Introduction

The genus Acroconidiella was proposed by Lindquist and Alippi (1964) to accommodate fungi with macronematous, mononematous, simple, or occasionally branched conidiophores with integrated, terminal, polytretic sympodial conidiogenous cells bearing solitary ellipsoidal, septate echinulate conidia (Ellis 1971). Baker (1947) made the first report of a disease in Tropaeolum majus (garden nasturtium—chagas, in Brazil) in California and considered that it was caused by a fungus belonging to Heterosporium, but without giving it a name. In the same year, Bond (1947) independently described the fungus on Tropaeolum majus in Ceylon (Sri Lanka) naming it Heterosporium tropaeoli. De Vries (1952), considered that Heterosporium should be regarded as analogous to Cladosporium. However, the fungus described as Heterosporium tropaeoli has morphological features which were recognized as clearly distinct from members of Cladosporium. The fungus on garden nasturtium has porospores instead of blastospores—typical of Cladosporium—and a scar morphology which is also distinct from those of Cladosporium. Lindquist mentioned a personal communication made to him by M. B. Ellis (Lindquist and Alippi 1964) “H. tropaeoli is closer to Alternaria, than to Curvularia and Cladosporium”. However, this fungus did not present muriform conidia in chains, thought to be diagnostic for Alternaria, at the time, and the conidia did not have its central cells larger than the apical cells, accepted as key for Curvularia at the time [now much changed after the works of Berbee et al. 1999 and Manamgoda et al. 2012]. Based on Ellis’ views, Lindquist and Alippi (1964) proposed a new genus Acroconidiella to accommodate this fungus and designated it Acroconidiella tropaeoli.

Presently, there are five species in the genus Acroconidiella: A. eschscholtziae, A. indicus, A. manoharacharii, A. tropaeoli, and A. trisepta (Lindquist and Alippi 1964; Ellis 1971, 1976; Muchovej 1980; Prasher and Verma 2015).

The lack of molecular studies for any of the species in this genus and the observations made by Vieira and Barreto (2002), particularly the in vitro production of short acropetal conidial chains, prompted a reappraisal of Acroconidiella. This was based on freshly collected material of two members of Acroconidiella, namely Acroconidiella tropaeoli and Acroconidiella trisepta.

Materials and methods

Sample collection processing and observation of fungus morphology

Samples of diseased foliage of Tropaeolum majus and dead branches of Glycine max were collected, the former from the original host of Acroconidiella tropaeoli, but the latter both from the substrate plant species and exactly from the type locality (Viçosa, state of Minas Gerais, Brazil). These were screened under a stereomicroscope, and parts of the samples bearing sporulating colonies of the fungi were selected and dried in a plant press. Fungal structures were scraped from the sample surface with a scalpel and mounted in lactophenol and lactofuchsin. Observations were made with an Olympus BX53 adapted with differential contrast lighting and equipped with digital capture system (Olympus Q-Color 3™). Representative specimens were deposited in the local herbarium (Herbarium Universidade Federal de Viçosa, VIC).

Isolations were performed by aseptic transfer of conidia from the leaf surfaces onto potato dextrose-agar (PDA) plates with a sterile fine-pointed needle. Culture descriptions were based on the observation of 7-day-old (A. tropaeoli) and 14-day-old (A. trisepta) colonies formed in plates containing either potato dextrose-agar (PDA) or potato carrot-agar (PCA), maintained at 25 °C under a 12-h daily/light regime (light provided by two white and one near-UV lamps placed 35 cm above the plates). The color terminology followed Rayner (1970).

Samples of dried material containing fungal structures were mounted on stubs with double-sided adhesive tape and gold-coated using a Balzer’s FDU 010 sputter coater. A Carl-Zeiss Model LEO VP 1430 scanning electron microscope (SEM) was used to analyze and generate images from the samples.

DNA isolation

Total genomic DNA was extracted from 7-day-old cultures formed on PDA by using Wizard® Genomic DNA Purification Kit (Promega Corporation, WI, USA) following the manufacturer’s instructions and the steps described by Pinho et al. (2012).

