Advancing automated identification of airborne fungal spores: guidelines for cultivation and reference dataset creation

Vol.: (0123456789) Aerobiologia (2025) 41:505–525 https://doi.org/10.1007/s10453-025-09864-y ORIGINAL PAPER Advancing automated identification ofairborne fungal spores: guidelines forcultivation andreference dataset creation NicolasBruffaerts · EliasGraf · PredragMatavulj · AsthaTiwari · IoannaPyrri · YanickZeder· SophieErb · MariaPlaza · SilasDietler· TommasoBendinelli· ElizabetD’hooge · BrankoSikoparija Received: 16 April 2025 / Accepted: 23 April 2025 / Published online: 2 June 2025 © The Author(s) 2025 due to diverse particle characteristics and limited training data availability, especially for fungal spores. This study aims to address this gap by outlining best practices for collecting reference material and creating tailored datasets for training algorithms. Using 17 fungal species from the Belgian fungi collection BCCM/IHEM, including five Alternaria species, key aspects such as in vitro cultivation, dry spore harvest, and aerosolization were addressed. Simple classification models were developed, achieving varying Abstract Airborne bioparticles, including fungal spores, are of major concern for human and plant health, necessitating precise monitoring systems. While a European norm exists for manual volumetric monitoring, there’s a growing interest in automated real-time methods. However, these methods rely heavily on machine learning, facing challenges Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s1045302509864-y. N.Bruffaerts(*)· A.Tiwari· E.D’hooge Mycology andAerobiology, Sciensano, Brussels, Belgium e-mail: nicolas.bruffaer[email protected] E. D’hooge e-mail: elizabet.dhoog[email protected] E.Graf· Y.Zeder Swisens AG, Emmen, Switzerland e-mail: [email protected] P.Matavulj Institute forData Science, University ofApplied Sciences North Western Switzerland, Windish, Switzerland e-mail: mata[email protected] I.Pyrri Biology Department, National andKapodistrian University ofAthens, Athens, Greece e-mail: ipyrr[email protected]r S.Erb Federal Office ofMeteorology andClimatology MeteoSwiss, Payerne, Switzerland e-mail: [email protected]h; [email protected] S.Erb Environmental Remote Sensing Laboratory, EPFL, Lausanne, Switzerland M.Plaza Faculty ofMedicine, University ofAugsburg, Augsburg, Germany e-mail: [email protected] S.Dietler· T.Bendinelli CSEM Alpnach, Centre Suisse d’Electronique et de Microtechnique, Alpnach, Switzerland e-mail: [email protected] E.D’hooge BCCM/IHEM, Sciensano, Brussels, Belgium B.Sikoparija BioSense Institute - Research Institute forInformation Technologies inBiosystems, University ofNovi Sad, NoviSad, Serbia e-mail: sikopari[email protected] 506 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) accuracies on different monitors. The Plair RapidE+ demonstrated accuracies ranging from 83.4% to 95.1% (macro average F1-score 0.61), with better recognition for Cladosporium spp. and Curvularia caricae-papayae. The SwisensPoleno Jupiter, initially achieving a macro average F1-score of 0.77 with holographic images of eight genera, improved to 0.83 when combined with fluorescence data. Accuracies ranged from 55 to 95%, with notable performance for Alternaria spp. and Curvularia caricae-papayae. Species differentiation was also shown to be possible for Cladosporium, but was more difficult for some Alternaria species, while the macro average F1-score remained good (0.72). Overall, this protocol paves the way for more efficient, standard, and accurate automatic identification of airborne fungal spores. Keywords Airflow cytometry· Automatic identification· Culture collection· Fungal spores· Machine learning 1 Introduction Bioaerosols impact both human and plant health in various ways, making their monitoring of great interest. Although a norm has been established (EN, 16868, 2019), based on the manual volumetric Hirst method, there is a growing demand from end-users, such as the medical community, allergic individuals, atmospheric modellers, and plant disease managers, for automated monitoring (Tummon et al., 2021, 2024). The diverse characteristics of bioaerosols in terms of size, origin and chemical composition (Després etal., 2012; Fröhlich-Nowoisky etal., 2016) make their monitoring challenging, particularly with respect to automatic real-time identification. Current approaches rely on physical or chemical properties, with some systems even employing a combination of both (Huffman etal., 2020). Despite these advances, challenges remain in differentiating fungal species due to overlapping fluorescence signatures. Studies such as those by O’Connor et al. (2011) and Saari etal. (2015) have highlighted the influence of species type, cultivation conditions, and exposure history on the fluorescence spectra of fungal spores. Moreover, foundational work on the biophysical basis of bioaerosol autofluorescence (Pöhlker etal., 2011; Hill etal., 2013) provides key insight into the molecular contributors to emission signatures detected by these systems. A recent evaluation comparing several operational bioaerosol monitors against the standard method has shown that airflow cytometry relying on morphological and chemical characterization performs comparably to digital microscopy (MayaManzano et al., 2023). A series of instruments relying on laser-induced fluorescence (LIF) and airflow cytometry-based technologies have also been employed extensively to characterize fungal spore composition and dynamics. The UV-APS, as evaluated by Kanaani et al. (2007), has shown sensitivity to the age and air exposure history of fungal spores, influencing fluorescence signal intensity and thus classification accuracy. WIBS systems have undergone substantial refinements for enhanced spectral resolution and particle characterization. For example, Markey et al. (2024) and Crawford etal. (2023) demonstrated the capabilities of WIBS-4 + in distinguishing fungal spores from pollen and other ambient particles under varying meteorological conditions. The BAA500, meanwhile, has recently been used with AI-driven classification approaches to detect Alternaria spores in real-time (González-Alonso etal., 2023). The quantity and complexity of data gathered automatically underscores the need for advanced data analytic methods to develop effective identification algorithms. Supervised machine learning methods are commonly used in automated systems for bioaerosol classification (Buters et al., 2022; Tummon et al., 2022), mostly relying on reference datasets upon which algorithms can be trained. As with training humans to identify particles by microscopy using reference slides of known material, these algorithms require datasets to be trained with appropriate labels, either generated from reference material or environmental samples. Obtaining such datasets is generally easier for pollen than for fungal spores. For pollen, reference material collection is relatively straightforward because the source plants can often be visually identified in the field with confidence, particularly for well-known anemophilous species. These species tend to produce large amounts of pollen, which facilitates sampling. A common technique involves harvesting mature inflorescences and placing them in paper envelopes or drying chambers, where the pollen is allowed to dehisce 507 