Re-examination of species limits in Aspergillus section Flavipedes using advanced species delimitation methods and description of four new species.

Since the last revision in 2015, the taxonomy of section Flavipedes evolved rapidly along with the availability of new species delimitation techniques. This study aims to re-evaluate the species boundaries of section Flavipedes members using modern delimitation methods applied to an extended set of strains (n = 90) collected from various environments. The analysis used DNA sequences of three house-keeping genes (benA, CaM, RPB2) and consisted of two steps: application of several single-locus (GMYC, bGMYC, PTP, bPTP) and multi-locus (STACEY) species delimitation methods to sort the isolates into putative species, which were subsequently validated using DELINEATE software that was applied for the first time in fungal taxonomy. As a result, four new species are introduced, i.e. A. alboluteus, A. alboviridis, A. inusitatus and A. lanuginosus, and A. capensis is synonymized with A. iizukae. Phenotypic analyses were performed for the new species and their relatives, and the results showed that the growth parameters at different temperatures and colonies characteristics were useful for differentiation of these taxa. The revised section harbors 18 species, most of them are known from soil. However, the most common species from the section are ecologically diverse, occurring in the indoor environment (six species), clinical samples (five species), food and feed (four species), droppings (four species) and other less common substrates/environments. Due to the occurrence of section Flavipedes species in the clinical material/hospital environment, we also evaluated the susceptibility of 67 strains to six antifungals (amphotericin B, itraconazole, posaconazole, voriconazole, isavuconazole, terbinafine) using the reference EUCAST method. These results showed some potentially clinically relevant differences in susceptibility between species. For example, MICs higher than those observed for A. fumigatus wild-type were found for both triazoles and amphotericin B for A. ardalensis, A. iizukae, and A. spelaeus whereas A. lanuginosus, A. luppiae, A. movilensis, A. neoflavipes, A. olivimuriae and A. suttoniae were comparable to or more susceptible as A. fumigatus. Finally, terbinafine was in vitro active against all species except A. alboviridis.

Introduction

Aspergillus is a large genus of filamentous fungi, which currently contains 446 accepted species and this number is rapidly rising. Aspergilli have traditionally been classified into subgenera and sections, and this classification has been recently revised and updated with the addition of series rank (Houbraken ). According to this most up-to-date overview, the accepted species are distributed over six subgenera, 27 sections, and 75 series. Thom & Church (1926) introduced the A. flavipes group and section Flavipedes was formally established by Gams . The section is close to sections Terrei and Jani (Kocsubé ) and is subdivided in four series: Flavipedes, Neonivei, Olivimuriarum and Spelaei (Houbraken ). Phylogenetic analysis performed by Peterson (2008) demonstrated the presence of undescribed species diversity and the need for a proper taxonomic revision. The section was revised by Hubka , who accepted 10 species, two of which, A. frequens and A. mangaliensis, are synonymous to A. micronesiensis and A. templicola, published independently during the same period (Visagie , Arzanlou ). In addition, Visagie introduced another species, A. capensis, a close relative of A. iizukae, isolated from house dust. Another four species were described since then, namely, A. urmiensis described from hypersaline soils in Iran (Arzanlou ), A. suttoniae from human sputum in the USA (Siqueira ), A. olivimuriae from olive brine in Italy (Crognale ) and A. sakultaensis from a water sample collected in Egypt (Zohri & Al-Bedak 2020). However, the last mentioned species was not validly described [Art. 40.8, Shenzen Code], and the study does not contain sufficient data to clearly classify A. sakultaensis into the current system. The isolate was not available for this study, but the DNA sequence of the internal transcribed spacer (ITS) generated by the authors was identical with some strains of A. templicola.

The species from section Flavipedes occur globally in soil (Klich 2002) and they can also grow as endophytes (El-Elimat ), cause food spoilage (Pitt & Hocking 2009), or contaminate the indoor environment (Visagie ). They are also occasionally isolated from clinical samples and infrequently cause opportunistic human or animal infections (Schultz , Gehlot , Siqueira ). Representatives of section Flavipedes are able to produce a wide range of metabolites, summarized by Frisvad & Larsen (2015). These include mycotoxins such as sterigmatocystin (Tuomi ) and citrinin (Greenhill ), or pharmaceutical drugs (established or potential) such as HMG CoA reductase inhibitor lovastatin (Valera ) and antiviral xanthones (Kang ). Section Flavipedes species are also studied for their biotechnological potential and various biological activities. For example, A. flavipes possesses the potential to act as a biocontrol agent (El-Sayed & Ali 2020), A. polyporicola and A. spelaeus may be employed in the remediation of crude oil contaminated soil (Al-Dhabaan 2021) and A. iizukae produces oxidative enzymes with an industrial application (Noman ).

In this study, we assembled a dataset of 90 strains belonging to section Flavipedes, which were newly isolated or originated from previous studies. We re-examined species boundaries of currently known species using modern species delimitation methods and discovered several new ones. The species delimitation and phylogenetic analyses utilized the DNA sequence data of three house-keeping genes. Phenotypic variability was examined in the species related to the newly discovered ones in order to find additional support for species hypotheses. The methodology of the species delimitation analysis follows up on previous studies within the genus Aspergillus (Sklenář , 2020, Hubka , 2018b) with notable changes in the species validation step, where we used a newly developed program DELINEATE as opposed to utilization of BPP (Bayesian Phylogenetics and Phylogeography) (Yang 2015) in previous studies.

Materials and methods

Strains

The newly isolated strains obtained from the indoor environment were isolated as described previously (Jurjević ) and the remaining strains were mostly obtained from collaborators or culture collections. Detailed information about provenance of the strains is listed in Table 1. Dried holotype and isotype specimens of the newly described species were deposited into the herbarium of the Mycological Department, National Museum, Prague, Czech Republic (PRM) and/or into the herbarium at the Westerdijk Fungal Biodiversity Institute (CBS H; Utrecht, the Netherlands). Nomenclatural novelties and descriptions were deposited in MycoBank (Crous ).

Table 1

Aspergillus strains from section Flavipedes examined in this study.

