Introduction

Twenty-one species in nine teleomorphic genera and eight species in two anamorphic genera of rust fungi have been listed for plants of the Vitaceae (Fungal Databases, Systematic Mycology and Microbiology Laboratory, ARS, USDA; http://nt.ars-grin.gov/fungaldatabases/, accessed 25 Feb 2015). In the course of a taxonomic study on vitaceous phakopsoroid fungi (Phakopsoraceae, Pucciniales), telia were found among uredinia on the abaxial leaf surface in the type specimen of Uredo cissi-debilis Vienn.-Bourg. on Cissus debilis Planch. from Côte d’Ivoire (Viennot-Bourgin 1958). The telia were aparaphysate and consisted of shortly pedicellate teliospores composed of two to four thin-walled cells. Another specimen identified as U. cissi-debilis on Cayratia gracilis (Guill. & Perr) Suess. from Sierra Leone (unpublished) was also found to form two-celled or three-celled, thin-walled teliospores on a short pedicel in paraphysate telia formed among uredinia. The two fungal specimens were morphologically similar to Kuehneola vitis (E. J. Butler) P. Syd. & Syd. on Ampelocissus latifolia (Roxb.) Planch. distributed in South Asia (Butler 1912; Mundkur and Thirumalachar 1943).

This paper describes the morphology of the Cayratia and Cissus fungi and discusses their taxonomic distinctness from previously described rust species on vitaceous plants and the taxonomy of the genus Kuehneola and its allies.

Materials and methods

Specimens examined: listed under the name and description of each species. Roman numerals placed before a herbarium accession number denote the presence of uredinia (II) or telia (III) in a specimen.

Microscopic observation: Small sorus-bearing pieces were cut out from the herbarium specimens and they were thin-sectioned with a razor blade under a binocular dissecting microscope. Spores and paraphyses were scraped from sori. Thin sections, scraped spores and paraphyses were mounted on a microscopic slide and treated as described elsewhere (Ono 2000). The slide preparations were then examined both by bright-field and differential interference contrast microscopy (DIC) with an Olympus BH2 or Olympus BX51 microscope (Olympus, Tokyo, Japan) and measurements were made with an ocular micrometer. Twenty selected spores and paraphyses were measured for each specimen. To observe urediniospore germ pores, the spores were mounted in lactic acid on a slide glass and heated at boiling point for a few seconds, and a drop of lactophenol solution with aniline blue was then added onto the boiled spores.

Results

Key to Kuehneola species parasitic on Vitaceae plants

1. Paraphyses absent in sori; urediniospores subglobose, broadly obovoid or broadly ellipsoid, 15–20 × 10–15 μm in size; teliospores 2–4-celled, 25–39 × 9–13 μm in size…....… Kuehneola cissi-debilis (Vienn.-Bourg.) Y. Ono

1. Paraphyses present at periphery of sori ………………..... 2

2. Paraphyses cylindrical, thin-walled, 30–46 × 7–12 μm in size; urediniospores subglobose, broadly obovoid or broadly ellipsoid, 17–25 × 14–22 μm in size; teliospores 2–3-celled, 24–40 × 10–15 μm in size…..… Kuehneola deightonii Y. Ono

2. Paraphyses broadly cylindrical, incurved, dorsally thick-walled (−3 μm), 26–50 × 9–15 μm in size; urediniospores subglobose, obovoid or ellipsoid, 17–25 × 11–20 μm in size; teliospores 2–5-celled, 28–65 × 7–14 μm…........ Kuehneola vitis (E. J. Butler) P. Syd. & Syd.

Taxonomy

Kuehneola cissi-debilis (Vienn.-Bourg.) Y. Ono, comb. nov. Fig. 1a, b; Fig. 2a.