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) naturally from the anthers under ambient or controlled conditions (Hoekstra, 1995). In contrast, reference fungal spore collection requires growing colonies in controlled conditions, with extraction methods often tailored to specific species (Drew Smith etal., 1961). So far, automatic identification of fungal spores relies on reference data obtained from exceptional events during operational measurements, i.e. when the fungal spore of interest is largely dominant in the air. This approach was applied for Alternaria spp. and total airborne spores using different measurement methods such as light microscopy imaging from an impactor (González-Alonso et al., 2023), airflow cytometry with holograms (Erb etal., 2023), and the combination of scattering data and laser-induced fluorescence (Simović et al., 2023). However, applications of this approach are limited by the scarcity of suitable episodes and by the difficulty to efficiently clean operational data. Therefore, the availability of large and clean reference datasets, encompassing the predominant spectrum of airborne fungal spores, is expected to facilitate the automated identification of a broad range of fungal diversity in ambient air, at least as allowed by the standard method (Anees-Hill etal., 2022). The aim of this study is to outline and describe the best practices for cultivating and collecting reference material, and creating datasets, in order to train algorithms for the classification of airborne fungal spores based on airflow cytometer measurements. Key aspects addressed here include: access to fungal strains whose spores are representative of the outdoor air mycobiome; controlled growing and sporulation of single species cultures; harvesting of clean and individual fungal spores from sporulating colonies; aerosolization of dry spores to challenge automatic monitors; and creation and cleaning of training datasets. In addition, basic classification models have been developed to assess the identification potential. 2 Fungal spore production 2.1 Fungal reference material The standard monitoring method EN 16868, which relies on direct microscopic observation of aerobiological samples, offers limited resolution for assessing the species diversity of airborne fungal spores. This limitation arises in part because aerobiological samples contain a wide range of morphologically similar spores, making fine taxonomic identification difficult without specialized mycological expertise. Moreover, many fungal taxa cannot be reliably distinguished based on spore morphology alone, as accurate identification often requires examination of mycelial structures that are absent in airborne samples. As for indoor mycological monitoring methods, more sensitive culture-based analysis from these aerobiological samples could offer a better representativity of its species diversity and therefore an opportunity to collect reference material. However, such an approach would require substantial effort in cleaning and isolating colonies to ensure only single-species isolates that can be cultivated for spore production. Additionally, given the rapidly changing fungal taxonomy and the consequent necessity to perform DNA identification, it is advisable to use reference strains from certified microorganism culture collections, which ensure reliability and reproducibility. In this study, 17 certified fungal strains (Table1), covering 11 different genera, were obtained from BCCM/IHEM (Sciensano, Belgium), a collection of yeasts and moulds of medical and veterinary interest that gathers more than 16,000 strains, representing more than 350 genera and 1,200 species of Ascomycetes, Basidiomycetes and Mucoromycetes. Therefore, the selection of fungal strains was primarily driven by the BCCM/IHEM public catalogue,1 which is limited to healthrelevant species and their sister taxa and species in the environment. In the future, the evaluation of strains of particular interest to agriculture or forestry (Entomophthoraceae, Pucciniomycetes, etc.) could be explored in collaboration with collections dedicated specifically to these fields. The overall abundance of fungal species in the outdoor air was taken into account for the strain selection as fungal spores and their mycelium fragments constitute up to 11% of atmospheric bioaerosols (Tordoni etal., 2021). Cladosporium and Alternaria are commonly known as the most abundant fungal genera in the ambient air as reported by morphological analyses and eDNA metabarcoding (Banchi 1 https:// bccm. belspo. be/ catal ogues/ ihemcatal oguesearch. 508 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) Table 1 List of the 17 fungal strains selected from the BCCM/IHEM fungal culture collection Label in this study Current species name Common synonym name(s) Strain name Alternaria alternata Alternaria alternata (Fries ex Fries) von Keissler Alternaria citri Ellis & Pierce emend. Bliss & Fawcett Alternaria destruens E.G. Simmons Alternaria soliaegyptiaca E.G. Simmons Alternaria tenuis Nees Alternaria tenuissima (Kunzeex Fries) Wiltshire IHEM 3327 Alternaria arborescens Alternaria arborescens E.G. Simmons – IHEM 18586 Alternaria botrytis Alternaria botrytis (Preuss) Woudenberg & Crous Ulocladium botrytis Preuss Stemphylium botryosum var. ulocladium Sacc Stemphylium botryosum var. botrytis (Preuss) Lindau IHEM 2964 Alternaria chartarum Alternaria chartarum Preuss Ulocladium chartarum (Preuss) E.G. Simmons Alternaria chartarum f. stemphylioides (Bliss) P. Joly Alternaria stemphylioides Bliss Sporidesmium polymorphum var. chartarum (Preuss) Cooke IHEM 20041 Alternaria terricola Alternaria terricola Woudenberg & Crous Ulocladium tuberculatum E. Simmons IHEM 4136 Botrytis cinerea Botrytis cinerea Persoon ex Fries Botryotinia fuckeliana (de Bary) Whetzel Sclerotinia fuckeliana (de Bary) Fuckel IHEM 27814 Chaetomium globosum Chaetomium globosum Kunze ex Fries Chaetomium rectum Sergeeva Chaetomium chartarum Ehrenb Chaetomium chlorinum (Sacc.) Grove Chaetomium kunzeanum Zopf Chaetomium olivaceum Cooke & Ellis IHEM 6003 Cladosporium cladosporioides Cladosporium cladosporioides (Fresen.) de Vries Cladosporium pisicola W.C. Snyder Penicillium cladosporioides Fresen Hormodendrum cladosporioides (Fresen.) Sacc IHEM 26980 Cladosporium herbarum Cladosporium allicinum (Fr.) Bensch, U. Braun & Crous Cladosporium bruhnei Linder Cladosporium herbarum f. hordei (Bruhne) Ferraris Cladosporium hordei (Bruhne) Pidopl IHEM 3260 Cladosporium sphaerospermum Cladosporium sphaerospermum Penzig – IHEM 25311 Curvularia caricae-papayae Curvularia caricae-papayae H.P. Srivast & Bilgrami – IHEM 21710 Epicoccum nigrum Epicoccum nigrum Link Epicoccum purpurascens Ehrenberg & Schlechtendahl IHEM 3433 Exserohilum rostratum Exserohilum rostratum (Drechsler) Leonard & Suggs emend. Leonard Drechslera rostrata (Drechsler) Richardson & Fraser Helminthosporium halodes Drechsler Helminthosporium leptochloae Nisikado & Miyake Helminthosporium rostratum Drechsler Bipolaris rostrata (Drechsler) Shoemaker IHEM 3524 Fusarium culmorum Fusarium culmorum (W.G. Smith) Saccardo Fusisporium culmorum Wm. G. Sm IHEM 3323 509 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) etal., 2020; Tordoni etal., 2021), as well as Epicoccum and Stemphylium are also relatively abundant. It should be noted that routine aerobiological analysis following the standard method EN 16868 relies