Species	Strain numbers1	Provenance (substrate, locality, year of isolation, collector/isolator)	GenBank/EMBL accession numbers2
			ITS rDNA	benA	CaM	RPB2
A. alboluteus	CBS 145855T = CCF 5695T = EMSL 2420T = IFM 66815T	USA, Pennsylvania, Philadelphia, outdoor air, 2014, Ž. Jurjević	MW448663	MW478497	MW478511	MW478532
	CBS 145859 = CCF 6201 = EMSL 3060	USA, Florida, Seminole, A/C Vent – swab, 2015, Ž. Jurjević	MW448662	MW478496	MW478510	MW478531
	CBS 145854 = CCF 4916 = EMSL 2311 = IFM 66816	USA, California, indoor air, 2005, Ž. Jurjević	MW448664	MW478498	MW478512	MW478533
	CCF 5849 = EMSL 2446 = IFM 66817	USA, Tennessee, Jackson, storage room - swab, 2014, Ž. Jurjević	MW448665	MW478499	MW478513	MW478534
	DTO 410-I8 = CBS 147065 = CCF 6551	Nigeria, Abia, Isiala Ngwa South, Obuba, multicrop farm, C.N. Ezekiel	MW448666	MW478500	MW478514	MW478535
A. alboviridis	CBS 142665T = FMR 15175T = CCF 6049T = IFM 66819T	Spain, Balearic Islands, Mallorca, Pollença, herbivore dung, 2016, J. Gené & J.P.Z. Siqueira	LT798909	LT798936	LT798937	LT798938
A. ardalensis	NRRL 62824T = CCF 4031T = CCF 4426T = CMF ISB 1688T = CBS 134372T	Spain, Andalucia, Ardales, near Cueva de Doña Trinidad, soil, 2008, A. Nováková	FR733808	HG916683	HG916725	HG916704
	IHEM 17781	France, Giens, hospital environment, 2000, J.-Ph. Bouchara	MW448667	LN909026	MW478515	MW478536
A. flavipes	NRRL 302T = CCF 3067T = IMI 171885T = ATCC 24487T = FRR 0302T	Received by Charles Thom in 1922 from Da Fonseca as Bainier’s culture of Sterigmatocystus flavipes	EF669591	EU014085	EF669549	EF669633
	NRRL 4852 = IMI 345934 = CCF 4836 (ex-type of A. archiflavipes)	Uruguay, dead beetle, received in NRRL from CBS as Blochwitz's strain of A. archiflavipes, W. Herter	LM999909	LM644261	LM644241	LM644260
A. iizukae	NRRL 3750T = CBS 541.69T = IMI 141552T = CCF 4548T	Japan, Gymna Prefecture, Fujioka, soil from stratigraphic drilling core, 1969, J. Sugiyama	EF669597	EU014086	EF669555	EF669639
	CBS 138188T = DTO 179-E6T (ex-type of A. capensis)	South Africa, Cape Town, house dust, 2010, E. Whitfield & K. Mwange	KJ775550	KJ775072	KJ775279	KP987020
	CanS-34A	China, Wuhan, oilseed rape (Brassica napus), between 2008-2010	MK072769	MK215220	MK215219	MK215221
	CCF 1895	Czechia, Most (brown lignite district), soil of spoil bank, 1984, M. Černý	FR727134	FR775336	HG916728	HG916707
	CCF 4033 = CMF ISB 1551 = NA16 = Y14	Czechia, Most (brown lignite district), soil of spoil bank tip, 2004, A. Nováková	FR733809	HG916686	HG916729	MW478537
	CCF 4032 = CMF ISB 1245	Germany, Weissagker Berg near Cottbus, Lusatian brown lignite district, soil of spoil bank, 1999, A. Nováková	HG915894	HG916687	HG916730	HG916708
	CMF ISB 2544	Romania, Dobrogea, Mangalia, soil near Movile Cave, 2011, A. Nováková	HG915895	HG916694	HG916731	HG916709
	CMF ISB 2417	Romania, National Park Apuseni Mountains, Meziad Cave, earthworm casts, 2009, A. Nováková	HG915896	HG916688	HG916732	HG916710
	CMF ISB 2616	Czechia, Ječmeniště, National Nature Monument, soil, 2012, A. Nováková	HG915899	HG916689	HG916733	HG916711
	CMF ISB 2617	Czechia, Kolby, National Reservation Pouzdřanská step, soil, 2012, A. Nováková	HG915897	HG916692	HG916734	HG916712
	CMF ISB 2618	Czechia, Kolby, National Reservation Pouzdřanská step, soil, 2012, A. Nováková	HG915898	HG916693	HG916735	HG916713
	CMF ISB 2619	Czechia, Kolby, National Reservation Pouzdřanská step, earthworm casts, 2012, A. Nováková	HG915900	HG916690	HG916736	HG916714
	CMF ISB 2620	Czechia, Kolby, National Reservation Pouzdřanská step, earthworm casts, 2012, A. Nováková	HG915901	HG916691	HG916737	HG916715
	NRRL 58963 = CCF 4843 = ZJ 1256	USA, Illinois, indoor air of a home, 2009, Ž. Jurjević	LM644237	LM644268	LM644245	MW478538
	CCF 4844 = ZJ 1817	USA, Idaho, Boise, indoor air of a home, 2012, Ž. Jurjević	LM644238	LM644269	LM644244	MW478539
	CCF 4845 = S746	Romania, Movile Cave, cave sediment, 2013, A. Nováková	LM999906	LM644270	LM644243	MW478540
	UTHSCSA DI14-219	USA, Illinois, human, bronchoalveolar lavage, 2012, D. Sutton	LT899477	LT899528	LT899579	LT899634
	FMR 15051	Spain, Catalonia, Els Ports Natural Park, herbivore dung, 2016, J. Gené	LT899475	LT798968	LT899577	LT899632
	FMR 15606	Spain, Catalonia, Els Ports Natural Park, herbivore dung, 2016, J. Gené	LT899476	LT798969	LT899578	LT899633
	CCF 5786 = EMSL 3408	USA, Florida, Saint Petersburg, bedroom floor – swab, 2016, Ž. Jurjević	MW448668	MW478501	MW478516	MW478541
A. inusitatus	DTO 121-G5T = CBS 147044T = CCF 6552T	Tunisia, Ras Rajel, soil in oak forest, 2009, C. Silva Pereira	MW448669	MW478502	MW478517	MW478542
A. lanuginosus	NRRL 4610T = IMI 350352T = CCF 4551T = IFM 66818T	Haiti, Fonds Parisien, soil	EF669604	EU014080	EF669562	EF669646
A. luppiae	NRRL 6326T = CBS 653.74T = CCF 4545T	France, Provence, near Aups, natural truffle soil, 1972, A.M. Luppi-Mosca	EF669617	EU014079	EF669575	EF669659
A. micronesiensis	CBS 138183T = DTO 267-D5T	Federated States of Micronesia, Yela of Kosrae Island, house dust, 2010, E. Whitfield & K. Mwange	KJ775548	KJ775085	KP987067	KP987023
	NRRL 4578 = ATCC 16805 = CBS 586.65 = IMI 135423 = CCF 4555 (ex-type of A. frequens)	Haiti, soil, 1960, J. Rabel	EF669602	EU014082	EF669560	EF669644
	CCF 2026	Czechia, Prague, archive material, 1986, O. Fassatiová	HG915893	HG916684	HG916726	HG916705
	NRRL 295 = ATCC 16814 = CBS 585.65 = IMI 135422 = CCF 4554 = FRR 0295	USA, Minnesota, dairy products, 1933, H. Macy	EF669588	EU014081	EF669546	EF669630
	CCF 4005	Czechia, Hradec Králové, hospital indoor air, 2005, V. Buchta	FR727135	HG916685	HG916727	HG916706
	NRRL 4263 = CCF 4556	India, Dehradun New Forest, soil, 1955, K.B. Bakshi	EF669600	EU014083	EF669558	EF669642
	NRRL 286 = ATCC 1030 = FRR 0286	Received in NNRL from Dr. J. Westerdijk (CBS)	AY373849	LM644262	LM644246	LM644258
	NRRL 26246 = CCF 4838	China, soil, 1944	LM999905	LM644263	LM644247	LM644257
	NRRL 58660 = CCF 4839 = ZJ 1111	Trinidad & Tobago, indoor air of a home, 2009, Ž. Jurjević	LM644239	LM644264	LM644248	MW478543
	NRRL 58682 = CCF 4840 = ZJ 1132	Puerto Rico, indoor air of a home, 2009, Ž. Jurjević	LM644240	LM644265	LM644251	MW478544
	NRRL 58899 = CCF 4841 = ZJ 1267	USA, New York, indoor air of a home, 2009, Ž. Jurjević	LM999903	LM644266	LM644249	MW478545
	NRRL 58598 = CCF 4842 = ZJ 1038	USA, New Jersey, indoor air of a home, 2008, Ž. Jurjević	LM999904	LM644267	LM644250	MW478546
	IHEM 18446	Belgium, Brussels, floor in hospital, 2001, BCCM/IHEM collection	MW448670	LN909029	MW478518	MW478547
	IHEM 662	Belgium, Brussels, indoor air in hospital, 1980, BCCM/IHEM collection	MW448671	LN909030	MW478519	MW478548
	IHEM 22506 = RV 21840	Belgium, Liège, human lung, 1967, University Hospital Liège	MW448672	LN909028	MW478521	MW478549
	IHEM 22505 = RV 42608	Belgium, Antwerp, human sputum (male), 1979, D. Van Vijver	MW448673	LN909027	MW478520	MW478550
	CBS 147045 = DTO 247-H3	Mexico, Sayulita, hotel room, house dust, 2009, A. Amend & E. Whitfield & K. Mwange	KP987079	KP987047	KP987062	KP987036
	IMI 357699 = DTO 305-B6 = IBT 23707 (ex-type of A. sunderbaniinom. inval.)	India, West Bengal, soil	KP987084	KP987052	KP987069	KP987026
	UTHSCSA DI14-214	USA, California, canine urine, 2012, D. Sutton	LT899480	LT899529	LT899582	LT899637
	FMR 15737	Spain, Canary Islands, Tenerife, 2016, J. Gené	LT899479	LT798971	LT899581	LT899636
A. movilensis	NRRL 62819T = CCF 4410T = CMF ISB 2614T = CBS 134395T	Romania, Mangalia, Dobrogea, soil near Movile Cave, 2011, A. Nováková	HG915904	HG916697	HG916740	HG916718
	CBS 139559 = CCTU 749 = DTO 203-C9 = IBT 32594	Iran, Urmia, Kabodan Island, soil, between 2011 and 2012, U. Ghosta & R. Samadi	KP987075	KP987043	KP987058	KP987032
	CBS 139562 = CCTU 788 = DTO 203-H3	Iran, Urmia, Kabodan Island, soil, between 2011 and 2012, U. Ghosta & R. Samadi	KP987078	KP987046	KP987061	KP987035
	S1040	Romania, soil above the Movile cave, 2014, A. Nováková	MW448674	MW478503	MW478522	MW478551
A. neoflavipes	NNRL 5504T = ATCC 24484T = CBS 260.73T = IMI 171883T = IFM 40894T = CCF 4552T	Thailand, Pak Thong Chai, cellulosic material buried in forest soil, 1968, C. Klinsukont	EF669614	EU014084	EF669572	EF669656
A. olivimuriae	NRRL 66783T = CCF 6208T	Italy, Viterbo, olive curing brine, 2012, S. Crognale	MH298877	MH492010	MH492011	MH492012
A. polyporicola	NRRL 32683T = CCF 4553T	USA, Hawaii, Hilo, Alien Wet Forest Zoo, basidioma of Earliella scabrosa (Polyporales), 2003, D.T. Wicklow	EF669595	EU014088	EF669553	EF669637
	NRRL 58570 = CCF 4828	USA, Hawaii, Alien Wet Forest, basidioma of Rigidoporus microporus (Polyporales), 2003, D.T. Wicklow	HQ288052	LM644274	LM644252	LM644254
	CCF 5427 = EMSL 2612	USA, New York, Holbrook, bedroom - settle plates, 2014, Ž. Jurjević	MW448675	MW478504	MW478523	MW478552
	CCF 6262 = EMSL 3169	USA, crawled space - settle plates, 2015, Ž. Jurjević	MW448676	MW478505	MW478524	MW478553
A. spelaeus	NRRL 62826T = CCF 4425T = CMF ISB 2615T = CBS 134371T	Spain, Andalusia, Nerja Cave, cave sediment, 2011, A. Nováková	HG915905	HG916698	HG916741	HG916719
	NRRL 62827 = CCF 544	Czechia, Bohemian Karst, Doutnáč hill near Srbsko, soil, 1961, O. Fassatiová	HG915906	HG916699	HG916742	HG916720
	CCF 4699 = CMF ISB 2659	Czechia, Hostěradice, National Nature Monument U Kapličky, Allolobophora hrabei intestine, 2012, A. Nováková	HG915907	HG916700	HG916743	HG916721
	CCF 4679 = CMF ISB 2663	Czechia, Ječmeniště, National Nature Monument, soil, 2012, A. Nováková	HG915908	HG916701	HG916744	HG916722
	CCF 4680	Spain, Andalusia, Nerja Cave, cave sediment, 2012, A. Nováková	HG915909	HG916702	HG916745	HG916723
	CCF 4697	Spain, Andalusia, Nerja Cave, cave air, 2012, A. Nováková	HG915910	HG916703	HG916746	HG916724
	EMSL 4874	USA, Georgia, Sandersville, crawlspace (swab), 2018, Ž. Jurjević	MW448677	MW478506	MW478525	MW478554
	CCF 4886 = S716	Spain, Andalusia, Nerja Cave, cave sediment, 2012, A. Nováková	LM999908	LM644272	HG916748	LM644259
	CCF 4829 = BMP 3043	USA, Arizona, Benson, Kartchner Caverns, speleothem surface, 2008, M. Vaughan	HQ832962	LM644273	LM644253	LM644255
	CBS 115952	Germany, dust, S. Ammermann	MW448678	MW478507	MW478526	MW478555
	UTHSCSA DI17-89 (UTHSCSA 04-3307)	USA, Missouri, human forearm, 2004, D. Sutton	LT899491	LT899538	LT899593	LT899648
	FMR 14606	Spain, Balearic Islands, Mallorca, soil, 2012, J. Gené	LT899488	LT899537	LT899590	LT899645
	FMR 15176	Spain, Balearic Islands, Mallorca, herbivore dung, 2016, J. Gené & J.P.Z. Siqueira	LT899489	LT798972	LT899591	LT899646
	FMR 15223	Spain, Balearic Islands, Mallorca, herbivore dung, 2016, J. Gené & J.P.Z. Siqueira	LT899490	LT798976	LT899592	LT899647
	CCF 6263 = EMSL 4125	USA, New Jersey, Marlton, black walnut (Juglans nigra), 2017, Ž. Jurjević	MW448679	MW478508	MW478527	MW478556
	CCF 6248 = EMSL 4140	USA, New Jersey, Marlton, black walnut (Juglans nigra), 2017, Ž. Jurjević	MW448680	MW478509	MW478528	MW478557
A. suttoniae	UTHSCSA DI14-215T = FMR 13523T	USA, human sputum, 2014, D. Sutton	LT899487	LT899536	LT899589	LT899644
A. templicola	CBS 138181T = DTO 270-C6T	Mexico, Sayulita, dust from church, 2010, E. Whitfield & K. Mwange	KJ775545	KJ775092	KJ775394	KP987017
	CBS 138180 = DTO 267-H4	Thailand, Bangkok, house dust, 2010, E. Whitfield & K. Mwange	KP987081	KJ775087	KP987064	KP987038
	NRRL 62825 = CCF 4698 = CMF ISB 2662 (ex-type of A. mangaliensis)	Romania, Mangalia, soil near Moville Cave, 2012, A. Nováková	HG915902	HG916695	HG916738	HG916716
	CCF 869 = NRRL 62823	China, industrial material, 1955, V. Zánová	HG915903	HG916696	HG916739	HG916717
	NRRL 4893 = IMI 343701 = CCF 4846	Japan, soil	LM999907	LM644271	LM644242	LM644256
	IHEM 14393	Belgium, Charleroi, furniture in hospital 1998, BCCM/IHEM collection	MW448681	LN909024	MW478529	MW478558
	DK-T43978	Denmark, Copenhagen, bronchoalveolar lavage, 2014	MW448682	LN909025	MW478530	MW478559
A. urmiensis	CBS 139558T = CCTU 742T = DTO 203-C2T = IBT 32593T	Iran, Urmia, Jade Darya (seaside), soil, 2011, U. Ghosta & R. Samadi	KP987073	KP987041	KP987056	KP987030
	CBS 139557 = CCTU 734 = DTO 203-B3 = IBT 32597	Iran, Jade Darya (seaside), soil, 2011, U. Ghosta & R. Samadi	KP987072	KP987039	KP987055	KP987029
	CBS 139766 = CCTU 743 = DTO 203-C3 = IBT 32598	Iran, Jade Darya (seaside), soil, 2011, U. Ghosta & R. Samadi	KP987074	KP987042	KP987057	KP987031

Acronyms of culture collections in alphabetic order: ATCC, American Type Culture Collection, Manassas, Virginia, USA; BMP, Barry M. Pryor laboratory culture collection, Tucson, Arizona, USA; CBS, Westerdijk Fungal Biodiversity Institute (formerly Centraalbureau voor Schimmelcultures), Utrecht, the Netherlands; CCF, Culture Collection of Fungi, Department of Botany, Charles University, Prague, Czech Republic; CMF ISB, Collection of Microscopic Fungi of the Institute of Soil Biology, Academy of Sciences of the Czech Republic, České Budějovice, Czech Republic; CCTU, Culture Collection of Tabriz University, Tabriz, Iran; FMR, Faculty of Medicine, Reus, Spain; FRR, Food Fungal Culture Collection, North Ride, Australia; IFM, Collection at the Medical Mycology Research Center, Chiba University, Chiba, Japan; IHEM (BCCM/IHEM), Belgian Coordinated Collections of Micro-organisms, Fungi Collection: Human and Animal Health, Sciensano, Brussels, Belgium; IMI, CABI’s collection of fungi and bacteria, Egham, UK; NRRL, Agricultural Research Service Culture Collection, Peoria, Illinois, USA; UTHSCSA, Collection of Fungus Testing Laboratory, University of Texas, Health Science Center, San Antonio, USA.

