerial mycelium, margin entire; surface butter yellow (12A4) in the centre, white (–A1) at the margin, reverse orange white (5A2); odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. tongrenense were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, and tef1 loci (Figure 3). Fusarium tongrenense is closely related to F. citri, F. weifangense, and F. caulendophyticum, but differs by 39 bp, 34 bp, and 30 bp in the three loci datasets, respectively. Morphologically, F. tongrenense differs from F. citri and F. weifangense in the number of septa in sporodochial conidia (1–6-septate in F. tongrenense vs. 3–5-septate in F. citri vs. 3–7-septate in F. weifangense) (Wang et al. 2019; Han et al. 2023), from F. caulendophyticum in the presence of sporodochia (absence in F. caulendophyticum) (Zhang et al. 2023).

Fusarium oxysporum species complex (FOSC)

Fusarium fabacearum L. Lombard et al., Persoonia 43: 24. 2018 [2019].

Material examined: See Table S1.

Notes: Fusarium fabacearum was introduced by Lombard et al. (2019), while F. wimaladesilvae was published by Tan et al. (2024). However, Tan et al. (2024) did not provide morphological description and only provided sequences of two loci (accession OR781002 for rpb2, and accession OR824385 for tef1) for F. wimaladesilvae. Sequence comparison revealed identical rpb2 sequences (MH484939, OR781002) between F. fabacearum (ex-type strain CBS 144743) and F. wimaladesilvae (ex-type strain BRIP 70752a), with differentiation based solely on tef1 sequence (MH485030, OR824385), showing 99% identity (612/615 base pairs). In this study, a larger sampling of F. fabacearum for the multi-locus phylogenetic analysis showed that the ex-type isolate (BRIP 47195a) of F. wimaladesilvae clustered within the F. fabacearum clade (Figure 5). Although phylogenetic and sequence similarity analyses strongly suggest that F. wimaladesilvae is a late synonym of F. fabacearum, we retain the name pending further taxonomic evaluation with morphology and additional loci.

Figure 5.

(Continued).

Fusarium hordeum S.L. Han, M.M. Wang & L. Cai, sp. nov. Figure 16

Figure 16.

Morphology of Fusarium hordeum (CGMCC 3.28878, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Aerial conidiophores and phialides. (d) Aerial conidia. (e) Chlamydospores. Scale bars = 10 μm.

Fungal names: FN 572690.

Etymology: Named after the host, Hordeum sp., from which the holotype was collected.

Typus: Australia, intercepted and isolated at Ningbo Customs, from Hordeum vulgare imported to China, unknown date, W.J. Duan (holotype HMAS 353953, ex-type living culture CGMCC 3.28878 = LC13745).

Description: Sporodochia not observed. Aerial conidiophores borne on aerial mycelium, straight or flexuous, erect or prostrate, smooth- and thin-walled, bearing terminal or lateral phialides. Aerial phialides mono- and polyphialides, subulate, subcylindrical to navicular, smooth- and thin-walled, periclinal thickening inconspicuous or absent, 4.5–18.0 × 2.3–4.1 μm. Aerial conidia falcate, the dorsal side curved than the ventral; with a blunt, conical to pointed apical cell and papillate basal cell, hyaline, smooth- and thin-walled, 0–3-septate: aseptate conidia: 5.7–11.8 × 2.2–4.0 μm (av. ± SD: 7.8 ± 1.5 × 3.1 ± 0.5 μm); 1-septate conidia: 11.1–21.9 × 2.6–4.7 μm (av. ± SD: 16.6 ± 3.0 × 3.5 ± 0.5 μm); 2-septate conidia: 18.3–27.0 × 3.2–5.1 μm (av. ± SD: 21.8 ± 2.8 × 4.0 ± 0.6 μm); 3-septate conidia: 22.8–37.9 × 3.1–4.9 μm (av. ± SD: 30.0 ± 3.9 × 4.1 ± 0.4 μm). Chlamydospores abundant, globose, subglobose to ovoid, subhyaline, rough-walled, terminal or intercalary, solitary, in pairs or forming long chains, 6.8–12.6 μm diam.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 79–82 mm diam. in 7 d; raised, rhizoid, with abundant aerial mycelia, colony margin lobate; surface pastel pink (11A4); reverse milk white (1A2) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 74–78 mm diam. in 7 d; raised, circular, with abundant aerial mycelium, margin entire; surface white (–A1), reverse mimosa (2B8); odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. hordeum were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, tef1, and tub2 loci (Figure 5). Fusarium hordeum is closely related to F. fabacearum and F. gossypinum, but differs from F. fabacearum in the absence of sporodochia (present in F. fabacearum), from F. gossypinum in the presence of abundant chlamydospores (absence in F. gossypinum) (Lombard et al. 2019).

