The accurate diagnosis of fungi is a prerequisite for many scientific fields, such as agriculture, ecology and medicine to enable rapid and efficient therapies, control strategies and to decipher the role of environmental and human effects on shaping the population structures of both beneficial and pathogenic fungal species. Despite the large number of species concept definitions in both plants and fungi (De Quieroz 2007; Giroud et al. 2008; Aldhebiani 2018; Jayawardena et al. 2021), the accurate identification through their morphological characters remains critically important. In their Handbook of Fungi (2007), Webster and Weber emphasized the importance of microscopic features in the identification of fungal species, before moving on to DNA identification, especially to introduce students to the ‘art’ of examining and identifying fungi. This is particularly true for fungal genera such as Fusarium and Trichoderma, or oomycetes like Phytophthora, consisting of many species that have been subjected to numerous revisions since they were first established. Species of the fungal genus Fusarium are ubiquitous, many of them are pathogenic to several important crops; in particular, two of them, F. graminearum and F. oxysporum, have been listed among the 10 most important fungal pathogens in plant pathology (Dean et al. 2012). Knowledge of the morphology of Fusarium is critical to an understanding of the disease cycle of Fusarium plant pathogens. For example, knowledge of the difference between thick walled chlamydospores and thickened hyphal and conidial walls is important to understand the persistence of some Fusarium pathogens such as F. pseudograminearum which causes crown rot of wheat and other cereals. This species does not form chlamydospores, but some isolates will form conidia and hyphae with slightly thickened cell walls in culture and soil (Wearing and Burgess 1977). This species persists primarily as hyphae in infested residues. Consequently, this is a key factor to be considered when developing disease management strategies such as the nature of crop rotations. The role of conidia in the biology and ecology of many Fusarium species, especially those not known to be pathogenic to plants or animals, is still poorly understood. For example, we do not know the role of the large globose conidia of F. beomiforme which form in culture. We do not know if they are formed in soil or in colonised plant material in nature. The morphological definition of a species is not a simple task, and it is ruled by the International Code of Nomenclature for algae, fungi, and plants (ICNafp, or simply, the Code) (https://www.iapt-taxon.org/nomen/main.php), revised on a six-year frequency at Nomenclature Section meetings of each International Botanical Congress (IBC), through the provision of detailed and precise formal requirements (Aime et al. 2021). The recent rapid and high throughput molecular identification of fungi and oomycetes has prompted proposals for the simplification of taxonomy, through the adoption of a “one name = one fungus” approach (Taylor 2011; Wingfield et al. 2012). This has also been applied to the genus Fusarium (Geiser 2013). Both morphological and molecular identification of fungi are useful tools: morphological and cellular structures often have direct biological significance, vital for understanding the survival, reproduction, infection and natural history of the organism. Nonetheless, DNA approaches are very good at identifying fixed allelic differences that are characteristic of the genetic isolation associated with species boundaries and for inferring evolutionary relationships among species. The importance of Fusarium molecular identification based on the analysis of several genes, such as the translational elongation factor 1-α (tef-1α), RNA polymerase 1 and 2 (RPB1 and RPB2), B tubulin (tub) or the nuclear ribosomal DNA intergenic spacer region (IGS rDNA), is well acknowledged (Geiser et al. 2004; O’Donnell et al. 2009, 2010, 2021). However, these analyses are costly and require specialized personnel that many laboratories cannot afford, especially those located in less developed countries. Therefore, morphological identification still has its importance, at least to provide an initial clue to the species, be subsequently confirmed by molecular approaches. The authors also consider that the requirement to include morphological details and illustrations, in addition to molecular data, in first reports of fungal pathogens or oomycetes in a country or world-wide, contribute to supporting the authenticity of the report. This can be important from an economic perspective as ‘First Reports’ may have a significant impact on trade in agricultural products. In recent years, different tools, using morphology as an important discriminant character, have been developed for the hierarchical classification of organisms like: PENIMAT (Kozakiewicz et al. 1993), and PENNAME (Pitt 1990) for Penicillium spp.; IdPhy (https://idtools.org/id/phytophthora/index.php) and FungId (Bouket et al. 2012) for oomycetes; XPER for nematodes (Palomares-Rius et al. 2022). A tool for Fusarium identification (FusKey) was developed using morphological features as discriminant characters. However, it covered only 17 Fusarium species and was not able to differentiate important Fusarium species such as F. graminearum, F. culmorum and F. crookwellense (syn: F. cerealis) because the author did not use macroconidia as a discriminatory character (Thrane 1991). More recently, new web-based technological tools, like FRIDA (Martellos 2010), have been implemented to generate identification keys for different organisms (Varese et al. 2010).

