Mycol. Res. 97 (4), 433-448 (1993) Pri/l/ed ill Great Bnfmll A re-evaluation of Cylindrocladiella, and a comparison with morphologically similar genera P. W. CROUS Department of Plant Pathology, University of Stellenbosch, 7600 Stellenbosch. South Africa M.J. WINGFIELD Deparhnent of Microbiology and Biochemistry, p.o. Box 339, University of the Orange Free State, 9300 Bloemfontein, South Africa Cylindrocladielln is confirmed as distinct from Cylindrocladium. Species of Cylindrocladium are primarily distinguished from Cylindroclndiella by septate stipes, more numerously branched conidiophores, only penicillate conidiophores and Calonecfria teleomorphs. Cylindrocladiella, hmvever, has non.septate stipes, small I-septate conidia, penicillate and subverticil1ate conidiophores, chlamydospores that are arranged in chains, phialides with extending, prominent coliareHes, conidia in slimy masses (as opposed to clusters), strong cultural odours and Nutria teleomorphs. Six Cylindrocladiella spp. are recognized. C. peruviana is placed in synonymy with C. wmelliae, while C. eIegans and C. lagenifonnis are described as new. Gliocladiopsis is distinguished from Cylindrocladiella by the absence of a stipe, the presence of more numerous conidiophore branches and the fonnation of conidia in slimy yeJlow masses. Gliocladiopsis is distinguished from Cylindrocarpol1 by having uniformly cylindrical conidia, more numerous conidiophore branches, and more numerous phialides per branch. Cyliruirocarpon le11l~eis shown to be beHer accommodated in Gliocladiopsis, and the name G. tenuis proposed. The taxonomic position of Acontiopsis is uncertain. due to its vague generic description and the absence of type material. The genus Cylindrocladiwll was established for the single Species of Cylilldrocladiella are frequently associaled with species C. scopariu»1 Morgan (1892). The primary disting- leaf spot and root rot symptoms of various hosts (Peerally, uishing charaderistic was the presence of cylindrical septate 1974; Boesewinkel. 1982; Sharma & Mohanan, 1982). conidia. The stipe extension above the conidiogenous Although there are conflicting reports regarding the patho- apparatus (referred to here as the stipe) was not mentioned in genicity of C. parva (Anderson) Boesewinkel (Sharma & the diagnosis, but its existence became knov.;n later (Anderson, Mohanan, 1982; Sobers & Alfieri. 1982; Crous, Phillips & 1918; Boedijn & Reitsma, 1950). Wingfield, 1992 b), other species in the genus have been In the first review of Cylindrocladiwl1, Candelospora Hav,..Iey shown to be pathogenic to various hosts «(rous et af., 1992b), apud Rea & Hav,'ley (1912) v,'as treated as a synonym, and all and some have been reported to cause significant damage to relevant new combinations made (Boedijn & Rietsma, 1950; economic plants (Mohanan & Sharma, 1985; Peer ally, 1991). Tubaki, 1958). Subsequently. Telracyllim Vanderwalle (1945) The stalus of Cylindrocladiella as a genus has recently been was also recognized as a synonym of Cylilldrocladiwtl contested (Peerally, 1991; Shanna & Mohanan, 1991). This (Subramanian, 1971; Domsch, Gams & Anderson, 1980; study was undertaken to consider the distinction between Brayford & Chapman. 1987). Cylindrocladiella and other morphologically similar genera Boesewinkel (I982) believed that several species of such as Cyli/1drocladiu/11(Peerally, 1991), Acontiopsis t\egru Cylindrocladium with small conidia were distinct from those (Kendrick & Carmichael. 1973; Peerally. 1991). Cylindrocarpon with larger conidia and septate stipes. On this basis he Wollenw. (syn. Moes:ia Bubak) (Tubaki, 1958; von Arx, 1970) transfeITed the names of these fungi to a new genus, and GliocladiopsisSaksena (Agnihothrudu, 1959; BaITon, 1968). Cylindrocladie/IaBoesewinkel. Species of Cylirzdrocladiellawere also distinguished from Cylindrocladium by cultural criteria, chlamydospore aITangement, the presence of penicillate as MATERIALS AND METHODS well as subverticillate conidiophores, Nectria (Fr.) Fr. as Evaluation of anamorphs opposed to Calonectria De Not. teleomorphs, non-septate, unbranched, central stipes and obvious collarettes (sensu Cultures derived from single conidia were plated on to Sutton, 1980) on the phialides, conidial arrangement on carnation-leaf agar (CLA) (Fisher el al.. 1982; Crous, Phillips & conidiophores, as well as the small. 1.septate conidia Wingfield, 1992 a), incubated at 25°C under nuv, and (Boesewinkel. 