PCR amplification

The large subunit of the nrDNA (nc LSU rDNA) and internal transcribed spacer (ITS) regions from each fungus included in the study were sequenced with the primers LSU1Fd (Crous et al. 2009) and LR5 (Vilgalys and Hester 1990) and IT5 + ITS4 (White et al. 1990), respectively. For Acroconidiella tropaeoli, two additional loci, polymerase II second largest subunit (RPB2) and translation elongation factor 1-α (TEF1), were amplified and sequenced with the primer pairs RPB2-5F2 (Sung et al. 2007) and fRPB2-7R (Liu et al. 1999) and EF1-728F + EF1-986R (Carbone and Kohn 1999). PCR amplifications were performed in a total volume of 12.5 μL containing 10–20 ng of template DNA, 1× PCR buffer, 0.63 μL DMSO (99.9%), 1.5 mM MgCl2, 0.5 μM of each primer, 0.25 mM of each dNTP, and 1.0 U BioTaq DNA polymerase (Bioline GmbH Luckenwalde, Germany). Conditions for PCR amplification consisted of an initial denaturation step of 5 min at 94 °C followed by 35 cycles of 30 s at 94 °C, 30 s at 48 °C, and 90 s at 72 °C for nc LSU rDNA, ITS and 40 cycles of 30 s at 94 °C, 30s at 52 °C/59 °C and 45 s at 72 °C for TEF1, and a final elongation step of 7 min at 72 °C. The partial RPB2 gene was obtained by using a touchdown PCR protocol of 5 cycles of 45 s at 94 °C, 45 s at 60 °C, and 2 min at 72 °C, followed by 5 cycles with a 58 °C annealing temperature and 30 cycles with a 54 °C annealing temperature. Amplicons were analyzed on 0.8% agarose electrophoresis gels stained with GelRed (InstantAgarose) in a 1× TAE buffer and visualized under UV light to check for amplification size and purity. PCR products were purified and sequenced by Macrogen Inc. (http://www.macrogen.com).

Phylogenetic analysis

DNA consensus sequences were generated and imported into MEGA v. 6 (Tamura et al. 2013) for initial alignment and the construction of sequence datasets. Sequences obtained from the GenBank (www.ncbi.nlm. nih.gov) and the novel sequences generated on this study were aligned and edited using DNA Dragon program (http://www.dna-dragon.com\index.php) (Table 1).

Table 1 Taxa and collections used for multi-gene phylogenetic analyses in this study
Full size table

Bayesian inference analyses were conducted, and the best-fit evolutionary model was determined by comparing different evolutionary models via the Akaike information criterion using PAUP (version 4.0b10, Sinauer Associates) and MrModeltest 2.2 (Nylander 2004). Posterior probabilities were determined by Markov chain Monte Carlo sampling (MCMC) in MrBayes v. 3.2.1 (Ronquist et al. 2012). Six simultaneous Markov chains were run for 10,000,000 generations, and trees were sampled every 100th generation and 10,000 trees were obtained. The first 2000 trees representing the burn-in phase were discarded, whereas the remaining 8000 trees were used for calculating the posterior probabilities. Bayesian posterior probabilities are presented on the left of each node. The analysis was hosted by CIPRES Science Gateway portal at San Diego Supercomputer Center (Miller et al. 2010). Phylogenetic trees were visualized with the program FigTree v1.3.1 (Rambaut 2009).

For maximum parsimony (MP), analyses were conducted using PAUP v. 4.0b10 (Swofford 2003) with phylogenic relationships estimated by heuristic searches with 1000 random stepwise addition sequences and tree bisection and reconstruction (TBR) branch swapping. Alignment gaps were treated as missing data, and all characters were weighted equally. Measures calculated for parsimony included tree length (TL), retention index (RI), consistency index (CI), rescaled consistency index (RC), and homoplasy index (HI). Statistical support for branch nodes in the most parsimonious trees was obtained by performing 1000 bootstrap replicates.

Maximum likelihood (ML) tree was generated with the nearest-neighbor-interchange (NNI) ML heuristic method and the Tamura-Nei substitution model as tree inference options in MEGA. The branch stabilities of the phylogenetic tree were assessed by using the bootstrap re-sampling strategy with 1000 bootstrap test replicates. The resulting tree topologies using the three methods (MP, ML, and BI) were then compared, and the phylogram was edited with CorelDRAW Graphics Suite 2017.

Sequences of Stemphylium herbarum (CBS 191.86 and BRIP 65181) and Stemphylium botryosum (CBS 714.68) were used as the outgroups in the Alternaria phylogeny, and sequences of Paradendryphiella arenariae (CBS 181.58) and Paradendryphiella salina (CBS 142.60) were used in the Dendryphiella phylogeny. Sequences derived in this study were lodged in GenBank (http://www.ncbi.nlm.nih.gov/genbank) (Table 1). The alignment and tree were deposited in TreeBASE (http://www.treebase.org) (study numbers S24915 and S24916) and taxonomic novelties in MycoBank (www.MycoBank.org).