Fig. 1
figure 1

Kuehneola species on Vitaceae plants. a, K. cissi-debilis (holotype, PC0167202). urediniospores; b, K. cissi-debilis (PC0167202, holoype). teliospores; c, K. deightonii (PURF15960, holoype). paraphyses; d, K. deightonii (PURF15960, holoype). urediniospores; e, K. deightonii (PURF15960, holoype). teliospores; f, K. vitis (PUR68020, isotype). paraphyses; g, K. vitis (IBAR7782). urediniospores; h, K. vitis (PURF15960, isotype). teliospores. Bars a–h 20 μm

Full size image
Fig. 2
figure 2

Pedicellate teliospores of Kuehneola species on Vitaceae plants. a, K. cissi-debilis (PC050286, holotype); b, K. deightonii (PURF15960, holotype); c, K. vitis (PUR68020, isotype). Pedicels are indicated by asterisks (*). Bars a–c 20 μm

Full size image

MycoBank no.: MB811793

Basionym: Uredo cissi-debilis Vienn.-Bourg. Uredineana 5: 230, 1958.

Spermogonia and aecia: not observed. Uredinia: formed on abaxial leaf surface, solitary to densely grouped, subepidermal in origin, soon becoming erumpent surrounded by torn epidermis, aparaphysate. Urediniospores: formed singly on a short pedicel, appearing almost sessile, mostly subglobose, broadly obovoid or broadly ellipsoid, 15–20 × 10–15 μm in size (Fig. 1a); the wall thin, colorless, echinulate with no apparent germ pore. Telia: formed on abaxial leaf surface, replacing uredinia, densely grouped, slightly raised, waxy, becoming reddish or dark brown, aparaphysate. Teliospores: composed of 2–4 linearly arranged cells formed on a short pedicel (Figs. 1b and 2a); two-celled spores 25–32 μm long; three-celled spores 28–38 μm long; four-celled spores 34–39 μm long; each cell broadly ellipsoid, 12–17 × 9–13 μm in size; the wall thin and colorless; metabasidia and basidiospores not observed.

Holotype: on Cissus debilis Planch. (Vitaceae): République de Côte d’Ivoire, near Bongouanou, August–September 1951, G. Viennot-Bourgin (II + III, PC0167202).

Notes: The type specimen bore abundant telia among uredinia on the abaxial leaf surface. Many uredinia were also replaced by telia. Being covered by urediniospores, the tiny telia were easily overlooked even though they were slightly raised and tinted differently from the host epidermal cells. Viennot-Bourgin (1958) described three to five obscure germ pores scattered on the urediniospore wall.

Kuehneola deightonii Y. Ono, sp. nov. Fig. 1c–e; Fig. 2b.

MycoBank no.: MB811794

Differs from K. cissi-debilis in forming paraphysate sori and from K. vitis in forming evenly thin-walled paraphyses and shorter teliospore.

Spermogonia and aecia: not observed. Uredinia: formed on abaxial leaf surface, scattered or loosely grouped, subepidermal in origin, soon becoming erumpent, peripherally paraphysate. Paraphyses: cylindrical, straight or weakly incurved, basally united, evenly thin-walled, colorless, 30–46 μm high, 7–12 μm wide (Fig. 1c). Urediniospores: formed singly on a short pedicel, appearing almost sessile, mostly subglobose or broadly ellipsoid, obovoid, 17–25 × 14–22 μm in size (Fig. 1d); the wall thin, colorless, echinulate, with no apparent germ pore. Telia: formed on abaxial leaf surface, white, cottony, peripherally paraphysate. Paraphyses: as in uredinia. Teliospores composed of 2–3 linearly arranged cells formed on a short pedicel (Fig. 2b); two-celled spores 24–33 μm long; three-celled spores 34–40 μm long; each cell broadly ellipsoid, 10–18 (−20) × 10–15 μm in size (Figs. 1e and 2b); the wall thin, colorless; germinating in situ to form four basidiospores on a four-celled metabasidium. Basidiospores: obovoid, 6–9 × 5–7 μm in size, the wall thin, colorless.

Holotype: on Cayratia gracilis (Guill. & Perr) Suess. (originally identified as Cissus gracilis Guill. & Perr.) (Vitaceae): Sierra Leone, Bonjema, Kori, 29 October 1954, F. C. Deighton (II + III, PURF15960); Isotype (II + III, IMI58574a).

Etymology: from F. C. Deighton, a dedicated mycologist on African fungi.