1982). examined after 7 d. Only material occurring on the leaves was ~YC9i A re-evaluation of Cylindroclndielln 434 examined. Mounts were prepared in lactopheno! cotton blue. Colours were taken from Methuen (Kornerup & Wanscher, All measurements vvere made under the (100 x) oil-immersion 1967) and Rayner (1970). objective. Wherever possible, fifty examples of each structure were measured. The data were analysed (Table 1) by Least Significant Difference (L.S.D.) at a probability level of P = 0'05. RESUL TS Where less mdterial was available, or when it was of insufficient quality, averages are not included. Cultures and Evaluation of anamorpl1s types are lodged at the National Collection of Fungi, Pretoria Six vesicle classes were recognized for CylindrocIadieIla spp.: (PREM). irregularly lanceolate, ellipsoid to lanceolate, ellipsoid to clavate, cylindrical, clavate to pyriform and lageniforrn to Vesicle shape and stipe lengtl1, Vesicle shape was ovoid (Figs 1-6). Although C. parva and C. novae-zelandiae detennined on CLA after 7 d incubation. Vesicles examined (Boesew.) Boesew. were both characterized by penicillate were all on stipes of conidiophores, with at least one primary conidiophores without obvious collarettes on dolijform to and one secondary branch bearing phialides. VesicIes that cyrnbifonn phialides, stipe lengths of C. parva were sig- showed signs of proliferation were ignored. Vesicle width was nificantly shorter (P measured at the widest point, and stipe length measured from = 0'05) than those of C. novae-zelandiae (Table 1).C.elegans sp.nov.,C.j11festans Boesew.and C. novae- the basal septum to the vesicle tip. zelandiae had very similarconidial dimensions when examined on CLA, making it difficult to distinguish these species solely Collarettes and conidiophores. Conidiophores were exam- on this basis. Similar conidial dimensions were also observed ined after 7 d on CLA to detennine if they were either for C. camelliae (Venkataramani& Venkata Ram) Boesew. and peniciIJate or subverticillate, or if both conidiophore types Clagenifonnis sp. novo (Table 1). In contrast to Cyli11drocladium, were present. Phialides were also examined for the presence or stipes were nearly always fanned at the centre of the absence of collarettes. conidiophore, and were never branched. Cultural characteristics Cultural characteristics Growth studies. To determine the maximum radial growth of Cylindroe!adiellaspp. could easily be distinguished from each species in culture, agar plugs from the periphery of 7-d-old other by their minimum and maximum temperature require- colonies (3 mm diam.) of each fungus were plated at the centre ments for growth. The ability to grow at either very high or of malt-extract agar (MEA) (20 g Oxoid malt extract, 15 g low temperatures enabled species to be grouped into either Difco agar, 1000 ml H20) plates, and incubated at 250 for 1 d high (growing above 30°) or low (growing below 10°) to ensure active growth. Growth after 1 d was marked, and temperature classes (Figs 7-12). Cylindrocladiella spp. could thereafter plates were placed at the respective temperatures also easily be distinguished in the same way from Gliocladiopsis under consideration. Growth was assessed for all isolates after sagarimsis Saksena, which was capabJe of growing at much 6 d of incubation in the dark at 5, 8, 10, 15,20,25,33 and 35°, higher temperatures than any species of Cylindrocladiella. with three replicate plates of each isolate at each temperature. As reported by Boesewinkel (1981, 1982), chlamydospore Radial growth was also determined after 3 d at 25 and 30°. fonnation can be used to further distinguish Cyli11droc/adiel1a Average growth was calculated from four radial measurements spp. from each other (Table 1).Some degree of variation was, from each of the three plates. The experiment was repeated at. however. found in different isolates of certain species, thus least once for all isolates. The rationale for including 8 and 33° reducing the value of this criterion. After 6 d on MEA, colony with the other 5° intervals was that the growth of many pigmentation in all spp. except C. lageniformiswas chiefly the species began to slow or stop between 5 and 10 and between result of chlamydospore production. In C. lagmiformis, 30 and 35°. These temperatures were therefore necessary to however, a reddish pigment diffused readily into the medium, detennine a finer distinction between minimum and maximum resulting in a darker colony. temperature requirements for growth. Chlamydospores and colony colour. Chlamydospore TAXONOMY measurements were found to be variable, and were therefore ignored. Production of chlamydospores and microsclerotia Cylindrocladiella Boesewinkel, Call.]. BDl. 60, 2289 (1982). was, however, rated for the extent of thickened, pigmented Conidiophores hyaline, single, subverticillate, as well as hyphae present after 6 d (viewed from the underside of penicillate, with primary and secondary branches. Phialides plates), at the completion of the growth studies. Dark brown, terminal, hyaline, in whorls of 2-4, with or without