Studies on European rust fungi, Pucciniales: molecular phylogeny, taxonomy, and nomenclature of miscellaneous genera and species in Pucciniastraceae and Coleosporiaceae

Using molecular phylogenetic analyses (ITS) and morphological data obtained from light and electron microscopy, some European and North American species and genera placed or formerly placed in the genus Pucciniastrum in the Coleosporiaceae and Pucciniastraceae (Pucciniales) were taxonomically revised. The ITS analyses confirmed recent familiar concepts based on less variable markers except for the genus Hyalopsora. The family Pucciniastraceae is characterized by Abietoideae (Abies, Tsuga) aecial hosts. Pucciniastrum is described as a genus that consists of host-alternating species forming aecia on needles of Abies hosts, with special features of aeciospore morphology, and Onagraceae telial hosts. Other genera in the Pucciniastraceae are Calyptospora, Melampsorella, and additional taxa, which are currently provisionally placed in Pucciniastrum, but must be revised in future studies. Pucciniastrum epilobii (s. lat.), the type species of Pucciniastraceae, represents at least two species with different life cycles and urediniospore characteristics and is lecto- and epitypified. The family Coleosporiaceae, characterized by Pinoideae (Pinus) and Piceoideae (Picea) aecial hosts, contains Rosaceae rusts from three well-supported clades represented by three genera, Thekopsora, Quasipucciniastrum, and Aculeastrum gen. nov. Aculeastrum is characterized by coarsely arcuate ostiolar peridial cells and infects Rubus spp. telial hosts. The following new taxonomic combinations are proposed: Calyptospora ornamentalis comb. nov., Quasipucciniastrum ochraceum comb. nov., Q. potentillae comb. nov, Aculeastrum americanum comb. nov., and A. arcticum comb. nov. The results are discussed with emphasis on future studies in Pucciniastrum and the P. epilobii complex and on nomenclatural changes necessary for rust fungi due to the Shenzhen Code.

In the present study, we sequenced the ITS marker from mainly European species placed or formerly placed in Pucciniastrum to re-assess their taxonomic (specific and generic) position within the two families Coleosporiaceae and Puccinistraceae s. str. Aime and McTaggart (2020). The basis for our studies is the higher rank phylogeny of Aime and McTaggart (2020) mentioned above. Furthermore, we studied urediniospore germ pores and spore surface features in Pucciniastrum epilobii s. lat. and Rosaceae rusts within Coleosporiaceae using light and electron microscopy. These features have recently been documented for species and sectional delimitation in Milesina (Bubner et al. 2019) but have not been considered in morphological studies of P. epilobii s. lat. (Klebahn 1900(Klebahn , 1914Bisby 1916;Pady 1933) and the rusts on Rosaceae (Ziller 1974;Helfer 2005;Qi et al. 2019). Finally, we carried out a detailed study on the taxonomy and nomenclature of these species with respect to the rules of the current "International Code of Nomenclature for algae, fungi, and plants" (Shenzhen Code, 2018), supplemented by the revised version of Chapter F of the Code, incorporating amendments approved by the Fungal Nomenclature Session of the 11th International Mycological Congress held in San Juan, Puerto Rico, in July 2018 (May et al. 2019).

DNA extraction, PCR, and sequencing
Samples for DNA extraction were prepared from herbarium specimens (Table 1) by excising a single sorus including the plant material. They were placed into microtubes with 8-12 ceramic beads, 1.4-mm diameter (Bio-Budget Technologies, Krefeld, Germany), frozen at −20°C overnight, and homogenized on a Bead Ruptor (Biolab Products, Bebensee, Germany) at a speed of 7.45 m/s for 25 s. After freezing the samples again for 10 min at −20°C, homogenisation was repeated. DNA was extracted with the DNeasy Plant Mini Kit (Qiagen, Hilden, Germany), following the manufacturer's protocol. Selected samples were homogenised with glass mini mortars and pestles (Roth, Karlsruhe, Germany) in 400 μl of the homogenisation buffer included in the extraction kit.
Molecular barcodes were generated for the ITS region by using the primers ITS1F (Gardes and Bruns 1993) and ITS4rust (Beenken et al. 2012(Beenken et al. , 2017. PCR was performed with the Accuprime Taq Polymerase System (Life Technologies, Karlsruhe, Germany) using the supplied buffer II and the following final concentrations: 2 mM MgCl 2 , 0.2 mM of each dNTP, and 500 nM of each primer. The PCR programme was as follows: 3 min denaturation at 94°C , 10 cycles with decreasing annealing temperature (94°C for 30 s, 60-50°C for 30 s, and 68°C for 60 s), 35 cycles with constant annealing temperature (94°C for 30 s, 50°C for 30 s, and 68°C for 60 s), and 7 min strand completion at 68°C. PCR products were visualised in 1.6% agarose gel.
After purification of the PCR products with QIAquick-PCR Purification Kit (Qiagen, Hilden, Germany), they were sent to GATC Biotech AG (Konstanz, Germany) for sequencing. Sequencing was performed with the same primers used for the PCR. Forward and reverse sequences were edited and assembled with the software package GENEIOUS 10.0 (Biomatters, Auckland, New Zealand).