Figure 5.

Phylogeny inferred based on the combined CaM-rpb2-tef1-tub2 gene regions of the Fusarium oxysporum species complex (FOSC). Fusarium udum (CBS 177.31) was used as an outgroup. Strains isolated in this study were indicated in red, while (likely) synonyms were highlighted in green. The RAxML bootstrap support values (RAxML-BS ≥  70%) and Bayesian posterior probabilities (BI-PP ≥  0.9) were displayed at the nodes (RAxML-BS/BI-PP). Ex-type and ex-epitype strains were indicated in bold with T and ET, respectively.

Fusarium inflexum R. Schneid., in Schneider & Dalchow, Phytopathol. Z. 82: 80. 1975.

Synonym: Fusarium callistephi L. Lombard & Crous, Persoonia 43: 15. 2018 [2019].

Material examined: See Table S1.

Notes: The strains of F. inflexum were not included in the analysis of Lombard et al. (2019) when F. callistephi was introduced. Our study reveals that these two taxa clustered in an indistinguishable terminal clade (Figure 5). Further analyses showed that the type strains of these two taxa had identical sequences of CaM (AF158366, MH484693), rpb2 (JX171583, MH484875), and tef1 (AF008479, MH484966). Given F. inflexum (pub. 1975) was published an earlier name than F. callistephi (pub. 2019), thus the latter was treated as a synonym.

Fusarium liaoningense S.L. Han & L. Cai, sp. nov. Figure 17

Figure 17.

Morphology of Fusarium liaoningense (CGMCC 3.28879, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Sporodochial conidiophores and phialides. (d) Aerial conidia. (e–g) Chlamydospores. Scale bars = 10 μm.

Fungal names: FN 572688.

Etymology: Named after the province, Liaoning, where the holotype was collected.

Typus: China, Liaoning Province, Shenyang City (123.36°E, 42.03°N), obtained from Phaseolus vulgaris, 14 Sep 2023, H.Y. Mao (holotype HMAS 353954, ex-type living culture CGMCC 3.28879 = LC21151 = HSL3576).

Description: Sporodochia not observed. Aerial conidiophores borne on aerial mycelium, straight or flexuous, erect or prostrate, smooth- and thin-walled, bearing terminal or lateral phialides. Aerial phialides mono- and polyphialides, subulate, subcylindrical to navicular, smooth- and thin-walled, periclinal thickening inconspicuous or absent, 2.5–20.2 × 2.2–4.6 μm. Aerial conidia falcate, the dorsal side curved than the ventral; with a blunt, conical to pointed apical cell and papillate basal cell, hyaline, smooth- and thin-walled, 0–4-septate. aseptate conidia: 4.9–12.4 × 2.1–3.6 μm (av. ± SD: 8.8 ± 2.1 × 2.7 ± 0.4 μm); 1-septate conidia: 9.5–21.6 × 2.8–4.0 μm (av. ± SD: 14.8 ± 3.9 × 3.3 ± 0.5 μm); 2-septate conidia: 23.8–26.9 × 4.1–4.5 μm (av. ± SD: 24.8 ± 1.4 × 4.3 ± 0.2 μm); 3-septate conidia: 26.0–36.4 × 3.8–5.8 μm (av. ± SD: 30.5 ± 3.1 × 5.0 ± 0.4 μm); 4-septate conidia: 38.2–39.6 × 5.2–5.4 μm (av. ± SD: 38.9 ± 0.7 × 5.3 ± 0.1 μm). Chlamydospores abundant, globose, subglobose to ovoid, subhyaline, smooth or rough-walled, terminal or intercalary, solitary, in pairs or forming long chains, 6.3–14.7 μm diam.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 68–69 mm diam. in 7 d; raised, rhizoid, with abundant aerial mycelia, colony margin erose; surface pink (12A4); reverse milk white (1A2) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 66–69 mm diam. in 7 d; raised, rhizoid, with abundant aerial mycelium, margin filamentous; surface purplish white (14A2) in the centre, white (–A1) at the margin; reverse mimosa (2B8); odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. liaoningense were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, tef1, and tub2 loci (Figure 5). This species failed to produce any sporodochia on the carnation leaf pieces, but produced abundant 3–4-septate conidia on the aerial mycelium, which is a unique character in the F. oxysporum species complex (Maryani et al. 2019a; Lombard et al. 2019).