Kuehneola vitis (E. J. Butler) P. Syd. & Syd., Monogr. Ured. 3: 321, 1914. Fig. 1f–h; Fig. 2c.

≡ Chrysomyxa vitis E. J. Butler, Ann. Mycol. 10: 158, 1912.

≡ Cerotelium vitis (E. J. Butler) Arthur, Bull. Torrey Bot. Club 44: 509, 1917.

≡ Catenulopsora vitis (E. J. Butler) Mundk. & Thirum. Ann. Bot. 7: 218, 1943.

Spermogonia and aecia: not observed. Uredinia: minute, scattered or loosely grouped on abaxial leaf surface, subepidermal in origin, soon becoming erumpent, peripherally paraphysate. Paraphyses: broadly cylindrical, moderately incurved, basally united, dorsally thick-walled (up to 3 μm thick), colorless, 26–50 μm high, 9–15 μm wide (Fig. 1f). Urediniospores: formed singly on a short pedicel, appearing almost sessile, mostly subglobose or broadly ellipsoid, obovoid, 17–25 × 11–20 μm in size (Fig. 1g); the wall ca. 2 μm thick, colorless, echinulate with no apparent germ pore. Telia: minute, loosely grouped on abaxial leaf surface, often replacing uredinia, white, cottony, peripherally paraphysate; paraphyses as in uredinia. Teliospores: composed of 2–5 linearly arranged cells formed on a short pedicel (Figs. 1h and 2c); two-celled spores 28–36 μm long; three-celled spores 33–53 μm long, four- or five-celled spores up to 65 μm long; each cell broadly ellipsoid, 10–19 × 7–14 μm in size; the wall thin and colorless; germinating in situ to form four basidiospores on a four-celled metabasidium. Basidiospores: obovoid, 6–9 × 5–7 μm in size; the wall thin, colorless.

Holotype: on Ampelocissus latifolia (Roxb.) Planch. (as “Vitis latifolia” in the original description), Bangladesh, Rangspur (as “Rangspur, Eastern Bengal” in the original description), 18 October 1909, S. N. Mitra (II + III, Herbarium Cryptogamae Indiae Orientalis, Delhi, India); Isotype in the Arthur Fungarium, Purdue University, USA (II + III, PUR68020).

Other specimen examined: on A. latifolia, India, Maharashtra, Ratnagiri, Dapoli, II, 14 October 1996, M. S. Patil (II, IBAR7782).

Notes: Describing K. vitis, Butler (1912) did not designate a holotype, but only mentioned “In foliis Vitis latifoliae, Dacca (A. L. Som) et Rangpur (S. N. Mitra), Indiae or.” Mundkur and Thirumalachar (1943) designated the holotype as above and specified another specimen as “Dacca, October 7, 1910, leg. A. L. Som.” They indicated the presence of spermogonia on the holotype specimen and a possible demicyclic life cycle of this species. Although Ragunathan and Ramakrishnan (1973) examined the holotype, they did not notice spermogonia on it. The isotype and an additional specimen examined in this study bore no structure like spermogonia.

Mundkur and Thirumalachar (1943) included two specimens on Cissus adnata Roxb. collected by E. J. Butler and Inayat in Noakhali, southeastern Bangladesh. Detailed observation was lacking for these specimens. Because these specimens were not available for the study, their taxonomic identity remains unsolved. Ragunathan and Ramakrishnan (1973) listed three specimens of V. vinifera L. as the host of this species. Because these specimens were also not available for study, their taxonomic identity was not determined. Unless the presence of teliospores is confirmed, the Southeast Asian Phakopsora species on V. vinifera (Pota et al. 2015) is easily mistaken as K. vitis.