Results

Phylogeny

Phylogenetic analysis using the ITS, nc LSU rDNA, RPB2, and TEF1 regions were based on 51 Alternaria strains, 2 isolates of Acroconidiella tropaeoli, and 3 outgroup sequences (Fig. 1). The combined alignment was comprised of 2731 characters with gaps (625 for ITS, 852 for nc LSU rDNA, 865 for RPB2, and 389 for TEF1). The phylogenetic analyses generated by maximum parsimony, maximum likelihood (ML), and Bayesian analysis indicate that Acroconidiella tropaeoli grouped within the genus Alternaria and formed a monotypic lineage near sect. Sonchi with a well-supported clade (99%/100%/1.00, MP/ML/BI supports, respectively).

Fig. 1
figure 1

Phylogenies based on combined nc LSU rDNA, ITS, RPB2, and TEF1 showing the relationship of Alternaria obtusa with other closely related species within Alternaria. Bootstrap support values (MP and ML) or Bayesian posterior probabilities higher than 70% or 0.90 are indicated above or below the thickened branches (“–” indicates lack of support). Isolates from this study are indicated by bold text

Full size image

Another phylogenetic analysis was performed using the ITS and nc LSU rDNA loci to clarify the phylogenetic position of Acroconidiella trisepta (Fig. 2). The alignment of combined ITS and nc LSU rDNA sequence data comprised the total of 2168 characters with gaps (772 for ITS, 1396 for nc LSU rDNA). The dataset comprised 30 strains including 1 newly sequenced taxon and 2 outgroup sequences, Paradendryphiella arenariae and P. salina (Pleosporaceae). The phylogenetic analyses generated by maximum parsimony (MP), maximum likelihood (ML), and Bayesian analysis indicate that Acroconidiella trisepta belongs within the genus Dendryphiella with high support (100%/98%/1.00, respectively). The isolate of Acroconidiella trisepta clustered together with 2 other Dendryphiella species. Additionally, Acroconidiella trisepta formed a distinct lineage and was a sister to a strain of D. eucalyptorum.

Fig. 2
figure 2

Phylogenies based on combined nc LSU rDNA and ITS showing the relationship of Dendryphiella trisepta with other closely related genera from family Dictyosporiaceae. Bootstrap support values (MP and ML) or Bayesian posterior probabilities higher than 70% or 0.90 are indicated above or below the thickened branches (“–” indicates lack of support). Isolates from this study are indicated by bold text

Full size image

Taxonomy

Alternaria Nees, Syst. Pilze (Würzburg): 72 (1816)

≡ Syn. nov. Acroconidiella J.C. Lindq. & Alippi, Darwiniana 13: 612, 1964

Alternaria obtusaB.W. Ferreira & R.W. Barreto, nom. nov. (Fig. 3).

Fig. 3
figure 3

Alternaria obtusa (VIC 44410). a Leaf spots on Tropaeolum majus. b Colony on PDA after 7 days (after incubation at 25 °C in 12 h light/dark cycle). c Detail of tretic conidiogenous cell bearing attached immature conidium and one mature conidium (note constrictions at septa of mature conidium) and verrucose surface on both immature and mature conidia. dIbid one-septate immature conidium. e Acropetal conidial chain. f SEM image showing smooth conidiophores arising through the stoma and bearing one verrucose conidium (collapsed). g SEM image of one verrucose conidium. Bars = 10 μm except g = 2 μm