Sequences generated in this study are designated by bold print.

Molecular studies

Total genomic DNA was isolated from 7-d-old cultures with ArchivePure DNAyeast (5 PRIME Inc., Gaithersburg, MD, USA) or NucleoSpin® Soil (Macherey–Nagel, Düren, Germany) isolation kits. The quality of the isolated DNA was verified using a NanoDrop 1 000 Spectrophotometer.

The ITS region of rDNA was amplified using forward primer ITS1 (White ) and reverse primers NL4 (O’Donnell 1993) or ITS4 (White ), a part of the β-tubulin gene (benA) was amplified using forward primers Bt2a (Glass & Donaldson 1995), T10 (O'Donnell & Cigelnik 1997) or Ben2f (Hubka & Kolařík 2012) and reverse primer Bt2b (Glass & Donaldson 1995), a part of the calmodulin gene (CaM) was amplified using forward primers CF1L, CF1M (Peterson 2008) or cmd5 (Hong ) and reverse primers CF4 (Peterson 2008) or cmd6 (Hong ) and a part of the RNA polymerase II second largest subunit gene (RPB2) was amplified using forward primer fRPB2-5F and reverse primer fRPB2-7CR (Liu ). Various primer pairs were used for the amplification of ITS, benA and CaM loci because the sequences were generated across various research groups. Thus, it was not due to the failure of PCR with some primer combinations.

The PCR reaction volume of 25 μL contained 1.2 μL (10-20 ng) of DNA, 1 μL of both primers ( 10 μM), 0.25 μL of DreamTaq DNA Polymerase (Thermo Scientific, Waltham, MA) and 2.5 μl of DreamTaq PCR buffer and 2.5 μl of dNTP. The ITS rDNA, benA and CaM fragments were amplified using following thermal cycle profile: 93 °C/2 min; 30 cycles of 93 °C/30 s; 55 °C/30 s; 72 °C/60 s; 72 °C/10 min. Partial RPB2 gene fragments were amplified using above-mentioned cycle or touchdown thermal-cycling: 93 °C/2 min; 5 cycles of 93 °C/30 s, 65–60 °C/30 s, 72 °C/60 s; 38 cycles of 93 °C/30 s, 55 °C/30 s, 72 °C/60 s; 72 °C/10 min. PCR products were purified with ethanol and sodium acetate in a 96-well plate; 2 μL of 3 M NaOAc and 60 μL of 96 % EtOH was mixed with 20 μL of PCR product. The plate was sealed, twisted several times and incubated in the refrigerator for 20 min. After incubation, the plate was centrifuged for 30 min at 4 °C and 3 000 rpm, the supernatant was removed and 85 μL 70 % EtOH was added. The plate was centrifuged for 15 min at 4 °C and 3 000 rpm, the supernatant was subsequently removed, the pellet was dried at room temperature for 20 min and resuspended in 10 μL of H2O.

Newly obtained DNA sequences were inspected in FinchTV v. 1.4 (available online https://digitalworldbiology.com/FinchTV) and assembled in Bioedit v. 7.0.5 (Hall 1999). Multiple sequence alignments were created in MAFFT v. 7 (Katoh & Standley 2013) using the G-INS-I strategy. Sequences were deposited into GenBank with accession numbers shown in Table 1. All alignments are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.dz08kprxj.

Phylogenetic analysis and species delimitation

In order to demonstrate the phylogenetic relationships within section Flavipedes, we calculated a Maximum Likelihood (ML) tree in IQ-TREE v. 2.0 (Nguyen ) using a concatenated alignment of all four loci as input. The inference was setup as partitioned analysis and the best-fitting model for each locus was determined using the Bayesian information criterion (BIC) in jModelTest v. 2.1.7 (Posada 2008). Selected models are listed in Table 2 together with the respective alignment statistics. To determine the branch support, the analysis ran for 1 000 bootstrap replicates.

Table 2

Alignment characteristics and substitution models according to Bayesian Information Criterion.1

Dataset (Series)	Locus	Alignment length	Variable sites	Parsimony informative sites	Substitution model
Flavipedes	benA	512	138	102	K80+G
	CaM	560	144	116	TrNef+G
	RPB2	1 013	161	129	TrNef+G
	ITS rDNA	539	14	12	F81+I
Spelaei	benA	507	127	108	K80+G
	CaM	731	193	164	TrNef+G
	RPB2	1 009	138	113	TrNef+G
	ITS rDNA	549	21	13	HKY

proposed by jModelTest v. 2.1.7 (Posada 2008).

For the purpose of species delimitation analyses, the dataset was split into two parts corresponding to series Flavipedes and Spelaei as designated by Houbraken . Series Neonivei and Olivimuriarum were excluded as these are single species series and phylogenetically distinct from series Flavipedes and Spelaei. The ITS rDNA region was excluded due to its low number of informative sites. Omission of the ITS region from the phylogenetic analysis is a common practice as discussed previously (Chen ). The following steps were performed separately for both datasets.

The haplotype function from R v. 4.0.2 (R Core Team 2015) package pegas (Paradis 2010) was used to retain only unique sequences in alignments. The best fitting models obtained in jModelTest v. 2.1.7 (Posada 2008) are listed in Table 2.

We used one multi-locus method, STACEY (Jones 2017), and four single-locus species delimitation methods, GMYC (Fujisawa & Barraclough 2013), bGMYC (Reid & Carstens 2012), PTP and bPTP (Zhang ), to create hypotheses about species boundaries. The multi-locus species delimitation analysis was performed in BEAST v. 2.6.3 (Bouckaert ) add-on STACEY v. 1.2.5 (Jones 2017) with the following parameters. The chain length was set to 1 × 108 generations, the species tree prior was set to the Yule model, growth rate prior was set to lognormal distribution (M = 5, S = 2), clock rate priors for all loci were set to lognormal distribution (M = 0, S = 1), PopPriorScale prior was set to lognormal distribution (M = -7, S = 2) and relativeDeathRate prior was set to beta distribution (α = 1, β = 1 000). The output was processed with SpeciesDelimitationAnalyzer (Jones 2017). The ultrametric input trees for the GMYC method (Fujisawa & Barraclough 2013) were calculated in BEAST v. 2.6.3 (Bouckaert ) with chain length 1 × 107 generations. The GMYC analysis was performed in R v. 4.0.2 (R Core Team 2015) with the package splits (Fujisawa & Barraclough 2013). One hundred ultrametric trees from the BEAST v. 2.6.3 inference were randomly selected using R v. 4.0.2 package ape (Paradis ) after discarding the initial 25 % of the trees as burn-in, and then used as input for the bGMYC method. The analysis was performed in R v. 3.4.1 with package bgmyc (Reid & Carstens 2012). One thousand maximum likelihood standard bootstrap trees were calculated in IQ-TREE v. 2.0 (Nguyen ) and used as input for PTP method. The analysis was run in Python v. 3 (van Rossum & Drake 2019) package ptp (Zhang ). The bPTP was also performed in Python package ptp (Zhang ). Recommended inputs for this method are trees from Bayesian inference, but without the requirement of ultrametricity. In this case, we used trees from a mcmc run in MrBayes v. 3.2.7 (Ronquist ). Phylogenetic trees generated during STACEY analysis were used for the presentation of species delimitation results analysis. The graphical outputs were created in iTOL (Interactive Tree Of Life) (Letunic & Bork 2016).

Finally, we tested different species boundaries hypotheses in DELINEATE (Sukumaran et al. 2021). The analysis was performed according to the manual available online. Briefly, the dataset was split into hypothetical populations with “A10” analysis in BPP v. 4.3 (Yang 2015) (Supplementary Table S1). The starBEAST (Heled & Drummond 2010) implemented in BEAST v. 2.6.3 (Bouckaert ) was used to estimate the species tree. The populations delimited by BPP were lumped into species based on the results of species delimitation methods and phenotypic characters, with several populations always left unassigned to be delimited by DELINEATE. Four models of species boundaries were set up for the Flavipedes series, 14 models for the Spelaei series and six models for both series analyzed together. The analysis was run in Python v. 3 (van Rossum & Drake 2019) package delineate (available online https://jeetsukumaran.github.io/delineate/).

Phenotypic studies

The macromorphology of colonies was studied on Czapek yeast autolysate agar (CYA; Fluka, Buchs, Switzerland), Czapek-Dox agar (CZA), malt extract agar (MEA; Oxoid, Melbourne, Australia) (Samson ), oatmeal agar (OA; Difco, La Ponte de Claix, France) and CYA supplemented with 20 % sucrose (CY20S). The strains were inoculated in three points on 90 mm Petri dishes and incubated at 25 °C in darkness. Cardinal temperatures were determined for A. movilensis and its relatives. The strains were grown on MEA for 14 d at 10, 15, 20, 25, 30, 35, 37 and 40 °C in darkness. For the description of colony colours, we used the hexadecimal colour codes and the names were assigned according to website https://coolors.co/. Colony details were documented using an Olympus SZX2-ILLT dissecting microscope (Tokyo, Japan) equipped with an Olympus DP27 digital camera.

Micromorphological characters were observed from 14-d-old colonies grown on MEA. Every character was measured 35 times for each isolate. Lactic acid (60 %) was used as the mounting medium. Photographs were taken using an Olympus BX51 microscope with an Olympus DP72 camera. The picture processing and preparation of photographic plates was done in Adobe Photoshop CS6.

Antifungal susceptibility testing (EUCAST method)

The determination of the minimum inhibitory concentrations (MICs) of antifungal agents was carried out according to the reference European Committee on Antimicrobial Susceptibility testing (EUCAST) guidelines (E.Def 9.3.2). Pure antifungal substances were stored in aliquots at -80 °C and stock solutions prepared in DMSO (5 000 mg/L; Sigma-Aldrich, Brøndby, Denmark). Cell-culture-treated NuncTM MicroWellTM 96-Well Microplates (ThermoFisher Scientific, catalogue no. 167008) were used throughout. Microtitre plates with 2-fold dilutions were prepared using serial dilution (with two pipette tip changes) and frozen at -80 °C for at least 24 h prior to use. The following antifungal agents (final concentration ranges) were included: amphotericin B (0.004–4 mg/L; Sigma-Aldrich, Germany), itraconazole (0.004–4 mg/L and 0.016–16 mg/L; Sigma-Aldrich), posaconazole (0.004–4 mg/L; MSD, Ballerup, Denmark), voriconazole (0.004–4 mg/L and 0.016–16 mg/L; Pfizer A/S, Ballerup, Denmark and Sigma-Aldrich), isavuconazole (0.008–8 mg/L and 0.016–16 mg/L; Basilea Pharmaceutica International Ltd, Basel, Switzerland), and terbinafine (0.004–4 mg/L; Sigma-Aldrich). Two concentration ranges for some antifungals were used because the isolates were analyzed in two batches and, as the breakpoints are low, we decided to skip the highest concentrations for the second batch and move the range toward lower concentrations. Plates were incubated at 37 °C (or 25 °C for species with insufficient growth at 37 °C) for 48 h. Aspergillus fumigatus ATCC 204305 was included as quality control (Arendrup ).