Fusarium odoratissimum Maryani et al., Stud. Mycol. 92: 159. 2019.

Material examined: See Table S1.

Notes: The ex-type strains of F. cili (HGUP 190024), F. odoratissimum (INACC F822), F. phialophorum (INACC F971), and F. rosae-roxburghii (HGUP 190111) (Maryani et al. 2019a; Zhang et al. 2023) cluster together in a well-supported clade within the F. oxysporum species complex (Figure 5). Their rpb2 and tef1 sequences share high similarity (99.8%), and their rpb1 sequences show approximated 98% similarity. F. odoratissimum (INACC F822) and F. phialophorum (INACC F971), published in Maryani et al. (2019a), did not provide CaM and tub2 sequences. Although our phylogenetic and sequence similarity analyses strongly suggest that F. cili, F. rosae-roxburghii, and F. phialophorum are late synonyms of F. odoratissimum, we decided to leave these names for further taxonomic evaluation when additional loci are sequenced.

Fusarium paraodoratissimum S.L. Han & L. Cai, sp. nov. Figure 18

Figure 18.

Morphology of Fusarium paraodoratissimum (CGMCC 3.28880, ex-type culture). (a) Colony on PDA. (b) Colony on OA. (c) Sporodochia. (d) Sporodochial conidiophores and phialides. (e) Sporodochia conidia. (f) Aerial conidiophores and phialides. (g) Aerial conidia. (h, i) Chlamydospores. Scale bars = 10 μm.

Fungal names: FN 572686.

Etymology: Referring to its resemblance to Fusarium odoratissimum.

Typus: China, Jiangsu Province, Xuzhou City (117.47°E, 34.4°N), obtained from Fragaria ananassa, 13 Sep. 2023, F.J. Wang (holotype HMAS 353955, ex-type living culture CGMCC 3.28880 = LC21097 = HSL3528).

Description: Sporodochia champagne (4B4), formed frequently on carnation leaves, and infrequently on agar. Sporodochial conidiophores densely and irregularly branched, bearing apical whorls of 3–4 phialides. Sporodochial phialides subulate to subcylindrical, smooth, thin-walled, 5.4–16.4 × 2.5–3.6 μm. Sporodochial conidia falcate, slightly slender, straight to slightly curved dorsiventrally, with a pointed to hooked apical cell and a poorly developed foot-shaped basal cell, hyaline, thin- and smooth-walled, 3–4-septate. 3-septate conidia: 26.6–41.3 × 3.7–5.8 μm (av. ± SD: 32.2 ± 3.6 × 4.5 ± 0.5 μm), 4-septate conidia: 36.5–40.7 × 4.3–4.9 μm (av. ± SD: 39.0 ± 1.8 × 4.7 ± 0.3 μm). Aerial conidiophores borne on aerial mycelium, straight or flexuous, erect or prostrate, smooth- and thin-walled, bearing terminal or lateral phialides. Aerial phialides mono- and polyphialides, subulate to subcylindrical, smooth- and thin-walled, periclinal thickening inconspicuous or absent, 6.1–14.1 × 2.4–4.7 μm. Aerial conidia obovoid to reniform, hyaline, smooth- and thin-walled, 0–1-septate. Aseptate conidia: 5.1–8.7 × 2.1–3.2 μm (av. ± SD: 6.3 ± 0.8 × 2.6 ± 0.3 μm); 1-septate conidia: 10.6–20.5 × 2.5–4.4 μm (av. ± SD: 14.9 ± 3.1 × 3.4 ± 0.5 μm). Chlamydospores abundant, globose, subglobose to ovoid, subhyaline, smooth or rough-walled, terminal or intercalary, solitary, in pairs or forming long chains, 4.9–11.9 μm diam.

Culture characteristics: Colonies on PDA incubated at 25 °C in the dark reaching 66–69 mm diam. in 7 d; convex, rhizoid, with abundant aerial mycelium, colony margin erose; surface milk white (1A2); reverse milk white (1A2) in the centre, white (–A1) at the margin; odour absent. On OA in the dark reaching 59–72 mm in 7 d; convex, circular, with abundant aerial mycelium, margin entire; surface white (–A1), reverse mimosa (2B8); odour absent.