Phylogenetic analysis
Sequences were aligned with the CrustalW algorithm implemented in the programme Bio Edit, version 7.1.3.0 (Hall 1999), using the standard parameters offered by the programme. Two sequences of Milesina kriegeriana (Bubner et al. 2019) were used as the outgroup. Alignments were used for phylogenetic reconstruction by three different methods:

i)
Maximum likelihood (ML) analysis: The alignment was analysed under the ML criterion on the Webbased RAxML black box (Stamatakis et al. 2008). The substitution model used was GTR without GAMMA correction for amongst-site heterogeneity.
Node support values were calculated from 100 bootstrap replicates.

ii)
Bayesian inference (BI) analysis: The DNAsubstitution model GTR+I+G was used for performing Bayesian analysis with the programme MrBayes 3.2 (Ronquist et al. 2012). Two independent MCMC runs were performed, each with four chains over 1,000,000 generations. Every 100th tree was sampled. Initial burn-in was 25% and summarisations were calculated after the standard deviation of split frequencies reached below 0.01. The resulting tree file contained posterior probability values for node support. iii) Maximum Parsimony (MP) analysis: Phylogenetic reconstruction was performed with the software package MEGA version 10.2.1 (Kumar et al. 2018). Node support values were calculated from 1000 bootstrap replicates.
Tree files resulting from the three methods were visualised using the programme TreeGraph 2 (Stöver and Müller 2010).

Light and scanning electron microscopy
For microscopical identification and morphological studies of Pucciniastrum epilobii s. lat., specimens were mounted in a mixture of lactic acid and glycerol (Kirk et al. 2001) and examined with a Zeiss Axioskop 2 plus microscope (Oberkochen, Germany) at a magnification of ×400 or ×1000. Specimens were identified as described by Klenke and Scholler (2015). Germ pores of urediniospores of Pucciniastrum epilobii s. lat. were visualised by the method described in Bubner et al. (2019). For scanning electron microscopy (SEM), uredinia and urediniospores of dried specimens of P. epilobii s. lat. were placed on a holder with conductive double-sided tape (Leit-Tabs, Plano GmbH). Scanning electron microscope images were obtained on a Philips XL 30 FEG environmental scanning electron microscope operated at an acceleration voltage of 12 kV at a chamber pressure of 133 Pa (1 Torr). To achieve better contrast and less charge effects, the samples were coated with a mixture of gold (80%) and palladium (20%) first (MED 020, BAL-TEC). SEM studies were carried out to study surface structures that are not visible by light microscopy. In addition, spine base diameters (30 per species) were measured with SEM and the software ImageJ 1.5.

Additional specimens examined morphologically
The following is a list of specimens studied morphologically in addition to (also sequenced) specimens listed in Table 1. Spore states of rust fungi are listed in Roman numerals. A Roman numeral in brackets indicates that this spore state was only formed sparsely. Acronyms of countries follow the International Organization for Standardization (ISO) alpha-2 country codes (https://www.iso.org/home.html). Fungus names are inferred from voucher specimens; host names are "current names" from the Catalogue of Life (COL | The Catalogue of Life) except for the genus Chamaenerion which we separate from Epilobium. Table 1 Names, specimens, and accession numbers for the ITS barcode. Specimens with lab numbers were sequenced for this study and submitted to GenBank. These specimens were also morphologically studied. Collection data are provided for these specimens

Phylogenetic analysis
Two major clades obtained by the analysis of ITS rDNA data of rust species placed or formerly placed in the genus Pucciniastrum ( Fig. 1) coincide with the families Pucciniastraceae and Coleosporiaceae as determined with the more conservative markers used by Aime and McTaggart (2020). Specimens named Pucciniastrum are mainly found in the Pucciniastraceae, but two species origin al l y i nc l u d ed in P u cc i ni as tr um c l u s t er i n th e Coleosporiaceae s. str. Aime and McTaggart (2020).