Conventional Fusarium identification procedures based on morphology are commonly undertaken by consulting several published synoptic keys (Gerlach and Nirenberg 1982; Nelson et al. 1983; Burgess et al. 1988, 1994; Leslie and Summerell 2006). The aim of FusaHelp is to facilitate the correct preliminary morphological identification of the most widespread Fusarium species by an open access, web-assisted interactive site, easily upgradable, and enriched by illustrations of the most relevant species. To generate FusaHelp, we have used LucidBuilder (https://www.lucidcentral.org/lucid-builder/), a platform based on JavaScript running on an independent browser application which runs on Windows, OSX and Linux and no additional software or plug-ins to run are required. The first step for the creation of FusaHelp was the thorough analysis of the literature concerning the morphological descriptions of new Fusarium species, along with those described in the above-mentioned synoptic keys. Where discordances among authors were encountered, or the description of some characters was missing, we examined cultures of the “true” species from International Culture Collections and confirmed by molecular analyses with tef-1α sequence. We performed a hierarchical clustering of the 82 Fusarium species of the dataset, using a binary distance matrix based on the Jaccard distance of the presence or absence of a certain character. The dendrogram was drawn with R version 4.2.1 (R Core Team 2022) using the “dendextend” package version 1.16.0 (Galili 2015). Except for F. graminearum/ F. pseudograminearum and F. equiseti/F. compactum, the hierarchical clustering performed on the entire dataset confirmed the unicity of the remaining species based on the combination of the morphological features observed (Fig. S1).

In developing FusaHelp, the following characters have been considered: the presence or absence of microconidia and their shape, the mode of formation of microconidia (single, in false heads or in chains); the type of conidiophores (monophialides/polyphialides); the presence or absence of the macroconidia and their shape (characteristic of the dorsal and ventral wall as well as the shape of the apical and basal cells); the presence and arrangement of chlamydospores. Finally, we considered the presence of pseudochlamydospores, of swollen hyphal cells and of coiled sterile hyphae as additional diagnostic features for some Fusarium species. We have not included the pigmentation on Potato Dextrose Agar (PDA) because this is a secondary character, that may only provide a first indication, and it has not been considered for diagnosis because it depends on the strain, type of media (commercial or homemade PDA) and cultural conditions. Before using the database, a special section has been created in which the user will be informed about which characteristics they have to consider and which to select to achieve a correct identification. We have provided a microscopic picture of each taxonomic character to help the user with the choice of the observed taxonomic character. The images used for the shape definition are reported in Fig. S2. Detailed data sheets have been created for 64 species, out of the 82 included in FusaHelp, in which additional information (important characteristics and differences with morphologically similar species) are provided. The list of Fusarium species included in the database can be found in Table S1.

For a correct and profitable use of FusaHelp, several rules must be followed. The morphological identification of Fusarium species must be based on pure single spore cultures (Fig. S3), with cultures grown on CLA (Carnation Leaf Agar) (Fisher et al. 1982) and/or SNA (Spezieller Nährstoffarmer Agar) (Nirenberg 1976). The morphological characteristics of the structures grown on PDA should not be considered as diagnostic characters. To verify the mode of formation of microconidia (false heads or chains), it is recommended to observe the culture directly under the 10× or 20× magnification, or to utilize a clear adhesive tape (Burgess et al. 1988; Harris 2000). Ideally the shape of macroconidia should be determined using macroconidia from sporodochia, when available. Therefore at least two slides are required: one from sporodochia and one from mycelium. Some authors describe the presence of “mesoconidia” as the macroconidia borne on mycelium with 3 or more septa. In this database we did not consider mesoconidia as a diagnostic criterion and we treat them as macroconidia.