Full size image

MycoBank: MB832351

Etymology: having blunt/rounded-ended conidia

≡ Heterosporium tropaeoli Bond, Ceylon Journal of Science 12: 185, 1947

≡ Acroconidiella tropaeoli (Bond) Lindquist & Alippi, Darwiniana 13: 613, 1964

On fruits or, most abundantly, on leaves forming irregular to subcircular brownish or purple amphigenous spots, up to 1 cm diameter coalescing and leading to the blight of large areas of the leaf with yellow periphery. Colonies predominantly hypophyllous. Mycelium immersed, branched, septate, hyaline, smooth, hyphae 3–7 μm diameter. Conidiophores arising singly or in small groups through stomata or breaking through epidermis, erect, flexuous, simple, or occasionally branched, up to 180 μm long, 5–10 μm diameter, septate, often geniculate, sometimes slightly swollen at the apex, pale to mid olivaceous brown, smooth, conidial scars similar to those formed in Drechslera and Curvularia. Conidiogenous cells polytretic, integrated, terminal, subcylindrical to cylindrical, sometimes geniculate with sympodial proliferation, 6–22 × 4–5 μm, with one locus on a broadly obtuse apex, with a pigmented alternarioid scar, 2–4 μm wide. Conidia tretic, solitary or in short (2–3) unbranched chains, ellipsoidal, 30–50 (av. 41) μm long, 15–27 (av. 21) μm wide in the broadest part, lacking longitudinal septa, (1–) 2 (−3) transverse septa, strongly constricted at the septa, olivaceous brown, thin-walled, verruculose.

In culture: Fast growing (4–7 cm diam after 7 days), umbonate, cottonose, either fimbriate (in PCA) or entire edged (in PDA), smoke gray becoming white, gray olivaceous with white edges reverse. No sporulation was observed after 7 days of incubation.

Material examined: Brazil: Rio de Janeiro, Nova Friburgo, Ponte da Saudade, on Tropaeolum majus, 23 May 2017, R. W. Barreto (VIC 44410 – neotype designated here, neotype culture COAD 2389, MBT 388771); Minas Gerais, Antônio Carlos, Sítio São Jorge, on Tropaeolum majus, 20 May 2019, B. W. Ferreira (VIC 47206, culture COAD 2799).

Notes: Heterosporium tropaeoli was originally described from a specimen collected from a non-designated site in Sri Lanka on living leaves of T. majus. No herbarium specimen is known for this specimen, and no ex-type culture was designated. It was hence considered necessary to designate a neotype and ex-neotype culture here. It might be argued that this should be obtained from Sri Lanka. Nevertheless, A. tropaeoli is a pathogen known only from T. majus, which is a plant native to South America. The fungus has been reported from many countries ranging from Australasia to Africa and the Americas, but not from Eurasia (Farr and Rossman 2019). In South America, it has been first reported by Vieira and Barreto (2002), in Brazil. It is likely that its original occurrence and original description from Sri Lanka is fortuitous, and the fungus is actually native to South America which spread to other regions of the globe in planting material of its host but without being noticed in its center of origin until recently.

The new name Alternaria obtusa is proposed above for Acroconidiella tropaeoli. It could not be recombined into Alternaria tropaeoli because this name is already in use for another valid (and distinct) species in this genus described from India on T. majus.

Dendryphiella trisepta(Muchovej) B.W. Ferreira & R.W. Barreto, comb. nov. (Fig. 4)

Fig. 4
figure 4

Dendryphiella trisepta (VIC 44409). aGlycine max debris at the type locality (experimental area of the campus of the Universidade Federal de Viçosa, Viçosa, state of Minas Gerais, Brazil). b SEM images of D. trisepta colony on the surface of dead soybean stems. c, d SEM images of conidial chains. e Conidial chains under light microscopy. f Group of typical predominantly triseptate conidia (note: thickened and darkened conidial scars). g Colony on PDA after 15 days (incubation at 25 °C in 12-h light/dark cycle). Bars = 10 μm except b = 20 μm

Full size image

MycoBank: MB832343

Basionym. Acroconidiella trisepta J.J. Muchovej, Mycologia 72: 1045 (1980)

Colonies bursting through the cuticle and forming dark spots amidst colonies of a range of other saprophytic fungi on rotting stems of Glycine max. Conidiophores erect, macronemateous, forming loose tufts, cylindrical, flexuous, geniculate, up to 140 μm, 5–6.5 μm diam, reddish-brown, smooth. Conidiogenous cells polytretric, cylindrical, terminal, proliferating sympodially, slightly swollen apically, 5–13 μm in width. Conidia tretic, cylindrical, straight, 22–27(–33) × 8–10(–12) μm, apex obtuse to rounded, base rounded, mostly 3-septate, in acropetal unbranched chains, pale to mid reddish-brown, echinulate, hila thickened and darkened, 1 μm diameter.

In culture: Slow-growing (4.5–4.8 cm diameter after 15 days), umbonate, felty, edges entire, either entirely white or buff centrally with white edges, honey reverse with white edges. No sporulation was observed after 15 days of incubation.