Pucciniastraceae
The phylogenetic analysis shows that all species in the Pucciniastraceae have aecial host species in the subfamily Abietoideae (Pinaceae). The position of Hyalopsora aspidiotus is not yet clear because the support values for the maximum likelihood and the maximum parsimony analyses are relatively low. The ITS phylogeny places this species in the Pucciniastraceae in contrast to the higher rank phylogeny presented by Aime and McTaggart (2020) which places this species in the Coleosporiaceae.
Among the European species of Pucciniastrum s. str., two species on Onagraceae can be distinguished with high support values, one on Chamaenerion and another one on Epilobium and Oenothera spp. They are not sister species. The reference sequence of P. circaeae on Circaea erubescens from Japan is clearly distinct from sequences of P. circaeae on Circaea lutetiana from Switzerland (7 different bases of 606 bp) and Pucciniastrum on Fuchsia. The Japanese species is considered an undescribed species (P. aff. circaeae), whereas the Fuchsia rust is conspecific with P. circaeae. In this respect, we follow Garibaldi et al. (2012) who previously placed P. fuchsiae in P. circaeae. The next clade consists of Thekopsora minima (host-alternating with Tsuga) and T. guttata (aecial host unknown), two species that are morphologically poorly known. An important feature in rust fungi, the number and position of pores in urediniospores, e.g., are not known for T. guttata (Klenke and Scholler 2015) or for T. minima (Arthur 1934). Both species as well as the basal Pucciniastrum species from Asia ( Fig. 1) should be placed in different genera, but more studies are necessary to describe the genera. We provisionally placed them in Pucciniastrum. Calyptospora is reported from Europe and North America. In the present phylogeny, the European and the American specimens can be differentiated as two species (Fig. 1). These two species are not closely related to the cluster containing Pucciniastrum s. str. Therefore, a placement in a separate genus (Calyptospora) is justified from the phylogenetic perspective. The European species was sequenced from both the aecial host, Abies alba, and the telial host, Vaccinium vitisidea, thus proving the host alternation of this rare species by molecular methods. The sister genus of Calyptospora, Melampsorella, is characterized by special symptoms (witches' brooms) and morphology and deserves a separate genus. Both genera have Abies aecial hosts.