Discussion

Distinction between Kuehneola, Phragmidiella, Catenulopsora and Cerotelium

Rust species forming uniseriate multiple cellular teliospores, which are thin-walled and laterally not adherent, are classified in Chrysomyxa Unger (Coleosporiaceae), Cerotelium Arthur, Phragmidiella Henn. (Phakopsoraceae) or Kuehneola Magnus (Phragmidiaceae) (Dietel 1928; Cummins 1959; Cummins and Hiratsuka 2003; Buriticá 1999). The family and genus assignment of a rust fungus forming this kind of teliospores is supplemented by the morphological types of spermogonium, aecium and uredininum, which the fungus produces during its life cycle. The host relationships in the life cycle are also considered to be important in some rust families and genera. If only telia and uredinia are known for a rust fungus to be classified, assignment of the species to a family and a genus is often arbitrary, except for Chrysomyxa, whose caeomatoid uredinia and the specific host relationships easily separate it from its morphological allies. This ambiguity of the generic classification is primarily due to different interpretations of the cell basal to a uniseriately arranged multiple cell. The basal cell may be interpreted as a meristematic cell from which one-celled teliospores are formed in basipetal succession; contrarily, it may be viewed as a pedicel on which an uniseriate, multicellular teliospore is formed by transverse septation. The two or more transversely septate cells on a pedicel are considered as integral parts of a single teliospore. This kind of teliospore was the characteristic of the genera Kuehneola (the type: K. albida (J.G. Kühn) Magnus on Rubus, Rosaceae) and Phragmidiella (the type: P. markhamiae Henn. on Markhamia, Bignoniaceae) (Hennings 1905). The two genera were distinguished by the presence (in Kuehneola) or absence (in Phragmidiella) of paraphyses in uredinia (Hennings 1905). It was also stressed that none of the species of Kuehneola occurred on the Bignoniaceae.

Subsequently, “pedicels” of teliospores of Kuehneola species became considered as meristematic sporogenous cells. Dietel (1912) said, “the so-called teliospores of Kuehneola are spore-chains, series of one-celled single spores, which are successively abstricted one after the other from the apex of common hypha, and remain united fast with one another” (cited from Arthur 1917). He restricted only species on rosaceous plants to accommodate in the genus and classified species of Kuehneola on non-rosaceous plants and of Phragmidiella in Cerotelium (Dietel 1928). Sydow and Sydow (1914) also defined Kuehneola as forming one-celled teliospores in elongated chains splitting apart upon maturity, and synonymized Phragmidiella under Kuehneola. They classified in the genus those species that occurred on non-rosaceous plants and which had been classified both in Phragmiella and Kuehneola. For them, taxonomic close host relationships were not important in defining rust genera.

Catenulopsora Mundk. was elected based on the type C. flacourtiae Mundk. & Thirum. on Flacourtia (Flacourtiaceae) (Mundkur and Thirumalachar 1943). In the type species, a laterally free, uniseriate multicellular teliospore developed from a teliospore initial arisen from a telial hymenium, from which a teliospore mother cell and a pedicel cell differentiated. Subsequently, the teliospore mother cell elongated and became transeversly septate to form a multicellular mature teliospore. Teliospore germination by continuous apical elongation was said to be unique to the genus (Mundkur and Thirumalachar 1943). However, teliospore germination (basidium production) by continuous apical elongation of teliospore cells was proven to be common in Kuehneola. Therefore, Catenulopsora was synonymized under Kuehneola (Thirumalachar 1960), despite Thirumalachar and Mundkur (1949) again characterizing Catenulopsora by pedicellate multicellular teliospores and Kuehneola by sessile catenate teliospores.

Cummins (1959) explicitly defined Kuehneola as forming pedicellate, transversely multi-septate teliospores and Phragmidiella as forming one-celled, sessile and catenate teliospores. The same taxonomic treatment was followed by Buriticá (1999) and Cummins and Hiratsuka (2003), except for Catenulopsora, i.e., being treated as distinct in the former and synonymized under Kuehneola by the latter authors. The ontogeny of teliospores in species of Kuehneola and Phragmidiella has not been precisely determined, except for K. flacourtiae (Mundk. & Thirum.) Thirum.. (= Catenulopsora flacourtiae Mundk. & Thirum.). Photomicrographs in some published studies on Kuehneola species (Bagyanarayana and Rao 1985, Fig. 2; Ono 2012, Fig. 2) and in this study (Fig. 2), however, showed that a cell basal to linearly arranged cells in telia was narrower than, and firmly attached to, upper cells. This basal cell is morphologically comparable to the teliospore pedicel of Hamaspora (Monoson 1969; Cummins and Hiratsuka 2003) or Fromeëlla (McCain and Hennen 1990; Cummins and Hiratsuka 2003). Therefore, the generic definitions of Kuehneola and Phragmidiella by Cummins and Hiratsuka (2003) is followed in this study.