Results

The species delimitation analyses were mostly divided into two parts corresponding to the series Flavipedes and Spelaei according to Houbraken . The analysis of each series consisted of two steps (for detailed description see Materials and Methods).

Species delimitation and validation in the series Flavipedes

The results of various species delimitation methods in the series Flavipedes (Fig. 1A) were in broad agreement and unequivocally supported the species A. flavipes, A. neoflavipes, A. micronesiensis, A. suttoniae and A. urmiensis. Aspergillus ardalensis was split into two tentative species by bPTP using RPB2 as input, but a broader concept was supported by all other methods (Fig. 1A). Similarly, A. templicola was split into two tentative species by GMYC using RPB2 as input and by bPTP using benA as input; but all other methods resolved A. templicola as a single species. Less stable results were obtained in the node comprising A. iizukae and A. capensis. All single-locus methods delimited “A. capensis lineage” represented by CBS 138188T, CCF 5786 and CanS-34A as separate species using CaM as input. When benA was used as input, the methods showed variable results. The bGMYC method lumped A. capensis and A. iizukae together, PTP and bPTP delimited several singleton species and GMYC delimited three species (with A. capensis consisting of CBS 138188T and A. iizukae strain NRRL 58963). These results together with unstable position of particular strains in single-locus trees (Fig. 2) suggest ongoing recombination within the clade formed by A. iizukae and A. capensis.

Fig. 1

Schematic representation of species delimitation results in the series Flavipedes (A) and Spelaei (B). The analyses were based on three loci (benA, CaM, RPB2) and utilized one multi-locus method (STACEY) and four single-locus methods (GMYC, bGMYC, PTP, bPTP). Only strains with unique haplotypes were used (strains with identical sequences are represented by one tip in the tree). The results are depicted by coloured bars with different colours indicating tentative species delimited by each method. The asterisk (∗) sign designates singleton species delimited by the particular methods. Ex-type isolates are highlighted with bold font. The phylogenetic tree was calculated during STACEY analysis and is used solely for the comprehensive presentation of the results from different methods.

Fig. 2

Single-locus Maximum Likelihood trees of clade containing isolates of Aspergillus iizuake and “A. capensis”. The trees were calculated in IQ-TREE v. 2.0 based on sequences of ITS region and benA, CaM and RPB2 loci. The ex-type isolates are designated by a superscript T. The strains with unstable position across phylogenies and thus causing incongruences are highlighted by colours.

Detailed results of STACEY (Fig. 3A and Supplementary Fig. S1 with different values of collapseheight parameter) with similarity matrix displaying the probability of assignment of strains to particular species suggest that the support for splitting A. iizukae/A. capensis into two species is lower than in other species, e.g., in A. templicola or A. micronesiensis which were not split by any single-locus species delimitation methods.

Fig. 3

The results of species delimitation by STACEY add-on of BEAST v. 2 in the series Flavipedes (A) and Spelaei (B) with the chosen collapseheight parameter = 0.001. The similarity matrices give the posterior probability of every two isolates belonging to the same multi-species coalescent cluster (tentative species). The darkest brown shade corresponds to a posterior probability of 1, while a white colour is equal to 0. The horizontal and vertical lines in the similarity matrices depict the species boundaries proposed by the analysis. Only strains with unique multilocus haplotypes were used in the analysis (strains with identical haplotype are represented by one tip in the tree). Ex-type isolates are highlighted with bold font. Presented phylogenetic trees were calculated in STACEY.

In the species validation step (Fig. 4A), we set up four different models focusing mainly on A. iizukae/A. capensis clade. The first model left all populations of A. iizukae and A. capensis delimited by BPP (Yang 2015) (Supplementary Table S1) unassigned into species, i.e., free to be delimited. These unassigned populations are depicted by brown coloured bars in the Fig. 4, while populations of other species were assigned according to the delimitation in the first step - grey bars. As a result, A. iizukae and A. capensis populations were lumped together into one species - depicted by red rectangle around bars (Fig. 4A). In the second model, populations of A. templicola were left unassigned and the results lumped them together within one species. Third and fourth model left unassigned all populations of A. templicola and several populations of “A. capensis”, A. iizukae and A. micronesiensis and the results always supported broad species definition of A. templicola, A. miconesiensis and A. iizukae (comprising A. capensis). Overall, the results supported all currently accepted species within series Flavipedes with the exception of A. capensis, which is therefore placed in synonymy with A. iizukae. This finding is also supported by morphological observations as A. iizukae and A. capensis are not distinguishable (Visagie ).

Fig. 4

The results of species validation using DELINEATE in the series Flavipedes (A) and Spelaei (B). The populations of each species were delimited by BPP (Supplementary Table S1) and the displayed tree was calculated in starBEAST. The bars on the right side of the tree depicts the setting and results of each model. The grey bars represent the predefined species (locked in the analysis), while the brown bars represent unassigned populations left free to be delimited. The red rectangles depict the resulting solution proposed by DELINEATE.

Species delimitation and validation in the series Spelaei

The species delimitation in the series Spelaei was less clear compared to series Flavipedes. Based on the results described below, we decided to introduce four new species, namely Aspergillus alboluteus, A. alboviridis, A. inusitatus and A. lanuginosus (see section Taxonomy).

In the first part of the analysis A. luppiae was the only species delimited by all methods without any exception (Fig. 1B). Aspergillus spelaeus and A. polyporicola were lumped together by PTP and bPTP using benA as input, and also by GMYC, bGMYC and PTP using CaM as an input. On the other hand, A. polyporicola was split into two species by bPTP using CaM and RPB2 as input. The four newly described species A. alboluteus, A. alboviridis, A. inusitatus and A. lanuginosus were all delimited by GMYC using benA and bPTP using all three genes (A. alboluteus was even divided into two species by bPTP using RPB2 as an input). The GMYC method with RPB2 as input delimited only one species in this clade, containing the four mentioned species and A. movilensis; and the remaining methods based on RPB2 delimited two species. If only these results would be considered, it might not be enough to support delimitation of five species, but there were striking morphological differences (see section Phenotype analysis) supporting the narrower species concept. This is also in agreement with the results of a multi-locus method STACEY which supported all five species in this clade, and at the same time supported the recognition of A. spelaeus and A. polyporicola as separate species (Fig. 1B). Especially the morphology of A. inusitatus does not allow its inclusion into any other species and denies the possibility of a broader concept in this clade.

Detailed results of STACEY show that when the collapseheight parameter is low enough to consider A. inusitatus a separate species (Fig. 3B and Supplementary Fig. S2B), the clade containing A. movilensis should be also split into three species (A. movilensis, A. alboviridis and A. lanuginosus). If the collapseheight parameter is too high (Supplementary Fig. S2A) to delimit A. inusitatus, also the support for A. polyporicola and A. spelaeus decreases.

In the species validation step (Fig. 4B), we set up 14 different models for testing the species hypotheses. The first model left unassigned populations of A. alboviridis, A. lanuginosus, A. inusitatus, A. polyporicola and population 1 of A. alboluteus. The results supported A. polyporicola but lumped together A. inusitatus/A. alboluteus and A. movilensis/A. alboviridis/A. lanuginosus. The results of the second model with similar setting except for predefined species status of A. polyporicola were identical. In the third and fourth model, A. inusitatus and A. alboluteus were defined as separate species. Aspergillus alboviridis and A. lanuginosus were left unassigned in the third model and in the fourth model, A. alboviridis, A. lanuginosus and A. movilensis were left unassigned. The results were similar in both cases, recognizing A. alboviridis, A. lanuginosus and A. movilensis as separate species. The fifth and sixth model were analogous to model 3 and 4, leaving A. alboviridis and A. lanuginosus (model 5); and A. alboviridis, A. lanuginosus and A. movilensis (model 6), respectively, free to be delimited. Unlike models 3 and 4, in case of models 5 and 6, A. inusitatus was predefined as a part of A. alboluteus. These models resulted in lumping of A. alboviridis, A. lanuginosus and A. movilensis into one species. The models 7–10 tested the opposite hypothesis, leaving A. inusitatus and either one population (models 7 and 9) or all populations (models 8 and 10) of A. alboluteus unassigned, and A. alboviridis and A. lanuginosus predefined either as separate species (models 7 and 8) or lumped together with A. movilensis (model 9 and 10). The final results corresponded with models 3–6. In case of A. alboviridis, A. lanuginosus and A. movilensis being defined as separate species, A. inusitatus and A. alboluteus were also delimited as separate species (models 7 and 8). If A. alboviridis, A. lanuginosus and A. movilensis were lumped together, A. inusitatus and A. alboluteus were also lumped into one species (models 9 and 10). Models 11–14 focused on A. polyporicola and A. spelaeus. Models 11–12 left all populations of A. polyporicola free to be delimited and models 13–14 left all populations of both A. polyporicola and A. spelaeus unassigned. Models 11 and 14 defined A. alboviridis and A. lanuginosus as a part of A. movilensis and A. inusitatus as a part of A. alboluteus. This resulted in lumping of A. polyporicola and A. spelaeus into one species. In models 12–13, A. alboviridis, A. lanuginosus, A. movilensis, A. inusitatus and A. alboluteus were all defined as separate species. This definition resulted in delimitation of A. polyporicola and A. spelaeus as two separate species.

Species validation in the section Flavipedes

Finally, we further validated the putative species in DELINEATE based on the combined dataset of both series using six different models (Fig. 5). The first model left unassigned A. alboviridis, A. lanuginosus, A. inusitatus, population 1 of A. alboluteus, A. polyporicola and “A. capensis” (strains CBS 138188, CanS 34A, and CCF 5786). The second model used the same assignment except for A. polyporicola. The results of both models lumped A. capensis together with A. iizukae; A. polyporicola was delimited as separate species; A. inusitatus was lumped with A. alboluteus; and A. alboviridis and A. lanuginosus were lumped with A. movilensis. The third model left unassigned A. alboviridis, A. lanuginosus, A. inusitatus and also all populations of A. movilensis and A. alboluteus. The results lumped together A. alboviridis/A. lanuginosus/A. movilensis and A. inusitatus/A. alboluteus. The fourth model left unassigned A. inusitatus and populations of A. alboluteus with A. alboviridis, A. lanuginosus and A. movilensis being defined as three separate species and it resulted in A. inusitatus/A. alboluteus lumped together. On the contrary, the fifth model left A. alboviridis and A. lanuginosus unassigned with A. inusitatus, A. alboluteus and A. movilensis predefined as separate species, and it resulted in delimitation of separate A. alboviridis and A. lanuginosus. The sixth model predefined separately all mentioned species except for populations of "A. capensis" and A. iizukae which were left free to be delimited. This model again resulted in lumping of these two species.

Fig. 5

The results of species validation using DELINEATE for the series Flavipedes and Spelaei analyzed together. The populations of each species were delimited by BPP (Supplementary Table S1) and the displayed tree was calculated in starBEAST. The bars on the right side of the tree depicts the setting and results of each model. The grey bars represent the predefined species (locked in the analysis), while the brown bars represent unassigned populations left free to be delimited. The red rectangles depict the resulting solution proposed by DELINEATE.