Coleosporiaceae
The ITS phylogeny shows that Coleosporiaceae species are restricted to hosts of the Pinoideae (Pinus) and Piceoideae (Picea). All species with Rosaceae telial hosts are placed in Coleosporiaceae confirming the results of Aime and McTaggart (2020).
Sequences retrieved from Pucciniastrum on Agrimonia spp. in Asia (China), Europe (Germany), and North America (USA) cluster together and coincide with each other, suggesting a single species, which is, however, phylogenetically distant from P. epilobii and thus Pucciniastrum s. str. (for nomenclatural changes see below and right column in Fig. 1). Pucciniastrum americanum clusters with Pucciniastrum on Agrimonia spp. and is phylogenetically closely related. The sequence of the holotype Quasipucciniastrum agrimoniae (voucher HMAS248095, GenBank MK208281, Qui et al. 2019) comprises only ITS2 and, therefore, has not been included in the phylogenetic analyses of Fig. 1. Sequence comparisons with the ITS2 showed that it is completely identical with the ITS2 of the newly generated sequences of Quasipucciniastrum ochraceum.
Our studies show that Pucciniastrum species on Epilobium s. lat. (Epilobium incl. Chamaenerion) may be separated into three morpho-taxa which correspond well with the phylogenetic results. These taxa are named Pucciniastrum epilobii (on Ch. angustifolium), P. pustulatum (on Epilobium spp. and Oenothera spp.), and P. epilobii-dodonaei (on Ch. dodonei). Puccinastrum epilobii-dodonaei is characterized by its systemic growth that deforms its host and prevents flowering. This species is known only from Chamaenerion dodonei. The species forms uredinia and telia amphigenously. Telia are small, 0.05-0.15 mm. Pucciniastrum epilobii is not systemic. The species forms uredinia and telia amphigenously. Telia measure 0.05-0.7 mm. Urediniospores measure 16.5-22.0 × 11.5-15 μm. The surface of spores is usually covered evenly with spines; we found spine-free areas only very rarely (Fig. 2c, d). Urediniospore germ pore number is 4-7. The third species, P. pustulatum, is not systemic and was verified on Epilobium (s. str.) spp. and on Oenothera spp. We did not observe any specimens with telia for this species. Urediniospores match the size, shape, and germ pore number of P. epilobii, but about 25% have at least one spine-free zone (Fig. 2e, f). Since all three species may occur on Ch. dodonei (Gäumann 1959), these features help to morphologically distinguish the species.  (1801), was the first name involved in the Pucciniastrum epilobii complex. Persoon (l.c.) separated this species into two unranked taxa, α. epilobii (on Epilobium montanum) and β. cerastii. De Candolle (1815) listed U. pustulata and its subspecific taxa according to Persoon (l.c.), and he determined them in the accompanying text as varieties. De Candolle used the names "epilobiorum" and "cerastiorum" (plural) since he cited in each case two hosts for the two taxa, but he did not intend to introduce new varieties, so that "epilobiorum" and "cerastiorum" should rather be considered orthographic variants of Persoon's names. Schlechtendal (1824) removed β. (var.) cerastii from U. pustulata and raised this taxon to species rank as Caeoma cerastii (Pers.) Schltdl., which is currently a synonym of Melampsorella caryophyllacearum (DC.) J. Schröt. In doing so, Schlechtendal (l.c.) confined U. pustulata to var. epilobii, which constituted a kind of "indirect lectotypification". Hence, since that time, the species name U. pustulata referred to the rust fungus on Epilobium spp., and it was later correctly used in this sense as basionym for several combinations, such as Erysibe pustulata, Melampsora pustulata, and Pucciniastrum pustulatum.
Otth (1861: p. 71) described the new genus Pucciniastrum with P. epilobii as type species. "(Pers.) G.H. Otth" was often referred to as author citation for the latter species name, suggesting that Otth (l.c.) had based this name on Uredo pustulata α. epilobii (conforming to var. pustulata), but this is incorrect. On p. 65, Otth (l.c.) Otth (l.c.) intended to introduce a new combination based on S. herbarum f. epilobii, as Otth's (1863: p. 85) later citation of its own name as "P. epilobii (Chaill.)" would suggest, or if he had in mind a new species, with S. herbarum f. epilobii as synonym (the latter name is a forma that does not have priority on species rank). Otth's treatment of his own name P. epilobii in "Wartmann & Schenk, Schweizerische Kryptogamen 615" (Fig. 2a) allows interpretation of Otth's intentions. In this exsiccata, he cited S. herbarum f. epilobii as a synonym of P. epilobii sp. nov., i.e. he had a new species in mind and did not intend to introduce a new combination (comb. et stat. nov.) based on Chaillet's forma.
In the original publication, Otth (1861) did not provide details for the specimens examined, but in his supplement (Otth 1863) he listed an undated collection from "Heimberg, on Epilobium angustifolium" which is undoubtedly original material (ICN, Art. 9.4) available for lectotypification. This material is preserved and deposited in the herbarium in Zürich as ZT Myc 43369 and the only preserved collection from Otth's herbarium under the name P. epilobii (the material distributed in "Wartmann & Schenk, Schweizerische Kryptogamen 615" was collected in 1869, i.e. it does not represent original material). Whether Otth had examined any material collected by Chaillet, which would also be original material, remains unclear. To stabilize available names within the P. epilobii complex, it is necessary to clarify the typification of the names involved and to propose epitypes that can be sequenced to get data for their phylogenetic characterization. The search for type material of Chaillet's taxon in several herbaria, including G, HAL (herb. Schlechtendal), UPS, and ZT, was not successful, suggesting that type material was probably not preserved.
Pucciniastrum pustulatum represents the oldest valid name for Pucciniastrum on Chamaenerion and Epilobium spp. and other species of genera of the Onagraceae and would be the correct name if the collections on all hosts were regarded as a single species. Most previous authors used the name P. epilobii in this broad sense. On the other hand, some previous authors, such as Kuprevich and Ul'yanishchev (1975), already recognized two species, in the latter case as P. abietischamaenerii and P. pustulata. However, taking into account the latest phylogenetic results for the P. epilobii complex (Hietala et al. 2008, the present work), connected with the phylogenetic differentiation of Epilobium s. lat. into two well-supported clades (Epilobium s. str. and Chamaenerion, see Baum et al. 1994, Wagner et al. 2007, as well as obvious differences in the development cycles, biological specialization of the taxa involved and morphological differences (see above), this complex must be split into two species, one of them confined to Chamaenerion spp. (= Pucciniastrum epilobii) and another polyphagous one to species of the genera Epilobium, Oenothera, and possibly Clarkia (Klenke and Scholler 2015) and Chamaenerion (Gäumann 1959 (Fig. 2b). Fuckel listed two different specimens under no. 300, viz., 300. I. (Uredo epilobii DC. on Epilobium montanum) and 300. II. (Melampsora epilobii Fuckel on Chamaenerion angustifolium). The examination of the urediniospores on Epilobium montanum (KR), identified by Fuckel as "Uredo epilobii", revealed that they belong to Uredo pustulata. Different interpretations of