The database FusaHelp is freely available at the following link: https://www.fusahelp.com. When entering the database, the first page that appears is divided into two screens. On the left frame are listed the seven morphological features available (microconidia; monophialides; polyphialides; chains; macroconidia; chlamydospores; other diagnostic features), while the right frame includes the list of 82 species in the database (Fig. 1). Each time a feature is chosen, the number of the possible species on the right will consequently be reduced. Identification starts by choosing one of the observed features. For instance, for microconidia, selecting if they are present or absent; if present, then it will appear “microconidial shape” and typing on it, the 13 different possible forms will appear. Choosing one or more of them, the search will be refined. As an example, if the “presence of microconidia” has been selected”, in the right part of the screen, the number of species will be reduced to 68 (Fig. 2).

Fig. 1
figure 1

Home page of FusaHelp

Full size image
Fig. 2
figure 2

Screen that appears after selecting the presence and shape of microconidia

Full size image

Please note that if you are not sure of the absence or presence of a character, it is better not to click on the box and leaving it blank.

Once the maximum number of the observed features has been correctly selected, the indication “identification complete” will appear and in the right part of the screen on the “Entities Remaining:1”, the corresponding species is identified (Fig. 3). For some species i.e., F. graminearum and F. pseudograminearum, morphologically indistinguishable, the “Entities Remaining” will be with these two species.

Fig. 3
figure 3

Example of the completed identification of F. fujikuroi

Full size image

If the advice “There are no entities remaining” appears, it means that the characters selected for identification are not sufficient for the identification of the species present in the database and some of them must be unselected or changed (Fig. 4). Please note that when selecting polyphialides, the presence of monophialides is also implied and you may not click the box with its presence. Furthermore, if you select monophialides “absent”, and “presence” in the box of polyphialides, the database does not lead to any identification (Fig. 5).

Fig. 4
figure 4

Example of a failed identification

Full size image
Fig. 5
figure 5

Example when the absence of monophialides and presence of polyphialides is selected

Full size image

For 64 species, a detailed description with additional photos is provided by clicking (icon with page image) the linked species (Fig. 6).

Fig. 6
figure 6

Example of a detailed description of one of the species present in the database

Full size image

Once a search is completed, or whenever needed, a new search is possible by simply clicking the icon on the top left-hand corner with two curved arrows arranged in a circle.

The classification and identification of microorganisms is constantly evolving and continuously updated according to the availability of even more advanced technologies. Currently, morphological species recognition (MSR), biological species recognition (BSR), and phylogenetic species recognition (PSR), are among the most widely adopted species concepts (Taylor et al. 2000). For many ascomycetes, including Fusarium, the use of one or more different approaches often leads to a long-lasting debate among research groups (Seifert et al. 2011; O’Donnell et al. 2020; Crous et al. 2021), but this is beyond the scope of the present work. Morphological identification of fungi is a key factor for the initial, immediate, and indicative diagnosis of many diseases in the agricultural, medical, veterinary sciences, as well as in food and feed industry. Furthermore, accurate morphological description of fungal structure opens the way for innovative, high throughput fungal identification techniques, like Digital Image Analysis (Papagianni 2014) and, more recently, deep learning and Artificial Intelligence approaches (Zieliński et al. 2020). FusaHelp is intended to provide a help for all those people who work with this important genus and need a quick clue on the identification, even incomplete, of the Fusarium species they are working with. It is also suited for student classes and for laboratories that do not have access to technologies based on fungal DNA determination. The versatility of use of FusaHelp with any portable device (tablet, mobile phone) could represent one of the strengths of FusaHelp that can be used for training courses. The programme is easily upgradable and will be implemented by the authors whenever any additional morphologically distinguishable Fusarium species will be requested and/or newly recognised important Fusarium species is formally reported.