Type material: Brazil: Minas Gerais: Viçosa, Universidade Federal de Viçosa, on dead branches Glycine max, 8 Jul 1979, J. J. Muchovej, Herbarium Universidade Federal de Viçosa (holotype lost); Minas Gerais: Viçosa, Universidade Federal de Viçosa, on dead branches Glycine max, 9 May 2017, R. W. Barreto (VIC 44409 – neotype designated here, ex-neotype culture COAD 2388, MBT 388772).

Notes: Dendryphiella trisepta was only known from soybean debris at the type locality. Colonies of the fungus were readily found in this substrate at the type locality in the first attempt of recollecting the fungus. As the original type material has been lost, and also because no ex-type culture was designated by Muchovej, a neotype and ex-neotype culture are designated here.

Species included in Acroconidiella

Acroconidiella eschscholziae (Harkn.) M.B. Ellis, nom. dub. More dematiaceous hyphomycetes. 407 (1976)

Basionym. Heterosporium escholtziae Harkn., Bull. Calif. Acad. Sci. 1: 38 (1884)

Colonies effuse, dark olivaceous. Mycelium immersed. Conidiophores solitary or in small groups developing between the epidermal cells, erect, flexuous, simple or occasionally branched, up to 85 μm long, 5–8 μm wide, often geniculate, pale olivaceous or golden-brown, smooth, with several dark-brown scars. Conidiogenous cells polytretic, sympodial. Conidia solitary or occasionally in short chains, mostly cylindrical, 28–90 × 9–19 μm, with 1–7 transverse and occasionally 1–2 longitudinal or oblique septa, often constricted at the septa, pale to mid-golden-brown, echinulate—as described in Ellis (1976).

Type material: USA: California, San Francisco, on leaves of Eschscholtzia californica, Jan 1884, Harkness (holotype destroyed)

Note: The presence of longitudinal septa in A. eschscholziae strongly suggests it to belong to Alternaria.

Acroconidiella indica I. B. Prasher & R. K. Verma, nom. dub. Journal on New Biological Reports 4: 111 (2015)

Colonies on natural substrate effuse, superficial on the substratum forming large stromatoid masses, made up of mycelium 3.2–6.4 μm wide, dark-brown, thick-walled, slightly roughened, extensively branched (branches close), short celled, bearing erect vertical conidiophores. Conidiophores 8–126 × 3.2–11.2 μm, brown, short or elongate, cylindrical, straight or slightly curved, septate, thick-walled, pigmented opaque, with a swollen basal cell, bearing conidiogenous cells. Conidiogenous cells rachiform, pale-brown to colorless, straight or flexuous, geniculated, geniculations thickened and minute, few (1–2), poroid. Conidia phaeo-, ceteri-, phragmo-, porosporous, acrogenous, 3–33.5 × 1.5–10.5 μm, brown to dark-brown, thick-walled, oval to elliptical or elongate, cylindrical, straight or slightly curved, dry, (1–6 celled), with transverse septa only, smooth, constricted at the septum: septa thick, distinct, apical cells rounded or sometimes pointed; basal cells more or less triangular and narrowed towards the hilum; hilum protruding, thickened. Germination of the conidia starts in situ or on the substrate by short germ tube which bears secondary conidia on them—as described in Prasher and Verma (2015).

Holotype: India: Himachal Pradesh: Solan, on dead twigs of unidentified tree, 10 Feb 2009, I. B. Prasher (PAN 30076)

Note: Prasher and Verma (2015) did not provide cultural or molecular data for A. indica and the morphology described for this fungus might place it in a range of similar genera such as: Dendryphiella, Paradendryphiella, or Alternariaster.

Acroconidiella manoharacharii I. B. Prasher & R. K. Verma, nom. dub. Journal on New Biological Reports 4: 113(2015)

Colonies on natural substratum black, minute, velvety, distributed throughout the substrate forming a scum. Mycelium immersed in the substrate, composed of branched, septate, brown, smooth-walled hyphae. Conidiophores 7–51 × 3–9.5 μm, branched, brown, short, cylindrical, straight or slightly curved, septate, with a swollen basal cell, pigmented opaque, thick-walled, bearing an apical conidiogenous cell. Conidiogenous cell rachiform, pale-brown, straight or flexuous, geniculated, geniculations thickened and minute, few (1–2) poroid. Conidia phragmo-, ceteri-, phaeo-, porosporous, borne singly, 8–20 × 3–8 μm, brown, thick-walled, oval to elliptic, straight or slightly curved, dry, (1–3 celled) with transverse septa, smooth, constricted at the septum; septa thick-walled, distinct; apical cell round or occasionally pointed; basal cell more or less triangular and narrowed towards the hilum; hilum protruding and thickened—as described in Prasher and Verma (2015).