Species of Cerotelium form one-celled teliospores successively from the basal sporogenous cells on a telial hymenium as in Phragmidiella. Unlike Phragmidiella species, however, the teliospore chains of Cerotelium species are compacted, even if laterally not adherent, and the upper part of teliospores becomes easily separated and irregularly arranged upon maturity (Arthur 1917; Ono et al. 1992). In addition to the telial characteristics, association of Milesia-type, Malupa-type or Physopella-type of uredinia with telia makes the genus distict from Phragmidiella (Ono et al. 1992; Buriticá 1999).

Taxonomic changes of Chrysomyxa vitis on Ampelocissus latifolia

A rust fungus on A. latifolia was orignally described under the name of Chrysomyxa vitis E. J. Butler. Two to four thin-walled teliospores were arranged in a uniseriate laterally free column and echinulate urediniospores appeared to be formed in chains (Butler 1912). Butler was “unable to find any character which separate this fungus from the genus Chrysomyxa,” even though he was well aware of the host difference between this species on the Vitaceae and other Chrysomyxa species occurring on Ericaceae and Pinaceae. Due to old and collapsed sori on the specimen he examined, however, Butler (1912) was unable to determine that the urediniospores were formed singly on a short pedicel. For similar reasons, Kühn (1883) might have erroneously described K. uredinis (Link) Arthur in the genus Chrysomyxa as C. albida J.G. Kühn.

Because of the changes in generic definitions due to different interpretations of teliospore ontogeny in Kuehneola, Phragmidiella, Catenulopsora and Cerotelium as briefly described above, C. vitis was classified differently later on. Sydow and Sydow (1914) classified the fungus on A. latifolia in the genus Kuehneola, as K. vitis (E. J. Butler) P. Syd. & Syd., a decision well supported by Butler’s (1912) description and illustration. Arthur (1917) affirmed that the teliospores of K. vitis and Kuehneola fici E. J. Butler (Butler 1914) were produced in the same way as in Cerotelium gossypii (Lagerh.) Arthur, in which species one-celled teliospores successively formed from basal sporogenous cells were laterally adherent (Arthur 1917), and therefore transferred both species to Cerotelium as well. This was in contradiction to the descriptions and illustration in Butler (1912) that clearly indicated what Arthur (1917) circumscribed for Kuehneola. On the other hand, Mundkur and Thirumalachar (1943), when establishing Catenulopsora, determined that K. vitis formed a laterally free, uniseriate multicellular teliospore by transverse septation on a pedicel and transferred it to Catenulopsora. This taxonomic treatment was soon replaced by the re-classification of C. vitis in Kuehneola with nomenclatural revision as K. vitis, which name is accepted herein.

Morphology-based distinction of the rust genera and classification of species may not reflect their history of evolutionary diversification (Aime 2006). The presence or absence of teliospore pedicels and host restriction may have different meanings in different rust genera and families. Molecular systematic analyses (Maier et al. 2003; Wingfield et al. 2004; Aime 2006) indicated that the family Phragmidiaceae to which Kuehneola belongs is a well-circumscribed monophyletic taxon, most, if not all, species occurring on rosaceous plants. If Kuehneola were proven to be restricted to the species parasitic only on the plants of Rosaceae, other species on non-rosaceous plants would be classified in Catenulopsora, or even in Phragmidiella or Cerotelium with the generic definitions revised. Molecular phylogenetic analyses at a generic level with a number of species representing Kuehneola, Phragmidiella, and Cerotelium species incorporated will resolve the current difficulty in the taxonomy of the four genera and their possible allies, such as Phakopsora, Mehtamyces and Newinia.