on spore identification based on their morphology, which usually limits identification up to the level of genus, often denoted as fungal spore type (Galan etal., 2021). The four previously mentioned fungal spore types are routinely measured by many European aerobiological monitoring networks (AneesHill et al., 2022), together with Botrytis, Chaetomium, Curvularia, Drechslera (Helminthosporium), Pithomyces, Torula and the Aspergillaceae family (Penicillium/Aspergillus spp.). All these taxa are relevant from the health point of view and were included in this study when possible. A strain of Exserohilum rostratum was selected to represent Drechslera, a morphologically classified spore type common in aerobiological analysis, as well as Pseudopithomyces palmicola for Pithomyces. The Torula strain tested in this study failed to achieve a satisfactory level of sporulation under the experimental conditions, thereby yielding insufficient material for analysis. In contrast, cultures of Penicillium chrysogenum and Aspergillus fumigatus strains exhibited prolific growth, facilitating the collection of high spore amounts with minimal hyphal contamination. This was evidenced by the material readily falling onto the glass wall of the slant tube. A gentle inversion onto the bench is enough to dislodge it from the culture surface, making a more complex harvesting system, as described in this study, unnecessary. Furthermore, the majority of spores in this family are smaller than 5µm and sphericalshaped. Hence, they present less interest when developing algorithms relying on holography data comparatively to other spore taxa. The Aspergillaceae taxon was accordingly excluded from this study despite its availability in the BCCM/IHEM catalogue. The Alternaria taxonomic group is broad and morphologically diverse, and it formally encompasses several taxa that are still commonly identified as separate entities in aerobiological monitoring, such as Pleospora, Ulocladium, and Stemphylium. Morphological species recognition in the Alternaria group is based on the branching patterns of conidial chains and on the morphological characteristics of conidia (Simmons etal., 2007). For this reason, in this study, we selected five strains belonging to different species and with contrasted conidia shapes, from spores with a beak (e.g., A. tenuissima and A. arborescens) to Ulocladium-like spores with non-apparent beak (A. botrytis and A. terricola). Recent molecular systematics have prompted the reassessment of species-group organization within the genus. As a result, certain species names have been subject to change, such as A. tenuissima being synonymized with A. alternata. Despite differences in conidia colour and beak size, Table 1 (continued) Label in this study Current species name Common synonym name(s) Strain name Fusarium pseudocircinatum Fusarium pseudocircinatum O’Donnell & Nirenberg –IHEM 18653 Pithomyces chartarum Pseudopithomyces chartarum (Berk. & M.A. Curtis) Jun F. Li, Ariyaw. & K.D. Hyde Pithomyces chartarum (Berk. & M.A. Curtis) M.B. Ellis Leptosphaerulina chartarum Cec. Roux, Trans. Br. mycol Piricauda chartarum (Berk. & M.A. Curtis) R.T. Moore Sporidesmium bakeri Syd. & P. Syd Sporidesmium chartarum Berk. & M.A. Curtis IHEM 15222 Stemphylium vesicarium Stemphylium vesicarium (Wallr.) E.G. Simmons Stemphylium herbarum E.G. Simmons Stemphylium sedicola E.G. Simmons Helminthosporium vesicarium Wallr Pleospora denotata (Cooke & Ellis) Sacc Pleospora pisi (Sowerby) Fuckel IHEM 3105 Each strain is listed with its current species name alongside a list of common synonyms, and the designation utilized in this study 510 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) these alterations have been supported (Woudenberg etal., 2015). Regarding the genus Cladosporium, which is one of the largest genera of dematiaceous hyphomycetes, strains representative of three major species complexes were selected, namely C. cladosporioides (belonging to the C. cladosporioides species complex), C. sphaerospermum (C. sphaerospermum species complex) and C. allicinum (C. herbarum species complex). Although these species may be identified thanks to their surface ornamentation (Bensch etal., 2015), holography images of current real-time airflow cytometers are not expected to allow efficient classification within this genus, as for the standard method by light microscopy. Finally, the genus Fusarium comprises a large number of human and plant pathogenic species, for example the F. fujikuroi species complex. A strain of F. pseudocircinatum, belonging to this complex, was selected for this study. However, the 25–50 µm long, fusiform macroconidia are abundantly mixed with ellipsoidal to cylindrical 5–10 µm microconidia (Ajmal etal., 2022), which may impede the creation of quality training data with uniform spore types. Therefore, a strain of F. culmorum was added in the selection. This last species is known to produce macroconidia without microconidia. Comparison of both strains would highlight the potential interference of mixed spore types within a taxon-specific training dataset. 2.2 Culture conditions for sporulating colonies All freeze-dried stock products were inoculated in slant tubes, following the BCCM/IHEM instructions of use (BCCM/IHEM, 2015), in order to reactivate the growth capacity of the strain. Then, the use of Petri dishes is recommended for subcultures to increase the sporulating surface and ease the access for harvesting. Here, 9 cm diameter Petri dishes (Greiner Bio-One, Belgium) were used. The size of the dish will determine the time required for complete confluence of the culture, which may affect growing dynamics and the sporulating phenotype. If the culture is left under confluence for too long, hyaline mycelium progressively grows on an upper layer (the so-called secondary mycelium). Although the literature contains optimal culture media for specific fungi groups (Atlas etal., 1993), we limited the set of broad spectrum media to Potato Dextrose Agar (PDA),2 diluted Sabouraud (S10)3 or V8 agar (V8) (5% V8 juice (Campbell Soup Co.), 1g MgSO4·7H2O, 1g KH2PO4, 20 g Pastagar, 1mL oligoelements for 1 L at pH 5.5), for all strains as detailed in Table 2. Sporulation yields were also tested on cultures grown on Malt Extract Agar4 and were similar (not shown). In general, the media mentioned here above should be suitable for most species, with the exception of V8 which is more commonly used for Alternaria species. V8 is reported to support abundant sporulation and reproducible sporulation patterns that are comparable to those found in nature (Simmons etal., 2007). Inoculation was performed either by cutting the initial culture into 3–5 agar pieces of 2 × 2 mm, loaded with fungal material, and aseptically placing it on the surface of the Petri dish, or by rubbing one of these pieces by striation over the entire surface of the Petri dish in order to ensure an even growth and sporulation over the whole surface (Li et al., 2020). All dishes were incubated without humidity control at 25 °C in the dark for about two weeks. It is recommended to avoid sealing Petri dishes, as the accumulation of water condensate may distort colonies and sporulation patterns. An incubation period of 5–10 days typically suffices to achieve a satisfactory sporulation yield, although sporulation may occur a few days earlier if inoculation is conducted via striation. The yield of sporulation can be assessed by examining the surface coverage of colourless mycelium. Subsequent to the incubation period, cultures of the 17 selected strains underwent both macroscopic (Supplementary Fig. 1) and stereoscopic inspections (Supplementary Fig.2). Notably, all strains exhibited satisfactory levels of sporulation (Supplementary Fig.1), obviating the need for further optimization of culture conditions like exposure to near-UV light or the use of more specific media (Su etal., 2012). Specifically, Alternaria spp. grew rapidly on artificial nutrient media 2 https:// bccm. belspo. be/ catal ogues/ ihemmediadetai ls? Cultu reMed iaID= 905. 