Holotype: India: Himachal Pradesh: Shimla, Tara Devi, on angiospermous sticks, 23 Sep 2010, I. B. Prasher (PAN 30077)

Notes: There are no DNA sequences available in public databases for A. manoharacharii or any other fungi described in Acroconidiella except for those in this publication. It was not possible to examine specimens of A. manoharacharii deposited in PAN herbarium, and it appears that no culture of this fungus was obtained by the authors, since no description of cultural features is provided in Prasher and Verma (2015). As for A. indica, morphological features alone would allow its placement in a range of dematiaceous genera. It is hence better to leave the two species as nomen dubium until the fungus is recollected allowing for a proper re-evaluation including molecular typing.

Discussion

Alternaria was originally described by Nees von Esenbeck (1816), based on A. tenuis. The genus grew in number of species and included an assemblage of dematiaceous hyphomycetes producing dark-colored phaeodictyospores ended in a tapering beak and usually forming conidial chains (Ellis 1976). Alternaria was monographed by Simmons (2007) who, based on morphological and cultural features, recognized 273 Alternaria species and segregated several similar genera such as Aternariaster, Chalastospora, Prathoda, and Teretispora. A modern re-evaluation of Alternaria and related genera is presently under way (Lawrence et al. 2013; Woudenberg et al. 2013) and is producing major changes in the delimitation of this important group of fungi. Our results provide yet another contribution by modern-day mycologists and clarified that Acroconidiella is not a valid genus since its type species is a member of the genus Alternaria.

Lindquist and Alippi (1964) treated the genus Acroconidiella as distinct from Alternaria because it did not produce conidial chains or longitudinal or oblique septa. This combination of features suggested that it might be inadequately placed in Alternaria. This appeared a logical decision, based on morphology, at the time. Nevertheless, modern studies of Alternaria have indicated that the absence of longitudinal or oblique septae alone is not sufficient for excluding a species from Alternaria. Some examples of species of Alternaria having solely transverse septa are as follows: A. leucanthemi, A. thalictrina, A. thalictricola, A. thalictrigena, and members of sect. Nimbya (Schubert et al. 2007; Lawrence et al. 2012; Woudenberg et al. 2013). As for the presence or absence of conidial chains, a morphological feature which had taxonomic weight for erecting Acroconidiella, Vieira and Barreto (2002) had documented the presence of short conidial chains in A. trapaeoli—also observed and illustrated in the present study and also confirmed to be a dominant feature in A. trisepta, as already observed in the original description by Muchovej (1980). Acroconidiella eschscholziae described by Ellis (1976) presents conidia with longitudinal or oblique septa (as for the majority of the members of Alternaria) which are occasionally produced in short chains. It is very likely that A. eschscholziae is yet another species to be recombined into Alternaria. Acroconidiella tropaeoli was found here to be phylogenetically close to Alternaria brassicae, Alternaria cinerariae, and Alternaria sonchi. None of these taxa is morphologically similar to A. obtusa. Although sharing the feature of the absence of longitudinal/oblique septae with A. obtusa neither A. leucanthemi nor A. thalictrigena grouped with A. obtusa. There are no A. thalictrina or A. thalictricola sequences in GenBank available for comparison with the fungus on T. majus. However, these latter taxa have clear morphological distinctions separating them from A. obtusa. Both have longer, wider, rostrate conidia (Schubert et al. 2007). According to Lawrence et al. (2013), section Sonchi includes Alternaria cinerariae and Alternaria sonchi and is characterized morphologically by producing large subcylindrical, broadly ovoid, broadly ellipsoid or obclavate, formed singly or in short chains, with multiple transverse and few longitudinal septa, slightly constricted at the septa, with a blunt taper which can form secondary conidiophores. Lawrence et al. (2013) included A. brassicae as the basal lineage in sect. Sonchi, which was supported by a monotypic lineage in the analysis of Woudenberg et al. (2013). Alternaria brassicae has conidia which are straight or slightly curved, obclavate, rostrate, with transverse septa and longitudinal or oblique septa. Conversely, conidia of A. obtusa are ellipsoidal with rounded tips and lack longitudinal septa. Additionally, two species of Alternaria, A. tropaeoli (Deshpande and Rajderkar 1964) and A. tropaeolicola (Zhang 2000), have been described on Tropaeolus majus from India and China, respectively. Although there are no DNA sequences in public databases for these two species, they are clearly morphologically distinct from Alternaria obtusa. Both A. tropaeoli and A. tropaeolicola have rostrate, non-verrucous conidia which have transverse and longitudinal septae (Deshpande and Rajderkar 1964; Simmons 2007).