Phylogeny

The Fig. 6 shows the best scoring Maximum likelihood (ML) tree based on the concatenated alignment of 90 strains from section Flavipedes representing known species diversity. All deep nodes gained bootstrap support of at least 95 % except for the lineage containing A. templicola and A. urmiensis. This node with support of 84 % is also the only site of incongruence between the concatenated tree and the tree generated by starBEAST during DELINEATE analysis. In the starBEAST analysis, A. iizukae formed a clade with A. suttoniae, A. templicola, and A. urmiensis (Fig. 5), while in the ML tree, A. iizukae was sister to the clade containing A. flavipes and A. ardalensis. Otherwise the topology of these two trees was identical. The only other node with lower support was the lineage comprising two singleton species, A. alboviridis and A. lanuginosus, with the bootstrap support of 90 %.

Fig. 6

Phylogenetic relationships of the section Flavipedes members inferred by Maximum Likelihood analysis in IQ-TREE v. 2.0 using concatenated alignment of four loci ITS, (benA, CaM, RPB2). The bootstrap support values are appended to nodes (only those supported by bootstrap value of 70 % or higher are displayed) with asterisks indicating the full support. Ex-type isolates are designated by bold font and superscript T. The names of species treated as synonyms are listed in parentheses. The source and locality of isolation are indicated by coloured circles and squares, respectively.

Apart from A. capensis (ex-type strain CBS 138188), which is discussed above, several other species names are treated here as synonyms in agreement with previous studies. The ex-type strain of A. archiflavipes NRRL 4852 is a synonym of A. flavipes; A. mangaliensis with the ex-type strain NRRL 62825 is a synonym of A. templicola; A. frequens with the ex-type strain NRRL 4578 is a synonym of A. micronesiensis; and finally, an invalid name A. sunderbanii based on the strain IMI 357699 is also included in the lineage of A. micronesiensis (Fig. 6).

Before the phylogenetic analysis itself, we tested the phylogenetic position of two relatively distant species, A. olivimuriae and A. neoniveus. Aspergillus olivimuriae was conclusively placed into the section Flavipedes, and therefore we included it as an outgroup in the phylogenetic analysis, but it was excluded from the species delimitation analyses. On the other hand, the position of A. neoniveus within the section Flavipedes did not gain sufficient support and its classification within Aspergillus sections remains uncertain. These findings are in agreement with Houbraken , who proposed series Olivimuriarum (containing only A. olivimuriae) and series Neonivei (containing only A. neoniveus), both within sections Flavipedes. In the phylogeny based on the three-gene dataset, the series Neonivei made the section Flavipedes paraphyletic with respect to section Terrei, while in the phylogeny based on nine genes, A. neoniveus was resolved within section Flavipedes (Houbraken ).

Phenotype analysis in relatives of A. movilensis

Selected culture and micromorphological characteristics relevant for species identification in section Flavipedes are summarized in Table 3. In the following paragraphs, we mostly focus on species related to A. movilensis because there was no clear consensus on species boundaries across molecular species delimitation methods used in this study.

Table 3

Overview of selected phenotypic characters for section Flavipedes members.1

Species	Growth parameters after 7 days					Prevailing colony colours on CYA and MEA	Conidia: diam (μm)4, surface	Vesicle diam (μm)	Stipe (μm)		Hülle cells2	Sex. state
	CYA	CZA	MEA	37 °C	40 °C				Length	Width
A. alboluteus	15–27	10–20	13–26	+	—	white, yellow	2.5–3.5, smooth	(6–)9–17(–20)	(500–)800–1200(–2000)	4–7(–10)	+	—
A. alboviridis	19–21	9–10	18–23	+3	—	light green, white, yellow	2.5–3.5, smooth	12–16(–20)	120–200(–550)	4.5–6(–8)	+	—
A. ardalensis	25–28	20–28	25–28	+	+	pale ochreous, yellow	2.5–3, smooth	(5–)7–19	commonly >1000	3–8(–9)	+	—
A. flavipes	25–35	25–28	25–35	+	—	white, yellowish-white, pale ochreous	2–3, smooth	7–11	occasionally >1000	3–6	—	—
A. iizukae	16–35	13–25	13–30	+	—	yellowish-white, yellow, ochreous, brown, greyish-brown	2–3(–3.5), smooth	(6–)14–20(–35)	200–1500	(3.5–)5–10(–13.5)	+	—
A. inusitatus	22–23	8–9	17–19	+	+3	dark green, yellow	3.5–4, echinulate	15–18	250–600	5–6	+	—
A. lanuginosus	26–27	21–22	20–21	+	—	white, light pinkish-brown, pale ochreous	2.5–3, smooth	10–12	600–1100	3.5–4.5	—	—
A. luppiae	18–20	17–21	20–22	+	—	yellow, white	2.5–3.5, smooth	11–16	100–220(–300)	3.5–5.5	+	—
A. micronesiensis	14–30	9–25	18–28	+	—	yellowish-white, yellow, pale ochreous, brown, grayish-brown	(2–)2.5–3.5(–4), smooth	(4–)6–16(–31)	250–1900	2–10	+	—
A. movilensis	22–25	19–20	25–30	+	—	white, pale ochreous, light yellow-green	2.5–3.3(–3.5), smooth	(5–)9–13(–16)	usually <400, occasionally >1000	3.5–6	+	—
A. neoflavipes	17–21	18–20	18–22	+	+3	yellow, white	(2–)2.5–3, smooth	13–19	250–950	5–7.5	+	+
A. neoniveus	15–16	12–14	13	—	—	yellow, white	2–2.5, smooth	9–11	150–300	4.5–6	+	+
A. olivimuriae	31–35	23–28	24–27	+	—	ochreous	2–2.5, smooth	8–10(–15)	100–150	5–6	—	—
A. polyporicola	17–27	14–20	20–32	—	—	pale ochreous, ochreous	2–3(–3.5), smooth	(6–)8–16(–20)	250–1000	3.5–6(–9)	—	—
A. spelaeus	16–30	6–26	15–32	—	—	pale ochreous, ochreous	2–3(–3.5), smooth	(5–)7–18(–23)	400–1000, occasionally >1000	3–7(–9)	+	—
A. suttoniae	24–25	20–22	24-25	+	—	yellowish-white, ochreous	2–3.5, smooth	(6–)12–17	180–420	4.5–6.5	—	—
A. templicola	21–32	21–28	23–30	+	+3	white, yellowish-white, pale ochreous	2–3, smooth	(6–)9–23	120–1400	3.5–10	+	—
A. urmiensis	28–32	20–24	23–27	+	—	white, ochreous	2–3, smooth	(17–)20–23(−30)	(350–)700–850(−1330)	(5–)8–10(−12)	—	—

"—" indicate no growth or absence of character/structure in culture.

Based on data from this study, Visagie , Hubka , Arzanlou , Siqueira , Crognale .

Production may vary between isolates and depend on cultivation conditions.

Very restricted growth (≤ 2 mm).

Conidia of all species are globose or subglobose, and only the longer dimension is given in case of subglobose conidia.

The colony colours of species from section Flavipedes are usually yellow, white, or in shades of brown. Two species newly described in this study are different in this regard. The colour of A. alboviridis colonies on some media is light green and A. inusitatus produces dark green colonies with yellow clumps of Hülle cells on all tested media. Aspergillus inusitatus is the only species with echinulate conidia in the whole section Flavipedes. In the study of Hubka , the production of accessory conidia has been observed in all section Flavipedes species. We expected to find them also in the newly described species, but they were only rarely present in some A. alboluteus strains and not observed in the other species. The sexual state has been only observed in A. neoflavipes and A. neoniveus, and it is produced in a homothallic manner. All other species from the section including all newly described species are presumably heterothallic and do not produce the sexual morph in culture under conditions used in this study.

To support species hypotheses and proposal of new taxa related to A. movilensis, we determined cardinal temperatures, compared the macromorphology on five media and the micromorphology of particular strains. Cardinal temperatures were determined on MEA after 14 d of cultivation in darkness at eight different temperatures ranging from 10 °C to 40 °C. The resulting colonies are compared in Fig. 7, which demonstrates phylogenetic pattern in growth rates and abilities to grow at different temperatures. Aspergillus lanuginosus grows faster than any other species at 20, 25 and 30 °C. Aspergillus alboluteus and A. insitatus grow faster at 37 °C than the other three species and unlike the other three species they do not grow at 10 °C (or grow very restrictedly - isolate CCF 6201). Aspergillus inusitatus is the only species capable of growing at 40 °C. The colony texture of A. lanuginosus is cottony or downy on the majority of media due to the production of abundant aerial mycelium, while the colonies of the other species are rather floccose.

Fig. 7

Temperature growth profile in the newly described species and Aspergillus movilensis after 14 d on MEA at temperatures ranging from 10 °C to 40 °C.

We observed morphological variation between the examined A. movilensis strains (Fig. 7). In general, white or grey colonies are produced, but the colonies of CBS 139562 are light green to yellow. We observed that colonies of some species change colour at suboptimal temperatures because of decreased sporulation (Fig. 7). There are differences between species in the ability to produce Hülle cells, which we observed in A. alboluteus, A. alboviridis, A. inusitatus, and A. movilensis, but they were not produced by A. lanuginosus. The length of stipes can be considered taxonomically important and may be used to distinguish A. alboviridis (mostly < 200 μm), A. movilensis (mostly < 400 μm), and A. lanuginosus (usually 600–1100 μm). Otherwise, the micromorphological characters were rather overlapping between the above-mentioned species.

Ecology

Based on the number of strains included in this study, A. micronesiensis, A. iizukae, A. spelaeus and A. templicola seem to be the most commonly encountered species. This is also in agreement with number of benA and CaM sequences deposited in GenBank for these species. In contrast to ITS region, the benA and CaM records for these species can be easily identified thanks to barcoding gap visible during BLAST analysis. The recorded numbers for benA /CaM loci in GenBank are as follows (accessed on June 10 2021) : A. micronesiensis 37 / 41, A. iizukae 24 / 23, A. spelaeus 18 / 18, and A. templicola 18 / 11.

The most common and diverse habitat for section Flavipedes members is undoubtedly the soil where 12 out of 17 species in our set of strains (excluding A. neoniveus) were found (Fig. 6). Our dataset contained five strains isolated from the hospital environment and six strains originating from clinical material. Only one of these strains belonged to the series Spelaei, specifically to A. spelaeus. The remaining strains were spread throughout the series Flavipedes, belonging to A. iizukae, A. suttoniae, A. templicola and A. micronesiensis. A significant number of species from both series originated from the indoor environment, namely A. iizukae, A. templicola, A. micronesiensis, A. polyporicola, A. spelaeus and A. alboluteus. In total three species, A. iizukae, A. spelaeus and A. alboviridis, were isolated from herbivore dung. Strains from food and feed were poorly represented in our dataset and restricted to A. iizukae, A. micronesiensis and A. spelaeus. None of the species which were represented by a high number of strains seem to be substrate specific.

The minimum inhibitory concentration (MIC) ranges and geometric mean (GM) values obtained by the EUCAST reference method for six antifungal agents are shown in Table 4 and more detailed results in Supplementary Table S2. Clinical breakpoints have been established for amphotericin B and the mould active azoles against A. fumigatus, which is the most common Aspergillus species in human infections. Clinical breakpoints have not been established for the Aspergillus section Flavipedes members. A general rule of thumb is that species that rarely cause disease in humans are less pathogenic, and therefore that adopting the breakpoints from the most common species in a genus for the rarer ones is clinically safe. With the caveat that some species were only represented with few strains, azole and amphotericin B MICs above the wild-type range for A. fumigatus were observed for A. ardalensis (all azoles and amphotericin B), A. micronesiensis (posaconazole and isavuconazole), A. polyporicola (posaconazole, voriconazole and isavuconazole), A. spelaeus (posaconazole, voriconazole, isavuconazole and amphotericin B), A. iizukae (isavuconazole and amphotericin B), and A. flavipes and A. urmiensis (amphotericin B) questioning the appropriateness of these drug bug combinations in clinical practice. In contrast, A. lanuginosus, A. luppiae, A. movilensis, A. neoflavipes, A. olivimuriae, and A. suttoniae were highly susceptible to amphotericin B and azoles. The MICs for the remaining species were comparable to those of A. fumigatus suggesting that these species are appropriate targets for the amphotericin B and azoles. Finally, terbinafine was active against all species except A. alboviridis.