3 https:// bccm. belspo. be/ catal ogues/ ihemmediadetai ls? Cultu reMed iaID= 908. 4 https:// bccm. belspo. be/ catal ogues/ ihemmediadetai ls? Cultu reMed iaID= 902. 511 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) covering the Petri dish surface with densely felted, dark olive brown to black colonies. Mycelium was partly submerged within the agar layer and partly aerial, in particular for the tested Alternaria alternata strain. Mature conidia germinated producing secondary mycelium in A. alternata, A. botrytis and A. chartarum. The three Cladosporium species produced slow growing, velvety, dark olive colonies with immersed stromatal hyphae. Botrytis cinerea covered the entire Petri dish surface with loosely floccose, mouse grey colonies. Curvularia caricae-papayae, Epiccocum nigrum, Pithomyces chartarum exhibited medium growth rate with colonies thinly floccose, black coloured. Chaetomium globosum spread rapidly, with sparse white to buff aerial hyphae when young, then without aerial hyphae, becoming greenish olivaceous owing to the aggregation of ascomata. Exserohilum rostratum was also relatively fast growing, covering the plate with radial fibery mycelium growth and dark brown colonies. Stemphylium Table 2 Relative score for ease of harvesting by suction with the cyclone collector (from 0 for the lowest yield to 3 for the highest yield), and microscopically significant presence of hyphae and/ or other particles in the harvested sample (Y: presence, N: absence). PDA: Potato Dextrose Agar; S10: Diluted Sabouraud Species Culture medium used Ease of harvesting score Presence of hyphae/other particles Alternaria alternata PDA 0 Y Alternaria arborescens PDA, S10, V8 2 N Alternaria botrytis PDA 2 N Alternaria chartarum PDA 1 N Alternaria terricola PDA 1 N Botrytis cinerea S10 0 N Chaetomium globosum S10 1 Y Cladosporium cladosporioides S10 2 N Cladosporium herbarum S10 2 Y Cladosporium sphaerospermum S10 2 N Curvularia caricae-papayae S10 1 N Epicoccum nigrum S10 1 N Exserohilum rostratum S10 1 N Fusarium culmorum S10 3 N Fusarium pseudocircinatum S10 1 N Pithomyces chartarum S10 3 N Stemphylium vesicarium S10 1 N Fig. 1 A Instructions for constructing a plastic cyclone collector vial using a micro-tube and two micropipette tips, B Picture of an assembled cyclone collector vial, C Picture of serially connected three cyclone collector vials for increasing the harvesting efficiency of large particles, D Scheme of the harvesting system to aspirate dry spores from fungi culture on Petri dish 512 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) vesicarium colonies developed relatively more slowly, having mostly white mycelium and sporulating sparsely. Finally, Fusarium spp. covered the entire Petri dish surface rapidly with colonies velvety, radiate, with relatively sparse aerial mycelium, white or reddish. In general, all tested species, except Chaetomium globosum, produced conidia on conidiophores. Alternaria alternata, A. arborescens, A. botrytis, A. chartarum, A. terricola and Cladosporium cladosporioides, C. herbarum, C. sphaerospermum produced conidia in branched chains while Curvularia caricae-papayae, Stemphylium vesicarium, Epicoccum nigrum, Pithomyces chartarum and Exserohilum rostratum produced spores singly at the top of simple, short conidiophores. Botrytis cinerea and Fusarium culmorum and F. pseudocircinatum had more complex, branched conidiophores whereas Chaetomium globosum produced ascospores inside perithecia that have hairy hyphae around the ostiole. The diverse modes of sporulation and mycelium quantity can be observed in Supplementary Fig. 2. Although sporulation yields may vary across different Petri dishes, culturing a larger number of replicates can mitigate this effect. Phenotypic variations can arise due to differences in micro-environmental conditions during culture and the inherent stability of the strain. 3 Fungal spore harvesting andaerosolization 3.1 Isolation from sporulating colonies In fungal taxonomy studies, a spectrum of methodologies are employed to either collect material for fungal inoculation or describe the morphology of spores (Senanayake, 2020). To enhance yield, spores are typically dislodged from colonies either through washing or by brushing and vacuuming onto a filter or into a cyclone particle collector (Drew Smith etal., 1961; Pogner etal., 2024). For airflow-based extraction, the location of fungal colonies relative to the airstream significantly impacts the number and trajectory of dispersed spores, emphasizing the importance of spatial configuration during lab experiments (Li et al., 2020). Methods that rely on washing spores from sporulating colonies with liquid (such as flooding and subsequent scraping or vortexing to detach spores from conidiophores) Fig. 2 Experimental laboratory setup for the aerosolization of fungal spores and their detection by the SwisensPoleno Jupiter. Homogenous aerosolization within the chamber of the SwisensAtomizer can be performed either by using an open square cuvette or the cyclone collector vial, filled with dry spore material. Arrows indicate the direction of air flows to and out of the container 513 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) are not ideal when analysing spores for specific molecules, as these could be released and dissolved in water. Indeed, biogenic secondary organic aerosols exhibiting their own autofluorescence properties may be subject to aqueous extraction and hence interfere with the fluorescence-based measurements of primary biological aerosol particles like fungal spores (Zhang etal., 2021). Furthermore, this method often leads to significant hyphal contamination (Jacques et al., 2021). Therefore, for creating reference material for airflow cytometry, dry extraction appears most suitable, as it is anticipated to maintain spores unaltered, akin to their natural emission and atmospheric dispersion processes, while also facilitating easier aerosolization. For certain fungi, such as those belonging to the Aspergillaceae family, spores detach easily and can be dislodged from the colony in substantial quantities. However, for many others, particularly when cultured in vitro, considerable force is necessary to detach and extract spores from the intricate network of hyphae. Brushing is a commonly employed method for extracting dry spores (Jacques etal., 2021), but it often inadvertently collects hyphae along with the spores. Hence, suctioning spores directly from the colony appears to be the optimal approach. We opted to vacuum-extract spores from the colony using a cyclone collector (Fig.1A) rather than a filter to prevent them from embedding in the filter matrix. To simplify fungal spore collection