the name M. epilobii are possible. Fuckel cited Uredo epilobii DC. under M. epilobii, which could be interpreted as citation of a synonym for the asexual uredinia/urediniospores on another host (Epilobium montanum), i.e. in this scenario, a valid name that should have been adopted according to the current Code would have been included in the protologue. With a different epithet, Fuckel's name would in this case be interpretable to be superfluous (nom. illeg. [nom. superfl.], Art. 52.1), but since he used the epithet "epilobii" Fuckel's name could also be considered a new combination based on de Candolle's Uredo epilobii, a species that does, however, not belong to the Pucciniastrum epilobii complex. In this case, Fuckel's M. epilobii would be a homotypic synonym of Uredo epilobii DC. (This name is still unresolved, but according to the original description, it might rather be a synonym of Puccinia pulverulenta Grev.) However, Fuckel (l.c.) indicated that he intended to introduce the name M. epilobii only for the telia found on Chamaenerion angustifolium ("II. Melampsora propria") [proprius = exclusively belonging to]. Thus instead, M. epilobii represents a separate valid species name confined to Chamaenerion angustifolium (Art. F.8.1), which has to be reduced to synonymy with Pucciniastrum epilobii. This name requires lectotypification: , the latter as "Uredo epilobiae" (see also Jørstad 1958: p. 9). All collections are undated. However, Persoon (1801) published Uredo pustulata α. epilobii on Epilobium montanum, whereas the host name for the two collections is given only as "Epilobium". Hence, it is unclear whether these collections are suitable as potential lectotypes and if they were collected prior to 1801. Three additional collections preserved at Leiden herbarium under Uredo pustulata, viz., L 0119130 (G.B. Balbis, s.n.), L 0119131 (J.F. Chailett), and L 0119132 (J.F. Chailett), were collected after 1801 and are unsuitable for lectotypification purposes (see also comments in Jørstad 1958). Therefore, the designation of a neotype for this taxon seems to be indicated, which also allows determining the genetic position of this species by ex-neotype sequences.
Another European Pucciniastrum species on Onagraceae (Oenothera), Pucciniastrum oenotherae Gaillard, is a chytrid (Synchytrium fulgens J. Schröt.) according to Sydow and Sydow (1915 Fig. 2b) published results of inoculation experiments, distributed several collections, and discussed the taxonomy and nomenclature of Calyptospora. He ascertained the synonymy of Aecidium columnare and C. goeppertiana, and, because of the priority of A. columnare, he introduced the new combination Calyptospora columnaris. We prefer to maintain the name Calyptospora columnaris for this rust fungus, based on the oldest basionym, Aecidium columnare, which has priority according to Art. F.8.1, Note 2, although there is not automatism in the nomenclature of pleomorphic fungi, i.e. it would also be possible to propose maintaining the younger teleomorph-typified name C. goeppertianum by protection or by adding it to a list of names proposed for protection, according to Art. F.2.1. However, Kühn (1869) accepted the oldest names and introduced the corresponding combination. Another problem concerns North American collections of C. goeppertiana s. lat. Two North American species are known, viz. Peridermium ornamentale (Arthur 1901), described based on an aecial host collection on Abies lasiocarpa from Washington, and Pe. holwayi (Sydow and Sydow 1903), introduced for aecial host spore stages (0 and I) found on Pseudotsuga menziesii in Canada, British Columbia. Arthur (1934) reduced both species names to synonymy with Pucciniastrum goeppertianum, but Faull (1939) disagreed and emphasized that collections of C. goeppertiana s. lat. are morphologically differentiated into two taxa, viz. C. goeppertiana s. str. in eastern North America and a taxon in western North America for which the name Pe. ornamentale is available. The existence of a different taxon of Calyptospora, possibly a species of its own, in western North America was already assumed by Weir (1926). In the western taxon, Pe. ornamentale, the aecia and spermogonia (type 3, according to Cummins and Hiratsuka 2003) are formed on current year needles in late summer or early autumn, whereas in P. goeppertianum from eastern North America aecia are formed in early summer, but spermogonia are lacking or sometimes developed only in degenerated form (Hiratsuka et al. 1967). Pe. holwayi has usually been reduced to synonymy with Pe. ornamentale, but the genuine identity of this species, based on an aecial host spore stage on Pseudotsuga menziesii, remains unclear and needs further detailed examinations, including phylogenetic analyses. Aecia of P. holwayi occur on Pseudotsuga, and are formed on previous year needles in early summer, in contrast to P. ornamentale growing on Abies and forming aecia on current year needles in late summer or early autumn. Vogler and Bruns (1998) generated sequence data retrieved from Calyptospora goeppertiana s. lat. (= Peridermium ornamentale) on Abies grandis (native to the Pacific Northwest and Northern California), belonging to the western taxon in North America. These sequences have been included in the present phylogenetic analysis. They cluster in sister position to the European sequences of C. columnaris. Thus, these preliminary data support Weir's (1926) andFaull's (1939) careful examinations and conclusions that there is a different Calyptospora species in western North America for which the name Pe. ornamentale is available. However, this can only be the very first step towards a revision of the C. goeppertiana complex in North America. The taxonomic and phylogenetic position of Calyptospora goeppertiana in eastern North America, although morphologically indistinguishable from European collections of this species, is still unproven, i.e. it remains open whether the eastern North American taxon is actually conspecific with C. goeppertiana s. str. or if a third species may exist.
The nomenclature of the two species involved in this complex can be summarized as follows: Notes: This species has been previously mostly incorrectly cited as "Melampsorella caryophyllacearum (DC.) J. Schröt.". This reference seems to be influenced by Schröter (1874), who cited "(D. C. Uredo)". However, the name "Uredo caryophyllacearum DC., 1805" does not exist. It was neither published on p. 85 (as cited in the databases), nor in the whole volume of de Candolle (1805), nor de Candolle, in de Lamarck (1806), or de Candolle (1815). Caeoma caryophyllacearum Link can also not be considered as a potential basionym for Melampsorella caryophyllacearum because Link's C. caryophyllacearum is an illegitimate superfluous name (Link cited two older available names, Caeoma stellariae Link and Uredo ovata F. Strauss, as synonyms in the protologue). Therefore, Melampsorella caryophyllacearum must be ascribed to Schröter (1874). Aecidium elatinum is the oldest valid species name for this rust, so that Melampsorella elatina is the currently correct name. Note: The combination Pucciniastrum circaeae is based on de Thümen's replacement name Melampsora circaeae (≡ Uredo circaeae Alb. & Schwein., nom. illeg.) and not on Uredo circaeae Schumach., as already correctly cited by Sydow and Sydow (1915: p. 445), so that a combination using the oldest valid name for this taxon, Uredo circaeae Schumach., is prevented.