Table 4

Antifungal susceptibility profiles of Aspergillus section Flavipedes members determined with EUCAST E.Def.9.3 method at 37 °C.1

Taxonomy

F. Sklenar, Jurjević, Ezekiel, Houbraken & Hubka, MycoBank MB 839382. Fig. 8.

Fig. 8

Macromorphology and micromorphology of Aspergillus alboluteus. A–E. Colonies after 14 d at 25 °C, left to right: CYA, CZA, MEA, OA and CY20S. F. Detail of colony edge on MEA. G. Detail of conidial heads and Hülle cells on MEA. H. Conidia. I. Conidia in air bubble. J, K. Accessory conidia. L, M. Hülle cells. N–P. Conidiophores. Scale bars: H–P = 10 μm.

Etymology: Named after white (sporulation) and yellow (clusters of Hülle cells) colours of the colonies on most media.

Typus: USA, Pennsylvania, Philadelphia, outdoor air, 2014, isolated by Ž. Jurjević (holotype PRM 952200, isotype PRM 952201, culture ex-type CBS 145855 = EMSL 2420 = CCF 5695 = IFM 66815).

Colony diam, 25 °C, 7 d (mm): CYA: 15–27; CZA: 10–20; MEA: 13–26; OA: 18–26; CY20S: 25–29.

Culture characteristics, 25 °C, 7 d: CYA: Colonies centrally raised; texture floccose; margin undulate to filiform; mycelial areas and sporulation white (#ffffff) to cream (#fffdd0) with icterine (#fcf75e) patches due to Hülle cell clumps; exudate absent; reverse centrally dark goldenrod (#b8860b) to light french beige (#c8ad7f), in margins dutch white (#f1ddb8). CZA: Colonies flat; texture floccose to granular; margin entire to delicately filiform; mycelial areas and sporulation white (#ffffff) with lemon yellow (#fff44f) patches due to Hülle cell clumps; exudate absent; reverse centrally flax (#eedc82), in margins naples yellow (#fada5e). MEA: Colonies slightly centrally raised; texture floccose to granular; margin entire to delicately filiform; mycelial areas and sporulation white (#ffffff) with lemon yellow (#fff44f) patches due to Hülle cell clumps; exudate absent; reverse centrally copper (#b87333) to saddle brown (#964b00), in margins bronze (#cd7f32). OA: Colonies flat to umbonate; texture granular; margin entire to delicately filiform; mycelial areas and sporulation white (#ffffff); exudate absent; reverse centrally khaki (#c3b091) to bone (#e3dac9) in margins. CY20S: Colonies flat to slightly centrally raised; texture floccose; margin slightly undulate to delicately filiform; mycelial areas linen (#faf0e6), sporulation white (#ffffff); exudate absent; reverse ecru (#c2b280).

Cardinal temperatures: Aspergillus alboluteus grows very restrictedly at 10 °C, and the optimum growth temperature is 30 °C. This species is able to grow at 37 °C but not at 40 °C (Fig. 7).

Micromorphology: Ascomata absent. Hülle cells present in strains CCF 5695 and CCF 4916 and absent in strains CCF 6201 and DTO 410-I8, elongated, branched, 20–30 μm long, forming yellow clumps. Conidial heads globose to radiate (remaining compact). Stipes smooth, brown (always hyaline under the vesicle), (500–)800–1 200(–2 000) × 4–7(–10) μm; vesicles hyaline, subglobose, (6–)9–17(–20) μm diam; metulae hyaline, cylindrical, 5–7 μm long, covering two thirds to entire surface of the vesicle; phialides hyaline, flask-shaped, 6–8 μm long. Conidia globose to subglobose, smooth, hyaline 2.5–3.5 (2.9 ± 0.1) × 2–2.5 (2.4 ± 0.1) μm. Accessory conidia absent or rare, globose to subglobose, on short, hyaline micro- to semimacronematous conidiophores.

Distinguishing characters: Aspergillus alboluteus is most closely related to A. inusitatus, but the latter is strikingly different from all related species by its green colonies and higher maximum growth temperature (40 °C). Phylogenetically, the next closest clade consists of A. alboviridis, A. lanuginosus and A. movilensis. Aspergillus alboluteus is phenotypically most similar to A. movilensis, that has similar colonies and also produces Hülle cells and accessory conidia with analogous morphology. However, these two species can be differentiated based on their conidiophore stipe lengths, vesicle diameters and colony sizes at 10 and 37 °C. The conidiophore stipes of A. movilensis rarely exceeds 400 μm, while stipes of A. alboluteus are (500–)800–1200(–2 000) μm long. The diameter of vesicles of A. movilensis rarely exceeds 13 μm, (5–)9–13(–16) μm, while vesicles of A. alboluteus are frequently larger, (6–)9–17(–20) μm. The colony diameters at 10 and 37 °C (on MEA, 14 d) slightly differ between these two species: at 10 °C, the colonies of A. movilensis attained 7 mm on average (the whole range was 6–8 mm), while colonies of A. alboluteus attained only 2 mm on average (1–2 mm); at 37 °C, A. alboluteus attained 18 mm on average (10–23 mm), while those of A. movilensis only 11 mm on average (6–19 mm).

J.P.Z. Siqueira, Gené, F. Sklenar & Hubka, MycoBank MB 821808. Fig. 9.

Fig. 9

Macromorphology and micromorphology of Aspergillus alboviridis. A–E. Colonies after 14 d at 25 °C, left to right: CYA, CZA, MEA, OA and CY20S. F, G. Detail of colony edge, conidial heads and Hülle cells on MEA. H. Conidia. I. Conidia in air bubble. J, K. Hülle cells. L–N. Conidiophores. Scale bars: H–N = 10 μm.

Etymology: Refers to the white and green colony colour.

Typus: Spain, Balearic Islands, Mallorca, Pollença, herbivore dung, 2016, isolated by J. Gené and J.P.Z. Siqueira (holotype CBS H-23128, isotype PRM 954607, culture ex-type CBS 142665 = FMR 15175 = CCF 6049 = IFM 66819).

Colony diam, 25 °C, 7 d (mm): CYA: 19–21; CZA: 9–10; MEA: 18–23; OA: 18–19; CY20S: 23–24.

Culture characteristics, 25 °C, 7 d: CYA: Colonies centrally raised; texture floccose; margin entire; sporulation centrally green sheen (#6eaea1) to white (#ffffff) in margins; large clear droplets of exudate on the colony surface; reverse centrally gold metallic (#d4af37) to flax (#eedc82) in margins. CZA: Colonies convex; texture floccose; margin slightly undulate; sporulation centrally pale spring bud (#ecebbd) to white (#ffffff) in margins with canary (#ffff9a) circle close to margin formed by Hülle cells; clear droplets of exudate on the colony surface; reverse centrally satin sheen gold (#ce9d41) to gold crayola (#e6be8a), in margins dutch white (#f1ddb8) to bone (#e3dac9). MEA: Colonies flat to umbonate; texture floccose; margin entire to filiform; sporulation centrally cambridge blue (#a3c1ad) to white (#ffffff) in margins with canary (#ffff9a) circle close to margin formed by Hülle cells; exudate absent; reverse centrally saddle brown (#964b00) to ochre (#cc7722), in margins earth yellow (#e1a95f). OA: Colonies slightly umbonate; texture floccose to granular, margin entire to delicately filiform; sporulation cenrally middle blue green (#8dd9cc) to white (#ffffff); clear droplets of exudate on the surface in the colony center; reverse centrally liver chestnut (#987456) to satin sheen gold (#ce9d41), in margins flax (#eedc82). CY20S: Colonies centrally raised, wrinkled; texture centrally velutinous to floccose in margins due to conidial heads; margin slightly undulate to filiform; sporulation centrally polished pine (#5da493) to white (#ffffff) in margins; exudate absent; reverse centrally antique bronze (#665d1e) to vegas gold (#c5b358), in margins dutch white (#f1ddb8).

Cardinal temperatures: Aspergillus alboviridis grows restrictedly at 10 °C, and the optimum growth temperature is around 25–30 °C. This species is able to grow very restrictedly at 37 °C (Fig. 7).

Micromorphology: Ascomata absent. Hülle cells elongated, frequently curved and branched, 20–30 μm long, forming yellow clumps. Conidial heads globose to compactly columnar. Stipes smooth, hyaline or brown (always hyaline under the vesicle), 120–200(–550) × 4.5–6(–8) μm; vesicles hyaline, globose to subglobose, 12–16(–20) μm diam; metulae hyaline, cylindrical, 6–7(–9) μm long, covering three quarters to entire surface of the vesicle; phialides hyaline, flask-shaped, 6–7.5(–8.5) μm long. Conidia globose to subglobose, smooth, hyaline 2.5–3.5 (2.9 ± 0.1) × 2–3 (2.5 ± 0.1). Accessory conidia not observed.

Distinguishing characters: Aspergillus alboviridis is most closely related to A. lanuginosus and A. movilensis. The colony colour of A. alboviridis is on some media in shades of green, most prominently on CYA, but also on MEA, OA and CY20S. Green coloured sporulation is not observed in A. lanuginosus and A. movilensis. Aspergillus alboviridis produces Hülle cells unlike A. lanuginosus. In contrast to A. movilensis, production of accessory conidia was not observed in A. alboviridis.

F. Sklenar, C. Silva Pereira, Houbraken & Hubka, MycoBank MB 839383. Fig. 10.

Fig. 10

Macromorphology and micromorphology of Aspergillus inusitatus. A–E. Colonies after 14 d at 25 °C, left to right: CYA, CZA, MEA, OA and CY20S. F. Detail of colony edge on MEA. G. Detail of conidial heads and Hülle cells on MEA. H–J. Hülle cells. K. Conidia. L. Conidia in air bubble. M–O. Conidiophores. Scale bars: H–O = 10 μm.

Etymology: Name refers to the strikingly different colony morphology in comparison with other members of section Flavipedes.

Typus: Tunisia, Ras Rajel, soil in oak forest, 2009, isolated by C. Silva Pereira (holotype PRM 954606, culture ex-type DTO 121-G5 = CBS 147044 = CCF 6552).

Colony diam, 25 °C, 7 d (mm): CYA: 22–23; CZA: 8–9; MEA: 17–19; OA: 17–18; CY20S: 25–26.

Culture characteristics, 25 °C, 7 d: CYA: Colonies flat, densely covered with exudate droplets; texture floccose; margin delicately undulate or delicately filiform; sporulation bottle green (#006a4e), in margins maximum blue green (#30bfbf) to white (#ffffff); large clear or golden brown (#996515) droplets of exudate on the entire surface of the colony, small clear droplets in margins; reverse centrally satin sheen gold (#cba135), in margins flax (#eedc82). CZA: Colonies umbonate with raised edge; texture floccose, cottony in central area; margin delicately undulate to filiform; sporulation middle blue green (#8dd9cc) to cadmium green (#006b3c) or white (#ffffff), canary (#ffff9a) ring at the colony edge formed by clumps of Hülle cells; exudate absent; reverse centrally olive green (#b5b35c), in margins vegas gold (#c5b358) to flax (#eedc82). MEA: Colonies centrally raised, richly permeated with clumps of Hülle cells; texture floccose; margin slightly undulate to delicately filiform; sporulation british racing green (#004225) to bottle green (#006a4e) to maximum blue green (#30bfbf) with clear boundaries between sectors, in margins middle blue green (#8dd9cc) to white (#ffffff), canary (#ffff9a) circle in the middle of the colony formed by Hülle cells; exudate absent; reverse centrally metallic sunburst (#9c7c38) to saddle brown (#964b00), in margins sage (#bcb88a). OA: Colonies centrally raised, densely covered with exudate droplets; texture floccose; margin entire to delicately filiform; sporulation cadmium green (#006b3c) to middle blue green (#8dd9cc), in margins deep jungle green (#004b49) to maximum blue green (#30bfbf) to white; clear droplets of exudate mainly in the colony center, scarcely on the edge; reverse centrally straw (#e4d96f) to artichoke (#8f9779), in margins sage (#bcb88a). CY20S: Colonies centrally raised; texture floccose with cottony patches; margin delicately undulate to delicately filiform; sporulation british racing green (#004225) to bottle green (#006a4e), in margins middle blue green (#8dd9cc) to white (#ffffff), canary (#ffff9a) circle close to the colony edge formed by Hülle cells; exudate absent; reverse centrally straw (#e4d96f) to vegas gold (#c5b358), in margins flax (#eedc82).