and keep it cost-effective, we aimed to use disposable cyclone collector vials. With this approach, separate cyclone collectors could be used for each sample. We constructed a collector using a single 1.5 mL Eppendorf vial and two 2–200 µL pipette tips (Fig.1B). One tip is positioned tangentially through the wall of the vial, while the other is inserted through the lid of the vial; both securely affixed using hot plastic glue. Alternatively, this type of collector vial could be fabricated through 3D printing techniques. The experiments confirmed that cyclone sampling is suitable for efficient extraction of fungal spores from sporulating colonies. Spore extraction varied in ease across different samples (Table 2; Supplementary Video 1), sometimes requiring surface scratching with the cyclone collector tip. It is noteworthy to emphasize that extracting spores from moist colonies (due to condensation on the Petri dish lid or exudate on hyphae) was not possible. Therefore, we ensured that excess moisture was evaporated by leaving the Petri dish lid opened inside the biosafety cabinet at room temperature until it was sufficiently dry, as confirmed under stereomicroscope. We empirically determined that a flow rate into the cyclone collector exceeding 5 L/min, corresponding to a pump inlet pressure of approximately 100 mmHg, provided sufficient force to detach spores from colonies while generally minimizing the co-harvesting of hyphae. To establish this, we began at the lowest flow rate permitted by our laboratory pump and gradually increased the pressure. This stepwise approach allowed us to identify the minimum critical pressure at which visible and consistent spore collection occurred across all selected taxa. Only once this threshold was reached did we proceed to score both spore yield and hyphal contamination. While this approach was not designed to formally quantify harvesting efficiency across the entire pressure gradient, it enabled us to identify a robust operating point suitable for most species tested. In practice, it was difficult to avoid contact between the cyclone collector tip and aerial hyphae for most species. In some instances, however, such contact facilitated spore release; nevertheless, it occasionally led to an increased collection of hyphae. This phenomenon was particularly notable for Cladosporium herbarum and Chaetomium globosum, where the delicate structures producing spores tended to fragment, resulting in the harvest of small pieces along with some hyphae. Additionally, increasing the airflow speed by narrowing the inlet could enhance spore detachment. Nevertheless, it is important that the inner diameter of the tube leading from the inlet to the body of the collector does not increase since it would result in deposition of particles on the inside walls, especially concerning small spores. It should be noted that the used cyclone collector vials have limited efficiency for capturing small particles, leading to a notable number of spores escaping and being drawn in by the pump. Efficiency could be substantially increased by connecting multiple cyclones in series (Fig. 1C). Despite this enhancement, the escape of some particles remains inevitable. Thus, it is essential to cover the pump outlet with a filter or place it within a biosafety cabinet to prevent environmental contamination with spores. It is also worth noting that sampling hyaline spores poses a challenge due to their transparent 520 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) (about 462 nm) detectors, which were susceptible to interference from the light of the scattering laser. Additionally, we aligned the fluorescence lifetime measurements so that the signal starts at the first maximum. To enhance data quality, the fluorescence spectrum and all three scattering images were smoothed using a Savitzky–Golay filter. Finally, we transformed both the fluorescence spectrum and the fluorescence lifetime modalities into image-like formats and then normalized them into a 0–1 range to facilitate uniform handling by the convolutional neural networks. For classification of selected fungal spores, we implemented a ResNet network, which showed superior performance in a previous study with the Plair Rapid-E device (Matavulj et al., 2023). The architecture employed in ResNet features that are termed as"shortcut connections". When an input, represented as x, passes through the convolutional layers of the network, it undergoes a transformation denoted by the function F. This function F(x) serves to highlight key features such as edges and shapes, which are critical to the accurate representation of the original image. The shortcut connection introduces x to the F(x), so the output x + F(x) represents a slightly altered or refined version of the original input x. We utilized a variation of an 18-layer ResNet model, where we implemented certain layers of ResNet. Specifically, we implemented a 4-block-layer ResNet for fluorescence lifetime, a 3-block-layer for scattering images, a 5-block-layer net for fluorescence spectrum, and a 2-block-layer ResNet for infrared. The blocklayers contained 3 convolutional layers (except for the first block-layer, which comprised 4 convolutional layers). To handle fluorescence lifetime and spectrum data, the first convolutional layer was customized to accept a monochrome image, configured with a kernel size of 5 × 5, a padding of 2 × 2, and without any stride to maintain the original spatial dimensions. To address our classification task with its defined number of classes, we adjusted the final fully-connected layer accordingly. The classification model was trained on 80% of the reference dataset, 10% was used for model validation during training to prevent overfitting, and 10% was used to test the classification performance. Fig. 6 Confusion matrix of the recognition performance of the convolutional neural networks trained on fluorescence data from seven fungal spore genera recorded by the Plair RapidE+. Alternaria spp. include the mixed species A. alternata, A. arborescens, A. botrytis, A. chartarum and A. terricola. Cladosporium spp. include the mixed species Cladosporium cladosporioides, Cladosporium herbarum and Cladosporium sphaerospermum. Percentage of accurately predicted and misclassified test particles is presented as a heat map 521 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) The classification performance of selected fungal spores measured with the Plair Rapid-E+ device yielded a macro average F1-score exceeding 0.61 (Fig. 6 and Supplementary Table 3). Notably, the model demonstrated best accuracy in identifying Curvularia caricae-papayae (accuracy 95.1%) and Cladosporium spp. (accuracy 92.7%), followed by Botrytis cinerea (accuracy 89.9%) and Epicoccum nigrum (accuracy 88.9%). Both Botrytis cinerea and Epicoccum nigrum had the most intense fluorescence after excitation at 337 nm and Cladosporium spp. was distinctively smaller than other tested spores. The analysis confirmed the potential of Plair Rapid-E+ measurements in the “middle” mode to discriminate between some of the most prevalent airborne fungal spores. However, since it is known that laser-induced data are prone to noise resulting from laser and detector sensibility (Huffman etal., 2020; Könemann etal., 2019; Robinson etal., 2017), more work needs to be done on cleaning training datasets to further improve the recognition performance. 