The fungus recollected from soybean debris was found here to be misplaced by Muchovej (1980). Acroconidella trisepta was found to clearly fit phylogenetically within the genus Dendryphiella. Dendryphiella was established by Ranojevic (1914) with the type species D. interseminata (Berk. & Ravenel) Bubák. Morphological characteristics used for delimitation of this genus included macronematous conidiophores with polytretic, integrated conidiogenous cells formed at the swollen tip and at intercalary swellings of conidiophores from which catenate or solitary conidia are formed (Ellis 1971; Matsushima 1971; Rai and Kamal 1986; Guo and Zhang 1999; Crous et al. 2014, 2016). Boonmee et al. (2016) performed phylogenetic analyses of Dendryphiella based on DNA sequence data for three loci (SSU, nc LSU rDNA, TEF1). The results showed that Dendryphiella represents a distinct genus within the Dictyosporiaceae. Different from other asexual morphs, genera belonging to this family—which form blastic conidiogenous cells—members of Dendryphiella have tretic conidiogenous cells.

Dendryphiella trisepta, as recombined in the present publication, has a distinct morphology from that of other species of Dendryphiella. Its conidiophores are shorter than those of D. eucaliptorum and D. aspera (up to 140 μm in length vs. up to 500 μm and 136–544 μm, respectively), and its conidia are larger [22–27 (–33) × 8–10 (–12) μm vs. (19–) 20–23 (–25) × 5 (–7) μm and 10–22 × 4–6 μm respectively]. In D. indica, D. paravinosa, D. eucalypti, D. vinosa, D. uniseptata, D. infuscans, and D. dregeae, conidiophores are mostly solitary, whereas in D. trisepta, these are formed in tufts. Dendryphiella broussonetiae and D. lycopersicifolia have conidiophores which are longer than those of D. trisepta (129–303 μm and 100–500 μm vs. 140 μm). Dendryphiella fasciculata has longer conidiophores (170–250 μm) and narrower conidia (4.3–7.4 μm) than those of D. trisepta.

Nuclear DNA of A. tropaeoli, the type species of Acroconidiella, was extracted and supported a phylogenetic evaluation of Acroconidiella performed here, for the first time. It demonstrated that the morphological delimitation utilized by the scientists who dealt with this genus in the past does not mirror the true phylogenetic placement of this taxon. As it has happened with other genera in the past, Acroconidiella became a “dumping ground” for somewhat morphologically similar taxa of uncertain placement, a fact demonstrated here by D. trisepta. By transferring A. tropaeoli to Alternaria, the generic name Acroconidiella can no longer be applied to the other species placed in the genus. There are doubts as for the correct taxonomic placement of A. eschscholziae, A. indicus, and A. manoharacharii. For D. trisepta, samples needed for re-evaluating the taxon were within the type locality, in the campus of the Universidade Federal de Viçosa where the work was performed. Nevertheless, it was surprisingly easy to recollect (in the first attempt) the material of A. trisepta from old soybean stems in the field, nearly forty years after it was collected and described by Muchovej (1980), suggesting that a stable population of this saprophytic fungus exists at this locality (an experimental area belonging to the Departamento de Fitotecnia of the Univesidade Federal de Viçosa). Although no other record exists of this fungus outside this area, it is likely that this fungus is present in soybean debris in other fields in Brazil and perhaps other parts of the world. Although it appears that this is a saprophytic species, no studies on its relationship with soybean or other plant species have ever been conducted (Baird et al. 1997, 2003; Almeida et al. 2001).

Recollecting, isolating and re-evaluating the other species in the former genus Acroconidiella is a pending challenge, among so many others posed to mycologists worldwide of recollecting, neotypifying, or epitipifying the fungal taxa of the past (Hyde and Zhang 2008). Until then, these species are to be regarded as Incertae sedis.