Cardinal temperatures: Aspergillus inusitatus grows at 15 °C but does not grow at 10 °C. The optimum growth temperature is 30–35 °C. This species is able to grow restrictedly at 40 °C but not at 42 °C (Fig. 7).

Micromorphology: Ascomata absent. Hülle cells elongated, frequently curved or branched, 20–30 μm long or subglobose to ovate, 9–12 × 8–10 μm, forming yellow clumps. Conidial heads compactly radiate. Stipes smooth, hyaline or dark brown (always hyaline under the vesicle), 250–600 × 5–6 μm; vesicles hyaline, subglobose to pyriform, 15–18 μm diam; metulae hyaline, cylindrical, 6.5–7.5 μm long, usually covering the entire surface of the vesicle; phialides hyaline, flask-shaped, 7.5–8.5 μm long. Conidia subglobose, echinulate, fern green (#4f7942), 3.5–4 (3.6 ± 0.2) × 3–3.5 (3.1 ± 0.2) μm. Accessory conidia not observed.

Distinguishing characters: Aspergillus inusitatus is most closely related to A. alboluteus. Phylogenetically, the next closest clade consists of A. alboviridis, A. lanuginosus and A. movilensis. Aspergillus inusitatus differs strikingly from all these species by its dark green colony colour. The optimum growth temperature of A. alboluteus, A. alboviridis, A. lanuginosus and A. movilensis is around 25–30 °C, while the optimum growth temperature of A. inusitatus is 30–35 °C. Unlike the four above-mentioned species, A. inusitatus is able to grow at 40 °C. Furthermore, unlike all other species in the section Flavipedes, A. inusitatus produces echinulate conidia.

F. Sklenar & Hubka, MycoBank MB 839384. Fig. 11.

Fig. 11

Macromorphology and micromorphology of Aspergillus lanuginosus. A–E. Colonies after 14 d at 25 °C, left to right: CYA, CZA, MEA, OA and CY20S. F. Detail of colony edge on MEA. G. Detail of conidial heads on MEA. H. Conidia. I. Conidia in air bubble. J–M. Conidiophores. Scale bars: H–M = 10 μm.

Etymology: Refers to the relatively rich production of aerial mycelium making the colonies downy on some media.

Typus: Haiti, Fonds Parisien, soil, unknown year of collection, unknown isolator (holotype PRM 954608, isotype PRM 954609, culture ex-type NRRL 4610 = IMI 350352 = CCF 4551 = IFM 66818).

Colony diam, 25 °C, 7 d (mm): CYA: 26–27; CZA: 21–22; MEA: 20–21; OA: 25–26; CY20S: 28–29.

Culture characteristics, 25 °C, 7 d: CYA: Colonies centrally raised; texture floccose; margin delicately filiform; sporulation centrally pale spring bud (#ecebbd) to beige (#f5f5dc) to champagne pink (#f1ddcf) in margins; clear droplets of exudate on the surface in the colony center; reverse centrally orange peel (ff9f00) to dutch white (f1ddb8) to gold crayola (e6be8a), in margins linen (#faf0e6). CZA: Colonies centrally raised; texture floccose; margin delicately filiform; sporulation centrally champagne pink (#f1ddcf), white (#ffffff) in margins; large amount of clear droplets of exudate on the surface in the colony center; reverse centrally camel (#c19a6b) to wheat (#f5deb3), in margins cream (#fffdd0). MEA: Colonies umbonate; texture cottony to floccose; margin delicately filiform; sporulation centrally champagne pink (#f1ddcf) to white (#ffffff) in margins; exudate absent; reverse centrally golden brown (#996515) to dark goldenrod (#b8860b), in margins maize crayola (#f2c649). OA: Colonies umbonate; texture floccose, cottony in the center; margin delicately filiform; sporulation centrally champagne pink (#f1ddcf) to white (#ffffff) in margins; exudate absent; reverse centrally antique bronze (#665d1e) to bistre brown (#967117), in margins flax (#eedc82). CY20S: Colonies flat to umbonate; texture floccose with small cottony patches in central areas; margin delicately filiform; sporulation centrally opal (#aac4c4) to linen (#faf0e6), in margins white (#ffffff); exudate absent; reverse centrally vegas gold (#c5b358) to bistre brown (#967117), in margins medium champagne (#f3e5ab).

Cardinal temperatures: Aspergillus lanuginosus grows at 10 °C, and the optimum growth temperature is 30 °C. This species is able to grow restrictedly at 37 °C but not at 40 °C (Fig. 7).

Micromorphology: Ascomata absent. Hülle cells absent. Conidial heads globose to radiate (remaining compact). Stipes smooth, hyaline or light brown (always hyaline under the vesicle), 600–1 100 × 3.5–4.5 μm; vesicles hyaline, subglobose to pyriform, 10–12 μm diam; metulae hyaline, cylindrical, 5–5.5 μm long, covering two thirds of the vesicle; phialides hyaline, flask-shaped, 5.5–6.5 μm long. Conidia globose to subglobose, smooth, hyaline 2.5–3 (2.7 ± 0.1) × 2–2.5 (2.3 ± 0.1) μm. Accessory conidia not observed.

Distinguishing characters: Aspergillus lanuginosus is most closely related to A. alboviridis and A. movilensis. In contrast to these two species, A. lanuginosus does not produce Hülle cells, its stipes are longer (A. lanuginosus 600–1 100 μm, A. alboviridis usually <200 μm, A. movilensis usually <400 μm) and the colony texture of A. lanuginosus on some media (MEA, OA, CY20S) is cottony, at least in the colony center, unlike the other species with mostly floccose colony texture.

Notes: The ex-type strain of A. lanuginosus NRRL 4610 was treated as A. carneus (section Terrei) by Raper & Fennell (1965). The phenotypic differences between ex-type strains of A. movilensis and A. lanuginosus were also observed by Hubka who provisionally treated both strains as A. movilensis. These two strains (NRRL 4610 and NRRL 62819) also showed unique PCR fingerprinting pattern using the phage M13-core oligonucleotide primer and primer 834t (Hubka ).

Sugiy., J. Fac. Sci. Univ. Tokyo, Sect. 3, Bot. 9: 390. 1967. MycoBank MB 326636.

Synonym: Aspergillus capensis Visagie et al., Stud. Mycol. 78: 105. 2014. MycoBank MB 809193.

Discussion

Species delimitation

In this study, we followed up on previous studies dealing with species delimitation in the genus Aspergillus (Sklenář , 2020, Hubka , 2018b). The species delimitation step employing four single-locus and one multi-locus method was performed in the same way as in Sklenář . In contrast to our previous studies, here we presented the results of STACEY in the form of similarity matrices showing the assignments of individuals into species using several different values of the collapseheight parameter. This parameter is critical for the delimitation and its value is chosen arbitrarily. Our goal was to find the lowest value of the parameter which correctly delimits the species with indisputable species boundaries. These “reference species” differ from their close relatives clearly by phenotypic characters and/or were conclusively delimited by vast majority of single-locus delimitation methods. In the series Flavipedes, we considered A. ardalensis, A. flavipes, A. suttoniae, A. templicola, and A. urmiensis as these “reference species”. The results of species delimitation methods in series Spelaei were less stable and only A. luppiae and A. inusitatus were regarded as “reference species”. Then we checked how the results changed depending on the increase or decrease of the collapseheight parameter. For series Flavipedes, we concluded that a sensible value is 1 × 10-3. Using this value, all currently known species except A. capensis were supported, with A. templicola having some probability of splitting into two species (Fig. 3A). A higher value of the collapseheight parameter (5 × 10-3) resulted in excessive lumping, supporting only four species in the series (Supplementary Fig. S1A). When we used a lower value (1 × 10-4), A. templicola and A. micronesiensis were split into several species, but A. capensis still formed one species together with strains of A. iizukae (Supplementary Fig. S1B).

The same value of the collapseheight parameter (1 × 10-3) was applied on the series Spelaei. Using this value, STACEY supported all species related to A. movilensis, i.e., A. alboviridis, A. lanuginosus, A. movilensis, A. inusitatus and A. alboluteus (Fig. 3B). All species were also delimited when a slightly lower value (7.5 × 10-4) was used, but at the same time this value resulted in increased support for additional splitting of A. polyporicola and A. spelaeus (Supplementary Fig. S2B). On the other hand, slightly higher value (2 × 10-3) resulted in lumping of A. alboluteus with A. inusitatus, and lower support for delimitation of A. movilensis/A. alboviridis/A. lanuginosus, and also A. polyporicola/A. spelaeus as separate species (Supplementary Fig. S2A).

The four single-locus methods gave mostly similar results within the series Flavipedes. The incongruences in the results were present among strains of A. iizukae and A. capensis, which are morphologically indistinguishable and the latter was described solely based on phylogeny (Visagie ). Especially the results of single-locus delimitation methods based on the benA gene together with STACEY results indicate that recombination is present in this clade. This fact is also evident in the single gene phylogenies (Fig. 2) proving that recognition of A. capensis is not supported by genealogical concordance phylogenetic species recognition concept (GCPSR) (Taylor ).

The results of single-locus methods within series Spelaei were more variable. The clade consisting of A. spelaeus and A. polyporicola was delimited as one species by 5/12 methods, as two separate species by 5/12 methods, and as three species by 2/12 methods (Fig. 1B). Aspergillus movilensis clade consisting of A. alboviridis, A. lanuginosus, A. movilensis, A. inusitatus and A. alboluteus was delimited as one species by 1/12 methods, two species by 7/12 methods, five species by 3/12 methods, and six species 1/12 methods. In this ambiguous situation, our conclusions were significantly influenced by the presence of species-specific characters. Most importantly, A. inusitatus was conspicuously different compared to all other species within the section (Fig. 7, Table 3) and was also considered one of the “reference species” for STACEY method. In all cases where A. inusitatus was delimited as separate species, A. alboviridis and A. lanuginosus gained support.

The GMYC method is sometimes considered prone to over-splitting (Lohse 2009, Talavera ), but it did not over-split the datasets in our previous study on the section Nidulantes (Sklenář ), neither the dataset in this study. The method delimiting the highest number of species was bPTP and the most conservative method (resulting in the lowest number of delimited species) was bGMYC.

Species validation

For the species validation step, we employed the program DELINEATE, which is used for the first time to delimit species within the fungal kingdom. Until now, the most commonly used software for species validation was BPP (Bayesian Phylogenetics and Phylogeography) (Yang 2015). In the past studies on Aspergillus, this method was not very useful for distinguishing between different hypotheses on species boundaries, because its results are usually very benevolent, supporting even more splits than proposed by species delimitation methods. This excessive acceptance of splitting scenarios was also noted in other organisms and described by Sukumaran & Knowles (2017), who proposed the usage of the protracted speciation model by software for species validation. DELINEATE is the first program to challenge BPP as a choice for species validation.