5 Discussion andconclusion The study elucidates effective methodologies for fungal spore harvesting and aerosolization, crucial steps in the creation of training datasets for automated monitoring systems. It emphasizes the importance of representative fungal strains, controlled sporulation, and harvesting of individual spores. Dry extraction of in vitro fungi cultures on agar medium using a cyclone collector proves to be an optimal approach for maintaining spore integrity, minimizing hyphal contamination, and allowing to produce a stock of reference spore material that can easily be stored and shared. Furthermore, several recommendations emerged from this protocol of spore production: 1. Careful strain selection The selection of specific fungal strains is paramount. We have shown that it may significantly affect morphology-based identification accuracy, in particular for large complexes like Alternaria and Cladosporium. The presence of mixed macroconidia and microconidia, as observed in Fusarium, is also a point that should be taken into account, as it may alter identification efficiency. 2. Anticipate phenotype variability Relatively variable results from a same fungal culture protocol is inevitable across different laboratories. Factors such as strain fitness, growth medium, and duration can influence outcomes. It is essential to allow sufficient growth time to ensure a diverse range of spore maturity and enhance the representativeness of the reference sample. However, for some strains, excessively long culture periods may lead to the growth of secondary mycelium, thus reducing the sporulation yield on the surface of Petri dishes. This can though be mitigated by an increased amount of dishes. 3. Dry spores from dry cultures While humidity control during the culture growth period may not be critical, it is recommended to dry Petri dishes just before the harvesting process in order to facilitate efficient spore collection. Cultures exhibiting high humidity levels, often indicated by water exudates within the mycelium, can impede the gentle harvest of spores, which is essential for minimizing the inadvertent sampling of hyphae. In particular, the spore dispersal strategy of some fungi (e.g., Acremonium spp., Stachybotrys chartarum, Fusarium spp., some myxomycetes, etc.) relies on the production of droplets or wet false heads at the apex of conidiophores (Magyar etal., 2016), which might be incompatible with the proposed protocol or would require some adaptation in regard to humidity level. Additionally, ambient humidity may influence the fluorescence intensity of biogenic secondary organic aerosols (Zhang etal., 2021), potentially posing a caveat to our proposed method for aerosolizing dry fungal spores. There is considerable potential for optimizing the method presented in this study. However, before assessing the classification performance in an operational setup against standard measurements such as EN 16868, several aspects pertaining to spore’s variability should be addressed by the scientific community in aerobiology. While the variability in spore morphology within a single culture is generally well-documented for reference fungal strains, there is limited understanding of how this cultured variability correlates with 522 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) the possible range of spore variability in natural environments. Single spores might not constitute the sole form suspended in the atmosphere. Moreover, most studies have only focussed on fresh bioaerosols for setting up datasets for machine learning. After long term transportation in ambient air, fungal spores might undergo changes in morphology and chemical properties due to ageing and interactions with atmospheric components such as pollutants, humidity and UV radiation. This represents a key research gap. Compounding this issue, spores cultivated in artificial nutrient media can exhibit diverse maturation states. For instance, immature spores of Alternaria may appear hyaline, ovoid, and one or two-celled, significantly different from their mature, multicellular forms. Such heterogeneity complicates the development of standardized and representative reference datasets for training reliable classification models. Moreover, the cleaning procedures employed will significantly influence the nature of the classification model and, consequently, its performance when compared to validation datasets like standard aerobiological measurements. For example, if standard measurements encompass both spores and hyphae under a specific spore type, cleaning protocols should refrain from discarding hyphae. Furthermore, for spores suspended in the atmosphere as aggregates (e.g., chains of Cladosporium conidia), retaining such formations in the training dataset becomes imperative. However, it is essential to note that when analysed under an upright microscope, a chain of three spores should be treated as a single particle rather than three separate ones, to ensure methodological comparability. Finally, exploring the chemical characteristics of spore cell walls from such reference material would be important, especially in cases where fluorescence measurements are utilized for identification. As previous observations showing that sporulation age significantly affects signal response (Kanaani etal., 2007; Saari et al., 2015), investigating chemical specificities given variable culture conditions would provide insights into the factors influencing algorithmic confusion and aid in the development of more accurate and reliable automated identification systems. In all, these considerations are crucial for the accurate interpretation and comparison of data obtained through different methodologies and instruments in the field of aerobiology. Moving forward, efforts should focus on expanding reference datasets to encompass a broader spectrum of fungal diversity found in the atmosphere. This will facilitate more accurate and comprehensive automated identification of fungal spores, ultimately enhancing our understanding of bioaerosol dynamics and their impact on human and plant health. Acknowledgements COST Action ADOPT through STSMs: E-COST-GRANT-CA18226-fa273a24, E-COST-GRANTCA18226-e04047ee, E-COST-GRANT-CA18226-d7 d22ed3, the Ministry Science, Technological Development and Innovations of the Republic of Serbia (Grant No. 200358), the Swiss National Science Foundation (Grant No. IZCOZ0_198117) and Horizon Europe project SYLVA (Grant No. 101086109). The research has received funding from the Chips Joint Undertaking and the Swiss Confederation under Grant Agreement No. 101095835. The project 23 NRM03 BioAirMet has received funding from the European Partnership on Metrology, cofinanced from the European Union’s Horizon Europe Research and Innovation Programme and by the Participating States. Authors contribution Conceptualization and supervision: NB, BS; Methodology, Resources and Investigation: NB, EG, AT, ED, BS; Software and Formal analysis: EG, PM, YZ, SD, TB; Writing original draft, editing and reviewing: all authors. Data availability The data that support the findings of this study are available from the corresponding author (NB) upon reasonable request. Declarations Conflict of interest Elias Graf and Yanick Zeder are employees of Swisens AG. All other authors have no conflict of interest to declare. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 523 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) References Ajmal, M., Hussain, A., Ali, A., Chen, H., & Lin, H. (2022). Strategies for controlling the sporulation in Fusarium spp. JoF, 9, 10. https:// doi. org/ 10. 3390/ jof90 10010 Anees-Hill, S., Douglas, P., Pashley, C. H., Hansell, A., & Marczylo, E. L. (2022). A systematic review of outdoor airborne fungal spore seasonality across Europe and the implications for health. Science of the Total Environment, 818, Article 151716. https:// doi. org/ 10. 1016/j. scito tenv. 2021. 151716 Atlas, R. M., & Parks, L. S. (1993). Handbook of microbiological media (pp. 1079–1079). CRC. Banchi, E., Ametrano, C. G., Tordoni, E., Stankovic, D., Ongaro, S., & ARPA working group, Tretiach, M., Pallavicini, A., Muggia, L.,. (2020). Environmental DNA assessment of airborne plant and fungal seasonal diversity. Science of the Total Environment, 738, Article 140249. https:// doi. org/ 10. 1016/j. scito tenv. 2020. 140249 BCCM/IHEM (2015). YouTube, uploaded by BCCM/IHEM, “BCCM/IHEM freeze-dried ampoules: instructions for use”, https:// www. youtu be. com/ watch?v= 9nKyM wOxTTA Bensch, K., Groenewald, J. Z., Braun, U., Dijksterhuis, J., De Jesús Yáñez-Morales, M., & Crous, P. W. (2015). Common but different: The expanding realm of Cladosporium. Studies in Mycology, 82, 23–74. https:// doi. org/ 10. 1016/j. simyco. 2015. 10. 001 Buters, J., Clot, B., Galán, C., Gehrig, R., Gilge, S., Hentges, F., O’Connor, D., Sikoparija, B., Skjoth, C., Tummon, F., Adams-Groom, B., Antunes, C. M., Bruffaerts, N., Çelenk, S., Crouzy, B., Guillaud, G., Hajkova, L., Seliger, A. K., Oliver, G., … Stjepanovic, B. (2022). Automatic detection of airborne pollen: An overview. Aerobiologia. https:// doi. org/ 10. 1007/ s1045302209750-x Crawford, I., Bower, K., Topping, D., Di Piazza, S., Massabò, D., Vernocchi, V., & Gallagher, M. (2023). Towards a UK airborne bioaerosol climatology: Real-time monitoring strategies for high time resolution bioaerosol classification and quantification. Atmosphere, 14(8), 1214. Després, V. R., Huffman, J. A., Burrows, S. M., Hoose, C., Safatov, A. S., Buryak, G., Fröhlich-Nowoisky, J., Elbert, W., Andreae, M. O., Pöschl, U., & Jaenicke, R. (2012). Primary biological aerosol particles in the atmosphere: A review. Tellus b: Chemical and Physical Meteorology, 64, 15598. https:// doi. org/ 10. 3402/ tellu sb. v64i0. 15598 Drew Smith, J., Crawley, W. E., & Lees, F. T. (1961). Collection and concentration of spores of Pithomyces chartarum from herbage and harvesting from ryecorn cultures. New Zealand Journal of Agricultural Research, 4, 725–733. https:// doi. org/ 10. 1080/ 00288 233. 1961. 10431 629 EN 16868, (2019). ‘Ambient air - Sampling and analysis of airborne pollen grains and fungal spores for networks related to allergy - Volumetric Hirst method’, Cen. Erb, S., Berne, A., Burgdorfer, N., Clot, B., Graber, M.-J., Lieberherr, G., Sallin, C., Tummon, F., & Crouzy, B. (2023). Automatic real-time monitoring of fungal spores: The case of Alternaria spp. Aerobiologia. https:// doi. org/ 10. 1007/ s1045302309780-z Fröhlich-Nowoisky, J., Kampf, C. J., Weber, B., Huffman, J. A., Pöhlker, C., Andreae, M. O., Lang-Yona, N., Burrows, S. M., Gunthe, S. S., Elbert, W., Su, H., Hoor, P., Thines, E., Hoffmann, T., Després, V. R., & Pöschl, U. (2016). Bioaerosols in the earth system: Climate, health, and ecosystem interactions. Atmospheric Research, 182, 346–376. https:// doi. org/ 10. 1016/j. atmos res. 2016. 07. 018 Galán, C., Smith, M., Damialis, A., Frenguelli, G., Gehrig, R., Grinn-Gofroń, A., Kasprzyk, I., Magyar, D., Oteros, J., Šaulienė, I., Thibaudon, M., Sikoparija, B., EAS QC Working Group. (2021). Airborne fungal spore monitoring: between analyst proficiency testing. Aerobiologia, 37, 351–361. https:// doi. org/ 10. 1007/ s1045302109698-4 González-Alonso, M., Boldeanu, M., Koritnik, T., Gonçalves, J., Belzner, L., Stemmler, T., Gebauer, R., Grewling, Ł, Tummon, F., Maya-Manzano, J. M., Ariño, A. H., Schmidt-Weber, C., & Buters, J. (2023). Alternaria spore exposure in Bavaria, Germany, measured using artificial intelligence algorithms in a network of BAA500 automatic pollen monitors. Science of the Total Environment, 861, Article 160180. https:// doi. org/ 10. 1016/j. scito tenv. 2022. 160180 Healy, D. A., O’Connor, D. J., Burke, A. M., & Sodeau, J. R. (2012). A laboratory assessment of the Waveband Integrated Bioaerosol Sensor (WIBS-4) using individual samples of pollen and fungal spore material. Atmospheric Environment, 60, 534–543. https:// doi. org/ 10. 1016/j. atmos env. 2012. 06. 052 Hill, S. C., Pan, Y. L., Williamson, C., Santarpia, J. L. & Hill, H. H., (2013). Fluorescence of bioaerosols: mathematical model including primary fluorescing and absorbing molecules in bacteria. Optics Express, 21(19), 22285–22313 Hoekstra, F. (1995). Collecting pollen for genetic resources conservation. In: L. Guarino, V. Ramanatha Rao, R. Reid, (eds.), Collecting Plant Genetic Diversity: Technical Guidelines, (pp. 527–550) Wallingford, UK. Huffman, J. A., Sinha, B., Garland, R., Snee-Pollmann, A., Gunthe, S., Artaxo, P., etal. (2012). Size distributions and temporal variations of biological aerosol particles in the Amazon rainforest characterized by microscopy and realtime UV-APS fluorescence techniques during AMAZE08. Atmospheric Chemistry and Physics, 12, 11997– 12019. https:// doi. org/ 10. 5194/ acp12119972012 Huffman, J. A., Perring, A. E., Savage, N. J., Clot, B., Crouzy, B., Tummon, F., Shoshanim, O., Damit, B., Schneider, J., Sivaprakasam, V., Zawadowicz, M. A., Crawford, I., Gallagher, M., Topping, D., Doughty, D. C., Hill, S. C., & Pan, Y. (2020). Real-time sensing of bioaerosols: Review and current perspectives. Aerosol Science and Technology, 54, 465–495. https:// doi. org/ 10. 1080/ 02786 826. 2019. 16647 24 Jacques, S., Lenzo, L., Stevens, K., Lawrence, J., & Tan, K.-C. (2021). An optimized sporulation method for the wheat fungal pathogen Pyrenophora tritici-repentis. Plant Methods, 17, 52. https:// doi. org/ 10. 1186/ s1300702100751-4 Kanaani, H., Hargreaves, M., Ristovski, Z., & Morawska, L. (2007). Performance assessment of UVAPS: Influence of fungal spore age and air exposure. Journal of Aerosol Science, 38, 83–96. https:// doi. org/ 10. 1016/j. jaero sci. 2006. 11. 003 524 Aerobiologia (2025) 41:505–525 Vol:. (1234567890) Könemann, T., Savage, N., Klimach, T., Walter, D., Fröhlich-Nowoisky, J., Su, H., Pöschl, U., Huffman, J. A., & Pöhlker, C. (2019). Spectral Intensity Bioaerosol Sensor (SIBS): An instrument for spectrally resolved fluorescence detection of single particles in real time. Atmospheric Measurement Techniques, 12, 1337–1363. https:// doi. org/ 10. 5194/ amt1213372019 Li, X., & Fu, H. (2020). Fungal spore aerosolization at different positions of a growing colony blown by airflow. Aerosol and Air Quality Research, 20, 2826–2833. https:// doi. org/ 10. 4209/ aaqr. 2020. 09. 0565 Lieberherr, G., Auderset, K., Calpini, B., Clot, B., Crouzy, B., Gysel-Beer, M., Konzelmann, T., Manzano, J., Mihajlovic, A., Moallemi, A., O’Connor, D., Sikoparija, B., Sauvageat, E., Tummon, F., & Vasilatou, K. (2021). Assessment of real-time bioaerosol particle counters using reference chamber experiments. Atmospheric Measurement Techniques, 14, 7693–7706. https:// doi. org/ 10. 