Other materials examined: See Table S1.

Notes: The isolates representing F. paraodoratissimum were resolved as a strongly supported and genealogically exclusive lineage in the phylogeny inferred from combined CaM, rpb2, tef1, and tub2 loci (Figure 5). Fusarium paraodoratissimum is closely related to F. odoratissimum, but differs by 16 bp in rpb2 combined tef1 sequences. Morphologically, F. paraodoratissimum could be distinguished from F. odoratissimum in the number of septa in sporodochial conidia (3–4-septate in F. paraodoratissimum vs. 0–6-septate in F. odoratissimum), and the number of septa in aerial conidia (0–1-septate in F. paraodoratissimum vs. 0–3-septate in F. odoratissimum) (Maryani et al. 2019a).

4. Discussion

4.1. Revisiting Fusarium forma specialis: insights from large-scale Fusarium wilt analyses

The concept of forma specialis, officially named “spezialisierten formen”, was initially applied to distinguish strains of Puccinia graminis with similar morphology but distinct host preferences (Eriksson 1894). In 1910, special forms (forma specialis, f.sp.) was proposed as an intraspecific rank for parasitic fungi at the 3rd International Botanical Congress (Proceedings of the 3rd International Botanical Congress, 1910, Brussels Chapter II, Recommendation I bis). While officially recognised by the International Code of Botanical Nomenclature in 1964, the concept remained loosely defined due to its reliance on physiological rather than morphological traits (Gordon 1965). Snyder and Hansen (1940) subsequently applied this concept to F. oxysporum, identifying 25 “biologic forms” based on host ranges, and by 1965, these were officially termed formae speciales. This concept has since become a cornerstone in plant pathology, where researchers classify Fusarium strains into many formae speciales based on morphology features and host specificity, facilitating studies in phytopathology (Gordon and Martyn 1997; Edel-Hermann and Lecomte 2019; Fayyaz et al. 2023).

Although the forma specialis concept has historically facilitated Fusarium research (Edel-Hermann and Lecomte 2019; Patel et al. 2023), it was developed during an era with limited molecular techniques and sparse data on Fusarium diversity and pathogenicity (Figure S5). Modern molecular and genomic analyses have exposed significant ambiguities in classifications based solely on morphology and host specificity. For instance, recent studies reveal polyphyly within each forma specialis, as demonstrated for F. oxysporum f.sp. apii, F. oxysporum f.sp. asparagi, and F. oxysporum f.sp. cubense (Qin et al. 2024). Additionally, many formae speciales exhibit broader host ranges than previously assumed. Edel-Hermann and Lecomte (2019) reviewed 106 documented formae speciales and found that nearly half lacked strict host specificity, occasionally exhibiting cross-pathogenicity. Our large-scale analysis of 659 Fusarium isolates from 37 host species further underscores the extensive phylogenetic diversity within Fusarium formae speciales. Nearly 80% of host plants examined contained two or more Fusarium species (Figure 8a), with an average of six Fusarium species per host. Pathogenicity tests confirmed that most species isolated from diseased samples exhibited pathogenic ability (Figures S3, S4). Collectively, these findings upturned the validity of forma specialis as intraspecific rank in Fusarium. Instead, Fusarium forma specialis function as general assemblages for a group of pathogens associated with certain hosts, lacking the taxonomic rigor required for precise classifications.

4.2. Re-examining quarantine-related Fusarium formae speciales: a call for updates

Global pathogen invasions present significant threats to ecosystems, agriculture, and economies, making effective quarantine measures essential (Capinha et al. 2015; Pyšek et al. 2020). Fusarium pathogens are prominent on quarantine lists, with at least 36 Fusarium species and 71 F. oxysporum formae speciales identified as quarantine targets by 81 countries (https://www.casbrc.org/pqfungi/globalQuarantine). However, most formae speciales, rooted in host-specificity hypothesis, have proven polyphyletic, undermining their reliability for quarantine purposes (Kim 1993; Gordon and Martyn 1997; O’Donnell et al. 1998, 2004; Lievens et al. 2009; van Dam et al. 2016).