In this study, BPP was only used as a tool for dividing the individuals from a given dataset into populations (Supplementary Table S1). The starBEAST was then used to calculate the “species tree” based on the populations. Finally, DELINEATE separated the populations from this tree into species. In contrast to BPP, at least some populations in DELINEATE need to be assigned to species (ideally those whose classification into species is clear) and others are left open to be delimited which offers opportunities to test a number of hypotheses. The limiting factor of the method are high computational demands that increase significantly with the number of unassigned populations. At the same time, a high total number of populations (unassigned and assigned into species) does not hinder the DELINEATE computation, but it can cause problems in starBEAST during species tree calculation.

Similarly to STACEY, where we applied several values of the collapseheight parameter, we generated several different models to be tested with DELINEATE. The results were again more complex in the series Spelaei, where we tested 14 models compared to only four models in series Flavipedes. The analysis of the model with all species in A. movilensis clade unassigned was not computationally feasible. Therefore, we had to generate separate models with different parts of the clade being unassigned (Fig. 4B, Fig. 5). When we left A. alboviridis, A. lanuginosus and A. inusitatus unassigned, as a result, they were lumped with their sister species. When A. inusitatus was defined as a separate species, A. alboviridis and A. lanuginosus were also separated from A. movilensis. Analogously, when A. alboviridis and A. lanuginosus were designated as separate species, A. inusitatus was also separated from A. alboluteus. Both parts of the analysis showed mixed support for the delimitation of A. inusitatus as separate species. But if A. inusitatus was designated as separate species, A. alboviridis and A. lanuginosus were consistently delimited as separate species too.

In the series Flavipedes, we mainly focused on the validation of species limits in the clade containing A. iizukae and A. capensis (Fig. 4A, Fig. 5). The results of all models were consistent. Aspergillus capensis was always lumped together with A. iizukae. Therefore, we consider A. capensis as a synonym of A. iizukae. Aspergillus templicola and A. micronesiensis were always delimited in the same way as suggested by majority of the species delimitation methods.

Furthermore, seven models of the whole section were generated to be tested in DELINEATE (Fig. 5). The results were congruent with the analyses performed separately in both series with one exception. The fourth model lumped A. inusitatus with A. alboluteus even though A. alboviridis and A. lanuginosus were specified as separate species. Based on our study, DELINEATE proved to give more relevant results which were more in agreement with putative species delimited in the first step of analysis compared to very benevolent BPP.

Similarly to our previous studies, we could not fully comply with the assumption of the species delimitation methods, i.e. not including species without variability (singleton species). It is a common issue in this kind of analyses and it has been discussed before (Ahrens ). Singleton species are frequently described within genus Aspergillus with the general idea of the endeavor to describe all discovered diversity. The possibility of the singleton species (A. alboviridis, A. lanuginosus and A. inusitatus) being part of other closely related species was tested, but the results supported the splitting scenarios.

In this study, we enriched the analyses by addition of new strains and thus variability, resulting in more robust and accurate knowledge of the species boundaries in comparison with the previous studies on the section Flavipedes. In general, the data supported currently known species and in one case demonstrated recombination leading to synonymization of A. capensis with A. iizukae. A similar case occurred for instance in A. parafelis and A. pseudofelis (section Fumigati) which were described based on several strains with very limited genetic variability (Sugui ) and later put into synonymy with A. felis when higher number of strains was available for analyses (Hubka ). The support for delimitation of A. polyporicola and A. spelaeus as separate species was also ambiguous (Fig. 1B). These two species were proposed by Hubka based on the differences in their ecology and massive production of accessory conidia by A. polyporicola in contrast to A. spelaeus. In this study, we obtained new strains demonstrating that the ecological factor is not relevant as a taxonomic character for A. polyporicola, which was considered by Hubka as a fungicolous species. We also examined the presence and abundance of accessory conidia in a broader set of A. spelaeus and A. polyporicola strains and we did not find any relevant differences between these two species – accessory conidia were rather rare or absent in all newly examined strains. Aspergillus polyporicola and A. spelaeus can be therefore considered cryptic species and there is a significant chance for their synonymization in the future when a larger dataset will be available in terms of the number of strains or loci. But in the present study, the splitting scenario gained overall higher support. There is also no conflict in the topology of single-gene phylogenies and both species are supported by GCPSR (data not shown).

Incorporating multispecies coalescent model-based methods into taxonomy of Aspergillus

The polyphasic approach is currently a standard for species delimitation in Aspergillus. It always involves the assessment of phylogenetic data, with the GCPSR most commonly claimed to be used as a method of choice for delineation of species boundaries. It is however seldomly used methodologically properly as originally described by Dettman . Using the methods employed in this study, the final decision regarding the species boundaries remains partly subjective as other components of the polyphasic approach are still included in the decision-making process. The species delimitation part based on molecular data is however largely free from subjectivity and the methods force the taxonomists to take a fresh and unbiased look at the species limits. Furthermore, these coalescent-based methods are at least to some degree able to accommodate phenomena such as incomplete lineage sorting, recombination, or non-reciprocal monophyly (Edwards 2009). We do not expect or encourage these methods to replace the polyphasic approach, but they should enhance and strengthen it.

Several studies including this study highlighted the fact that the intraspecific genetic variability in Aspergillus is probably higher than expected and that species range sizes are unequal across aspergilli (Geiser , Hubka ). Consequently, it can be expected that the species numbers in some extensively studied groups with a narrow species concept will be reduced in the future. The resulting taxonomy should become clearer, which is demanded by user community given the number of studies (from clinical, biotechnological, or industrial background) struggling with the identification of cryptic Aspergillus species in the various species complexes (Negri , Pantelides , D'hooge , Imbert , Mincuzzi ).

The ultimate future goal is to bring the majority of species in Aspergillus to the same level from the population genetics standpoint, which will certainly result in a delimitation of new species and synonymization of some others, with the synonymization being more likely in species groups which have been extensively studied and have a narrower species concept. An important prerequisite for these steps is a shift of taxonomy from a discovery phase, where species are described based on a small number of strains, to a phase where species limits are assessed based on large numbers of strains from different localities. It is already possible to obtain comprehensive datasets from databases for some species or species groups such as A. fumigatus, series Nigri, Versicolores, Flavi or others which can be re-analyzed using similar approaches.

Another important factor relevant for species concept in Aspergillus is the inevitable advent of species delimitation based on whole-genome sequences (Matute & Sepúlveda 2019). In this aspect, it is crucial not to exploit this vast source of genetic variability to flood the genus with cryptic species most likely corresponding to natural populations of a broader species, but to utilize it to learn more about the mechanisms of species boundaries formation and preservation, eventually leading to the establishment of meaningful rules for species delimitation (biologically meaningful species concept).

Ecology and clinical relevance

The ecology and clinical relevance of section Flavipedes members has been extensively reviewed by Hubka . The members of the section are distributed worldwide in soil, especially in tropical and subtropical regions (Klich 2002, Domsch , Choochuay ). Some representatives are osmotolerant and xerophilic and they are frequently isolated from arid or saline environments, e.g. sea sand (Lee ), saline soil (Kang ) or salterns (Chung ). There are also several reports on species adopting life strategy of plant endophytes (Luyen , Qin , El-Hawary ). Other studies report another type of symbiosis of these fungi, namely with marine animals, sea cucumbers (Nerva ) and marine sponges (Wiese ). The same authors, Nerva , also reported that A. spelaeus harbors mycovirus Aspergillus spelaeus polymycovirus 1 (AsPMV1). Another habitat where section Flavipedes members are present is the cave environment. Numerous A. spelaeus strains used in this study originated from cave air or sediment and there are also reports of A. iizukae and A. movilensis isolated from caves (Hubka , Nováková ). Similarly to many other Aspergilli, species from section Flavipedes spoil various food and feed (Pitt & Hocking 2009) and are frequently isolated from the indoor environment (Visagie , Sánchez Espinosa , this study). Even though hospitals can be also considered indoor environments, we highlighted these strains separately in Fig. 6, because some species have the potential to be pathogenic and hospital environment monitoring has usually a different purpose than monitoring indoor fungi in other buildings.

Some section Flavipedes members are uncommon human and animal pathogens. The majority of well documented cases of aspergillosis were attributed to A. flavipes and comprise diverse clinical manifestation including cutaneous aspergillosis (Barson & Ruymann 1986), onychomycosis (Gehlot ), otomycosis (Stuart & Blank 1955), osteomyelitis (Roselle & Baird 1979), diskospondylitis (Schultz ), pulmonary aspergillosis (Katou ) and cerebral aspergillosis (Masih ). However, the species identification of these strains according to the current taxonomy is mostly unclear except for the isolate associated with onychomycosis (GenBank EU515154) which represents A. micronesiensis. Even though there are sequences available for the strain VPCI 631/P/15 (GenBank KX455808, KX455766), a cause of cerebral aspergillosis (Masih ), their information is contradictory. The benA sequences represent A. flavipes s. str., while the CaM sequence represents A. micronesiensis. These two species are phylogenetically distant (Fig. 6) and one of these sequences was therefore most likely incorrectly deposited. Apart from mentioned cases, A. micronesiensis was also reported by Siqueira from canine urine sample. The same authors also reported A. iizukae and A. suttoniae from human sputum and bronchoalveolar lavage, respectively, and A. spelaeus from human forearm. However, detailed information about clinical relevance of these isolates was not included. In this study, we examined two strains (IHEM 22505 and IHEM 22506) originally identified as A. flavipes from human sputum and lung, respectively, which were reidentified as A. micronesiensis. Both strains were isolated from patients in Belgium and preserved in BCCM/IHEM collection. Another clinical strain examined here was obtained from bronchoalveolar lavage of a Danish patient and was identified as A. templicola. The available anamnestic data for these cases are limited and do not allow confirmation of the etiological relevance of strains.

The antifungal susceptibility patterns in section Flavipedes are largely unknown and restricted to few clinical isolates without clear species identification (A. flavipes s. l.) and examined by various methods (Del Carmen Serrano , Martin-Mazuelos , Masih ). In this study, we evaluated antifungal susceptibilities to six antifungal agents in a large set of reliably identified strains across section Flavipedes. Our results showed notable differences in susceptibility pattern between the species as the MIC distributions for itraconazole, posaconazole, and isavuconazole spanned six two-fold dilutions and those for amphotericin B, voriconazole and terbinafine spanned seven two-fold dilutions across the 67 strains. In general, species-specific MIC distributions for wild type strains span three two-fold dilutions when tested in a single laboratory with a single batch of plates. Therefore, the difference in susceptibility among the species included herein cannot be explained by inherent variation of the test but must reflect intrinsic differences in susceptibility among the members of the Aspergillus section Flavipedes. About half of the species were susceptible to all agents whereas the remaining species were resistant to one or several agents. Therefore, species identification and susceptibility testing is important in clinical infections.

Conclusions

The revision of section Flavipedes led to an increase of its known variability. We performed phylogenetic and phenotypic analyses utilizing strains from previous studies and newly isolated strains. As a result, four species were newly described and one species, A. capensis, was put in synonymy. So far, the species from the series have been known to produce white, yellow, and brown colonies. The newly described species extend this heterogeneity, namely, the A. inusitatus deviates significantly by its dark green colonies and also colonies of A. alboviridis are light green on some media. The species delimitation analysis performed here builds upon recent studies on the genus Aspergillus and expands the spectrum of methods in the species validation step. Antifungal susceptibility across species diversity of the section Flavipedes showed elevated MICs to amphotericin B or azole derivates in significant part of the tested species, especially in those that are potentially clinically relevant. This finding emphasizes the need for species identification and susceptibility testing in clinical strains.