5194/ amt1476932021 Magyar, D., Vass, M., & Li, D. W. (2016). Dispersal Strategies of Microfungi. In D. W. Li (Ed.), Biology of microfungi. Fungal biology. Springer. https:// doi. org/ 10. 1007/ 978-331929137-6_ 14 Markey, E., Clancy, J. H., Martínez-Bracero, M., SardaEstève, R., Baisnée, D., McGillicuddy, E. J., Sewell, G., Skjøth, C., & O’Connor, D. J. (2024). Spectroscopic detection of bioaerosols with the wibs-4+: Anthropogenic and meteorological impacts. Science of the Total Environment, 943, 173649. Matavulj, P., Panić, M., Šikoparija, B., Tešendić, D., Radovanović, M., & Brdar, S. (2023). Advanced CNN architectures for pollen classification: Design and comprehensive evaluation. Applied Artificial Intelligence, 37, 2157593. https:// doi. org/ 10. 1080/ 08839 514. 2022. 21575 93 Maya-Manzano, J. M., Tummon, F., Abt, R., Allan, N., Bunderson, L., Clot, B., Crouzy, B., Daunys, G., Erb, S., Gonzalez-Alonso, M., Graf, E., Grewling, Ł, Haus, J., Kadantsev, E., Kawashima, S., Martinez-Bracero, M., Matavulj, P., Mills, S., Niederberger, E., … Buters, J. (2023). Towards European automatic bioaerosol monitoring: Comparison of 9 automatic pollen observational instruments with classic Hirst-type traps. Science of the Total Environment, 866, Article 161220. https:// doi. org/ 10. 1016/j. scito tenv. 2022. 161220 O’Connor, D. J., Iacopino, D., Healy, D. A., O’Sullivan, D., & Sodeau, J. R. (2011). The intrinsic fluorescence spectra of selected pollen and fungal spores. Atmospheric Environment, 45(35), 6451–6458. Pogner, C. E., Graf, E., Niederberger, E., & Gorfer, M. (2024). What do spore particles look like - use of realtime measurements and holography imaging to view spore particles from four bioaerosol generators. Aerosol Science and Technology, 58(7), 779–795. https:// doi. org/ 10. 1080/ 02786 826. 2024. 23385 44 Pöhlker, C., Huffman, J. A., & Pöschl, U. (2011). Autofluorescence of atmospheric bioaerosols–fluorescent biomolecules and potential interferences. Atmospheric Measurement Techniques Discussions, 4(5), 5857–5933. Robinson, E. S., Gao, R.-S., Schwarz, J. P., Fahey, D. W., & Perring, A. E. (2017). Fluorescence calibration method for single-particle aerosol fluorescence instruments. Atmospheric Measurement Techniques Discussions, 10, 1755– 1768. https:// doi. org/ 10. 5194/ amt1017552017 Saari, S., Mensah-Attipoe, J., Reponen, T., Veijalainen, A. M., Salmela, A., Pasanen, P., & Keskinen, J. (2015). Effects of fungal species, cultivation time, growth substrate, and air exposure velocity on the fluorescence properties of airborne fungal spores. Indoor Air, 25(6), 653–661. Šaulienė, I., Šukienė, L., Daunys, G., Valiulis, G., Vaitkevičius, L., Matavulj, P., Brdar, S., Panic, M., Sikoparija, B., Clot, B., Crouzy, B., & Sofiev, M. (2019). Automatic pollen recognition with the Rapid-E particle counter: The first-level procedure, experience and next steps. Atmos. Meas. Tech., 12, 3435–3452. https:// doi. org/ 10. 5194/ amt1234352019 Sauvageat, E., Zeder, Y., Auderset, K., Calpini, B., Clot, B., Crouzy, B., Konzelmann, T., Lieberherr, G., Tummon, F., & Vasilatou, K. (2020). Real-time pollen monitoring using digital holography. Atmos. Meas. Tech., 13, 1539– 1550. https:// doi. org/ 10. 5194/ amt1315392020 Scheermeyer, E., & Agranovski, I. E. (2009). Design and evaluation of a new device for fungal spore aerosolization for laboratory applications. Journal of Aerosol Science, 40(10), 879–889. Senanayake, I. (2020). Morphological approaches in studying fungi: Collection, examination, isolation, sporulation and preservation. Mycosphere, 11, 2678–2754. https:// doi. org/ 10. 5943/ mycos phere/ 11/1/ 20 BSikoparijaPMatavuljISimovicPRadisicSBrdarVMinicDTesendicEKadantsevJPalamarchukMSofiev2024Classification accuracy and compatibility across devices of a new Rapid-E+ flow cytometerAtmospheric Measurement Techniques10.5194/egusphere-2024-187Sikoparija, B., Matavulj, P., Simovic, I., Radisic, P., Brdar, S., Minic, V., Tesendic, D., Kadantsev, E., Palamarchuk, J., & Sofiev, M. (2024). Classification accuracy and compatibility across devices of a new Rapid-E+ flow cytometer. Atmospheric Measurement Techniques. https:// doi. org/ 10. 5194/ egusp here2024187 Simmons, E. G. (2007). Alternaria: An Identification Manual. CBS Fungal Biodiversity Centre, Utrecht, NL. Simović, I., Matavulj, P., & Šikoparija, B. (2023). Manual and automatic quantification of airborne fungal spores during wheat harvest period. Aerobiologia, 39, 227–239. https:// doi. org/ 10. 1007/ s1045302309788-5 Su, Y.-Y., Qi, Y.-L., & Cai, L. (2012). Induction of sporulation in plant pathogenic fungi. Mycology, 3, 195–200. https:// doi. org/ 10. 1080/ 21501 203. 2012. 719042 Tan, M., Le, Q. V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. https:// doi. org/ 10. 48550/ ARXIV. 1905. 11946 Tešendić, D., Boberić Krstićev, D., Matavulj, P., Brdar, S., Panić, M., Minić, V., & Šikoparija, B. (2022). RealForAll: Real-time system for automatic detection of airborne pollen. Enterprise Information Systems, 16, 1793391. https:// doi. org/ 10. 1080/ 17517 575. 2020. 17933 91 Tordoni, E., Ametrano, C. G., Banchi, E., Ongaro, S., Pallavicini, A., Bacaro, G., & Muggia, L. (2021). Integrated eDNA metabarcoding and morphological analyses assess spatio-temporal patterns of airborne fungal spores. 525 Aerobiologia (2025) 41:505–525 Vol.: (0123456789) Ecological Indicators, 121, Article 107032. https:// doi. org/ 10. 1016/j. ecoli nd. 2020. 107032 Tummon, F., Arboledas, L. A., Bonini, M., Guinot, B., Hicke, M., Jacob, C., Kendrovski, V., McCairns, W., Petermann, E., Peuch, V., Pfaar, O., Sicard, M., Sikoparija, B., & Clot, B. (2021). The need for Pan-European automatic pollen and fungal spore monitoring: A stakeholder workshop position paper. Clinical & Translational All, 11, Article e12015. https:// doi. org/ 10. 1002/ clt2. 12015 Tummon, F., Bruffaerts, N., Celenk, S., Choël, M., Clot, B., Crouzy, B., Galán, C., Gilge, S., Hajkova, L., Mokin, V., O’Connor, D., Rodinkova, V., Sauliene, I., Sikoparija, B., Sofiev, M., Sozinova, O., Tesendic, D., & Vasilatou, K. (2022). Towards standardization of automatic pollen and fungal spore monitoring: Best practices and guidelines. Aerobiologia. https:// doi. org/ 10. 1007/ s1045302209755-6 Tummon, F., Adams-Groom, B., Antunes, C. M., Bruffaerts, N., Buters, J., Cariñanos, P., Celenk, S., Choël, M., Clot, B., Cristofori, A., Crouzy, B., Damialis, A., Fernández, A. R., González, D. F., Galán, C., Gedda, B., Gehrig, R., Gonzalez-Alonso, M., Gottardini, E., … de Weger, L. A. (2024). The role of automatic pollen and fungal spore monitoring across major end-user domains. Aerobiologia. https:// doi. org/ 10. 1007/ s1045302409820-2 Wounderberg, J. H. C., Seidl, M. F., Groenewald, J. Z., de Vries, M., Stielow, J. B., Thomma, B. P. H. J., & Crous, P. W. (2015). Alternaria section Alternaria: Species, formae speciales or pathotypes? Studies in Mycology, 82, 1–21. https:// doi. org/ 10. 1016/j. simyco. 2015. 07. 001 Zhang, M., Su, H., Li, G., Kuhn, U., Li, S., Klimach, T., Hoffmann, T., Fu, P., Pöschl, U., & Cheng, Y. (2021). Highresolution fluorescence spectra of airborne biogenic secondary organic aerosols: Comparisons to primary biological aerosol particles and implications for single-particle measurements. Environmental Science & Technology, 55(24), 16747–16756. https:// doi. org/ 10. 1021/ acs. est. 1c025 36 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.