For instance, the 2021 “List of Quarantine Pests of Import Plants in the People’s Republic of China” includes five Fusarium formae speciales, namely F. oxysporum f.sp. apii, F. oxysporum f.sp. asparagi, F. oxysporum f.sp. cubense, F. oxysporum f.sp. elaeidis, and F. oxysporum f.sp. fragariae. However, based on nomenclature, morphological characteristics, and molecular phylogenetic analysis, Qin et al. (2024) confirmed that F. oxysporum f.sp. elaeidis actually refers to F. elaeidis, while, the other four formae speciales do not constitute monophyletic group. For example, three isolates previously identified as F. oxysporum f.sp. apii (Henry et al. 2020b), were reclassified in this study as two distinct species, F. cugenangense and F. curvatum (Figure S2). Additionally, two F. oxysporum f.sp. apii strains were found intermixed with F. oxysporum f.sp. fragariae and F. oxysporum f.sp. cubense within the F. cugenangense clade (Figure S2).

Further evidence from our study and related works shows that quarantine-significant pathogens like F. oxysporum f.sp. cubense, once thought confined to the FOSC, actually appear across FFSC, FIESC, and FSSC complexes (Maryani et al. 2019b). Our findings from samples collected from Musa nana identified 19 distinct species in four species complexes contributing to the same disease (Figure 9). This broad genetic diversity within each forma specialis challenges traditional taxonomic boundaries and complicates current quarantine practices. To enhance quarantine efficacy, Fusarium forma specialis must integrate molecular diagnostics and phylogenetic insights, refining classifications to reflect true genetic and ecological relationships. Such updates would improve pathogen identification, safeguarding agriculture and ecosystems against complex multi-species threats.

4.3. Fusarium pathobiome: moving beyond host-specific forma specialis

The Fusarium pathobiome concept, proposed by Han et al. (2023), emphasises that host-pathogen dynamics are shaped by community-based interactions rather than being limited to single-pathogen and host-specific relationships. Our large-scale investigation revealed that 57% of wilt samples involved co-infections by different Fusarium species (Figure 10). Pathogenicity tests further demonstrated that many Fusarium species, historically regarded as host-specific, could induce wilt across different crops (Figures S3, S4). These findings reinforce the pathobiome framework as a more accurate paradigm for understanding Fusarium wilt than the traditional forma specialis model.

This pathobiome perspective highlights ecological flexibility and adaptive infection strategies in Fusarium: when crop rotation between maize and tomato, previously non-dominant pathogens (e.g., F. nirenbergiae) adapted to maize shifted to tomato, becoming highly pathogenic. Such ecological adaptability negates the assumption of strict host-pathogen specificity. Although some species exhibit low pathogenicity to particular hosts (Figures S3, S4), their broad host ranges serve as a survival strategy, enabling them to switch hosts as needed. This adaptability poses challenges for disease management strategies, particularly those relying on crop rotation or resistance breeding, which are theoretically grounded in host-specific interactions (Edel-Hermann and Lecomte 2019).

Adopting the pathobiome perspective is essential for advancing the management of Fusarium wilt diseases. This holistic view, which considers the interactions among multiple species within the disease context, has also been proven beyond Fusarium. For example, studies on Cercospora leaf spot in sugar beet (Beta vulgaris) and common bean have uncovered substantial cryptic diversity and the involvement of multiple co-occurring Cercospora species in disease development (Vaghefi et al. 2018; Bakhshi et al. 2021). Such findings challenge the traditional single-pathogen disease paradigm, instead supporting an emerging understanding where pathogenicity may emerge from synergistic interactions among closely related fungal taxa. Applying these conceptual advances to Fusarium wilt research could fundamentally transform our comprehension of host-pathogen dynamics and facilitate the development of more ecologically sustainable disease management approaches.

Funding Statement

The work was supported by the National Natural Science Foundation of China [U24A20343], the Key R&D Programs of Xizang Autonomous Region in China [XZ202201ZY0011N], and the Survey of Wildlife Resources in Key Areas of Xizang (Phase II) [2023JXAUHX186].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

Shiling Han: conceptualisation, validation, formal analysis, resources, data curation, writing – original draft, writing – review & editing, visualisation. Peng Zhao: writing – review & editing, funding acquisition. Mengmeng Wang: resources, writing – review & editing. Xiaoyu Jiang: validation. Fang Liu: resources, writing – review & editing, funding acquisition. Junen Huang: resources. Lei Cai: conceptualisation, resources, writing – review & editing, supervision, project administration, funding acquisition. All authors read and approved the final manuscript.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21501203.2025.2528352