Pucciniastrum
Pucciniastrum minimum (Schwein.) Arthur, Résult. Sci. Congr. Bot. Wien 1905: 337, 1906. Note: Pucciniastrum minimum is a combination based on Uredo minima. Caeoma azaleae is a superfluous name, which was published with reference to Uredo minima (as "Syn. Carol. 470"), a name that should have been adopted in Caeoma. The species forms a separate clade, needs further investigation, and may be placed in a different genus. Note: Melampsora guttata is the nomenclaturally correct basionym for this species. Caeoma galii Link is an illegitimate younger homonym of C. galii (Pers.) Schltdl., which was only validated in 1881 as Melampsora galii G. Winter. Uredo galii Duby is not a replacement name for Caeoma galii Link since Duby (1830) only cited the latter name as a questionable synonym (with question mark), whereas Winter (l.c.) did not cite Duby's name at all. The species forms a separate clade in our ITS phylogeny. It needs further investigation and may be placed into a different genus.
This also applies to P. fagi G. Yamada and P. hydrangeaepetiolaris Hirats. fil. which form a basal clade in our ITS phylogeny (Fig. 1).

Coleosporiaceae
Taxonomy, morphology, and nomenclature of genera and species with Rosaceae hosts In the higher rank phylogeny of Aime and McTaggart (2020), rusts on Rosaceae in the Coleosporiaceae formerly placed in Pucciniastrum form a single clade with species on Agrimonia, Potentilla, Prunus, and Rubus with two well-supported subclades formed by species on Rubus and Prunus. The authors placed all species in the genus Thekopsora based on phylogenetic data exclusively. Our ITS phylogeny (Fig. 1) c o n f i r m s t h a t s p e c i e s o n R o s a c e a e b e l o n g t o Coleosporiaceae. There are three highly supported clades, one clade with Agrimonia as host which forms a sister clade with Rubus rusts and one clade with Prunus as host. Since there are remarkable differences between species representing the three clades, we propose placing them in three different genera, Quasipucciniastrum for species on Agrimonia and Potentilla, Thekopsora for species on Prunus (represented by two species, the type species T. areolata (Fr.) Magnus and T. pseudo-cerasi Hiratsuka fil.), and for species on Rubus we propose a new genus. Differences between them concern host range, life cycle, symptomology, and morphology and are listed in Table 2. This requires nomenclatural changes: Table 2 Features of the the three rosaceous rust genera Aculeastrum, Quasipucciniastrum, and Thekopsora (Coleosporiaceae) formerly placed in Pucciniastrum. Features of uredinia and urediniospores are inferred from this study; all other data are from Fischer (1904) (1887), erroneously introduced the combination "Uredo agrimoniae (DC.) J. Schröt." that must be corrected to Uredo agrimoniaeeupatoriae (DC.) J. Schröt. because Uredo potentillarum var. agrimoniae-eupatoriae was cited as a basionym. Dietel (1890) provided the first description of telia/ teliospores of this species. He named this species "Thekopsora agrimoniae (DC.)", suggesting that he based this name on Uredo potentillarum var. agrimoniae-eupatoriae as basionym and intended a new combination, based on de Candolle's (1815) variety. However, he did not use the name "Thekopsora agrimoniae-eupatoriae" but T. agrimoniae [the name "Thekopsora agrimoniae" was previously, under the rules of the Codes before 2012, prior to the Melbourne Code, ascribed to Dietel as teleomorph-typified name, and in the event that the original author intended to introduce a new teleomorph-typified name, which can be assumed in this case, T. agrimoniae can actually be ascribed to Dietel and represents a legitimate name, according to the current Code (Art. F.8.1)]. The nomenclature of Dietel's (1890) name influences the nomenclature of Pucciniastrum agrimoniae (Dietel) Tranzschel (Tranzschel 1893) as well. Later, Tranzschel (1939 probably realized that Caeoma agrimoniae represents the oldest name in this taxonomic-nomenclatural complex and introduced the new combination Pucciniastrum agrimoniae (Schwein.) Tranzschel.
Recent phylogenetic analyses showed that Pucciniastrum on Agrimonia spp. in Asia must be excluded from Pucciniastrum s. str. Qi et al. (2019) introduced the new genus name Quasipucciniastrum with its type species Q. agrimoniae X.H. Qi, P. Zhao & L. Cai. However, there are two basic problems in the latter publication lying in the missing phylogenetic comparison of the new genus with Pucciniastrum s. str., determined by its type species, P. epilobii, from which the new genus has been segregated, and a missing comparison with sequence data of Pucciniastrum on Agrimonia eupatoria from Europe for which older species names are available. In a comprehensive phylogenetic study on rust fungi, Aime and McTaggart (2010) showed that Q. agrimoniae and the fungus previously referred to as Pucciniastrum agrimoniae cluster together with sequences retrieved from Thekopsora areolata, the type species of Thekopsora, i.e. the species on Agrimonia spp., was considered to be congeneric with Thekopsora. However, the present phylogenetic examinations revealed that a single species in the Northern hemisphere (Asia, Europe, and North America) is involved (see "Introduction"). These results raise the question as to the correct naming of the species involved. The oldest epithet ("agrimoniae"from Caeoma agrimoniae) is not available in Thekopsora since "agrimoniae" is preoccupied in this genus by Thekopsora agrimoniae. Coleosporium ochraceum represents the next name within the synonymy that is available and would have priority in Thekopsora. However, because our own phylogenetic examinations support the conclusion that the former Pucciniastrum/Thekopsora agrimoniae on Agrimonia spp. deserves a genus of its own, we accept Quasipucciniastrum as introduced by Qi et al. (2019). However, the oldest epithet "agrimoniae" (from Caeoma agrimoniae) is also not available in Quasipucciniastrum and preoccupied by the recently introduced Q. agrimoniae, so that the next available name, Coleosporium ochraceaum, takes priority: ≡ Uredo agrimoniae-eupatoriae (DC.) J. Schröt. [as "agrimoniae (De Candolle)"], in Cohn, Krypt.-Fl. Schlesien (Breslau)  Notes: Herbarium material authentic for Bonordon's (l.c.) name is not preserved, but he provided an original illustration, which is original material in the sense of the Code (Art. 9.4) that has to be taken into consideration for a lectotypification (Art. 9.3). The application of Bonordon's name is established by epitypification. Note: Pucciniastrum potentillae, a species distributed in the northern hemisphere, resembles Q. ochraceum but has smaller urediniospores and teliospores (Arthur 1934). No ITS sequence is available but Aime and McTaggart (2020) proved in their higher rank phylogeny that it is closely related to Q. ochraceum (Thekopsora pseudoagrimoniae). Because of this and because of its morphological similarity to Q. ochraceum, we suggest placing this species in Quasipucciniastrum as well.

Pucciniastrum on Rubus
Pucciniastrum americanum and P. arcticum on Rubus spp. (host-alternating with Picea glauca) have not only several features in common with Quasipucciniastrum and with Thekopsora but also features which differentiate them from the latter genera. All features concerning host range, geographical distribution, symptomology, life cycle, and morphology are listed in Table 2. Besides the host plants, Rubus rusts (P. americanum and P. arcticum) have several morphological features that are missing in Thekopsora and Quasipucciniastrum and all other species so far placed in Pucciniastrum: Uredinia have four to six ostiolar knob-like and conspicuously spiny cells (Fig. 3a, b; for comparison see uredinia of Q. agrimoniae in Fig.  3c), the germ pore number and position are different (Table 2) and finally echinulation (shorter spines on central surface) differs from the other two genera (Fig. 3c). Also, spermogonia and aecia are formed on Picea needles, not on cones as in Thekopsora. Based on differences in morphology, symptomology, life cycle features, and host range, supported by two rDNA phylogenies, we propose placing Pucciniastrum americanum and P. arcticum in a different genus: Aculeastrum M. Scholler, U. Braun & Bubner gen. nov.
Aculeastrum differs from related Coleosporiaceae genera on Rosaceae telial hosts, Quasipucciniastrum and Thekopsora, in having a different telial host range (Agrimonia, Potentilla), by coarsely spiny ostiolor cells, by fewer (mostly 4-6) strictly bipolar urediniospore germ pores, by shorter urediniospore spines in the center of the spore, and by forming spermogonia and aecia on needles of the aecial host (Picea).
Note: This species strongly resembles A. americanum. The main feature of Aculeastrum species, the coarsely arcuate ostiolar cells are well-developed in this species. Aculeastrum articum differs from A. americanum in its host range (other Picea and Rubus species), by uredinia which only scarcely protrude above the epidermis, and by its peridial cell morphology (Moss 1926;Arthur 1934;Hiratsuka 1936;Ziller 1974).
Key to the rust genera in Coleosporiaceae with Rosaceae telial hosts:

Use of ITS rDNA data and delimitation of genera
This study is based on ITS rDNA data because the aim was to explore species delimitation in the genus Pucciniastrum s. lat. A previous study in rust fungi showed that for species delimitation, ITS rDNA data are more suitable because it is more variable than, for instance, the marker 28S rDNA.
However, phylogenetic analyses in Pucciniastrum have already shown that this genus in its previously wide circumscription is a heterogeneous, non-monophyletic complex, i.e. in a phylogenetic context some species should be placed in different genera (Maier et al. 2003;Aime et al. 2018;Qi et al. 2019). The most recent and most comprehensive study that includes members of the genus Pucciniastrum s. lat. is the study of Aime and McTaggart (2020). Because it is a study on higher rank classification, authors used the markers 28S, 18S, and CO3. Aime and McTaggart (2020) showed that Puccininastrum in its wide circumscription comprises species that are so distantly related that they are even placed in different families, Coleosporaceae and Pucciniastraceae. Despite the large genetic distance, our study of Pucciniastrum s. lat. and related genera based on ITS showed almost the same basic structure in the monophyletic groups as the study of Aime and McTaggart (2020). There is only one genus, Hyalopsora, which in our ITS phylogeny is placed in Pucciniastraceae, whereas Aime and McTaggart (2020) placed it in the Coleosporiaceae. Considering coevolutionary arguments and thus also host plants of Hyalopsora (Abies, fern), our ITS phylogeny is more convincing because this host-parasite combination occurs in the Pucciniastraceae but does not occur in species of Coleosporiaceae. Our study is a complement to the study of Aime and McTaggart (2020) by adding specimens of Pucciniastrum species on different Onagraceae hosts and specimens of species that are now placed in the genera Thekopsora, Quasipucciniastrum, and Calyptospora. ITS is a suitable marker to resolve terminal nodes. Thus, Pucciniastrum epilobii could be resolved to comprise the species P. epilobii restricted to a limited host spectrum (Chamaenerion angustifolium) and P. pustulatum with a wider host spectrum in the Onagraceae. By including more specimens of Thekopsora areolata and for the spec i e s k n o w n a s P u c c i n i a s t r u m a g r i m o n i a e a n d P. americanum, the separation of the two genera Quaspuccinastrum and Thekopsora was confirmed. This is the only major difference on a generic level to Aime and McTaggart (2020), where Pucciniastrum agrimoniae, P. americanum, and Thekopsora areolata are assigned to the genus Thekospsora.
The general difficulty of marker sequencing from rust vouchers is a further argument for concentrating on ITS rDNA in this study. We tried different markers (28S, nad6) but were not able to generate a complete set for all samples (unpublished data). In some cases, nad6 or 28S worked for a sample, but not ITS. Therefore, it was not possible to build a concatenated tree and it would have been necessary to build three separate phylogenies all comprising different sets of specimens. Because ITS worked for the majority of the samples, the much smaller trees with nad6 and 28S would have provided only limited additional information. Combining all the above listed arguments, we consider it suitable and sufficient to base all conclusions in the scope of this paper on an ITS phylogeny.

The genus Pucciniastrum
One basic issue has thus far been unresolved, viz., the phylogenetic circumscription of Pucciniastrum s. str., based on its type species (P. epilobii), which is essential for the further phylogenetic splitting of this complex. This problem has been addressed in the present work for European species.
Phylogenetic analysis of collections of P. epilobii (s. lat.) on various hosts revealed that this taxon is composed of two species. One is P. epilobii (s. str.) with Abies spp. as aecial hosts and Chamaenerion spp. as telial hosts. The other is P. pustulatum, a non-host-alternating, plurivorous species on Epilobium, Oenothera spp., and possibly on Chamaenerion and Clarkia usually without formation of telia. Whether Chamaenerion is indeed a host of P. pustulatum remains doubtful since the host Ch. angustifolium of the pertinent reference from Norway (DQ445906) (Hiatala et al. 2008) not exist (H. Solheim, pers. com.). Therefore, it is still not certain whether the host range of the two species overlaps. Differences in the phylogenetic placement were supported by differences in urediniospore morphology.
Our delimitation of Pucciniastrum from related genera remains provisional and only applies to European species and follows mainly morphological and life cycle criteria. Besides species of the Pucciniastrum epilobii complex, we maintain P. minima and P. guttata, formerly placed in Thekopsora, in Pucciniastrum because of similar morphological features. The taxa in the group consisting of Calyptospora and Melampsorella differ strongly from latter taxa in aeciospore and teliospore features and life cycle. Therefore, we decided to keep them separate from Pucciniastrum. However, this remains a provisional "European solution" as well. As mentioned above, additional species, particularly from Asia, must be stud ied for a futu re g ene ric c onc ep t w ithin P u c c i n i a s t r a c e a e . T h e r e f o r e , w e d i d n o t p l a c e P. hydrangeae-petiolaris and P. fagi in new genera, although our phylogeny (Fig. 1) and other phylogenies (e.g. Aime and McTaggart 2020) favour such a conclusion.

Rosaceae rusts in Coleosporiaceae
The hitherto sequenced species of Thekopsora belong to two families, namely Coleosporiaceae and Pucciniastraceae, as formerly shown by Aime and McTaggart (2020). Thekopsora areolata, the type species, is host-alternating between Picea [cones] (Pinaceae) and Prunus s. lat. (Rosaceae), a n d i s a m e m b e r o f t h e C o l e o s p o r i a c e a e . Our Coleosporiaceae ITS phylogeny differs from the 28S-18S-CO3 Coleosporiaceae phylogeny presented by Aime and McTaggart (2020) in one aspect. In our phylogeny, species Baden-Württemberg, Germany (project "Pilzflora Wilder See", grant to M. Scholler). The funders had no role in the design of the study, collection and interpretations of data, and in writing the manuscript.
Data availability Datasets generated during and/or analyzed during the current study are available on GenBank, from the corresponding author M. Scholler (morphological data) and Ben Bubner (lab protocols) upon request.

Declarations
Ethics approval Not applicable

Conflict of interest Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.