Shoemaker et al.: Rhexocercosporidium carotae 362

Rhexocercosporidium carotae (Årsvoll) U. Braun, Mycotaxon, 51: 65. 1994.� Acrothecium carotae Årsvoll, Acta Agric. Scand. 15(2): 112–113. 1965.� Pseudocercosporidium carotae (Årsvoll) de Hoog and Oorschot, Stud. Mycol. 26: 103. 1985.

LESIONS dark brown to black, moist, mostly circular spots 0.5 to 1.2 cm diam. that extend only to 2 mm in the root, occurhaphazardly, and number 20 to 25 per carrot root. CONIDIOPHORES formed in culture at 25°C are mycelioid or lack-ing, unbranched or branched sympodially, hyaline. CONIDIOGENOUS CELL terminal or lateral, or intercalary in myce-lium, clavate, flask-shaped, or cylindrical, hyaline, 20–30 × 3–6 µm, connected to the conidium by a short isthmus,finally with a truncate apical scar 2 µm wide bearing a thin frill. CONIDIA formed in culture at 25°C are initiallyobovoid, but extend and become septate and more or less cylindrical, straight, smooth, often rounded at the apex andsometimes slightly wider near the apex, (1)3(5)-euseptate, hyaline, 25–45 × 5–6.5 µm, with granular contents, withoutguttules, basally obconic, with a slightly protruding, truncate scar, 2 µm wide, with a faint marginal frill. Germination oc-curs promptly from any cell. CULTURES after 11 days at about 25°C exposed to daylight, the colony on potato dextroseagar plus antibiotics (tetracycline hydrochloride 0.05 g/L plus streptomycin sulphate 0.1 g/L) was appressed, white,10 mm diam., with some hyaline conidia. On oatmeal agar, under the same conditions, colony was appressed, 13 mmdiam., with yellow pigment in submerged mycelium, with a pale yellow to olive surface overlaid with white conidia. On

Can. J. Plant Pathol. 24: 359–362 (2002) 359

Fungi Canadenses No. 344 RHEXOCERCOSPORIDIUM CAROTAE

Figs. 1–9. Rhexocercosporidium carotae. (1) Conidia, conidiogenous cell and isthmus (×450); (2) conidium initial (×1100); (3–6)conidia (×1100); (7–9) symptoms on carrot roots.

Synthetischer nährstoffarmer agar with added filter paper (Nirenberg. 1976. Mitt. Biol. Bundesanst. Land. Forstwirtsch.169: 1–117) the fungus produced a pale yellow colony 5 mm diam. on the filter paper under the same conditions.

DISTRIBUTION: Quebec, also in Norway and the Netherlands and possibly in the United States (New York).

COLLECTIONS: DAOM 226960, isolated from roots of Daucus carota L., grown and stored at Île-d’Orléans, Quebec, coll.Jean Coulombe, isol. Mario Tesolin (99-4845), comm. Michel Lacroix, det. R.A. Shoemaker (2000-M21), 23 March2000. The diseased carrots had been preserved in cold storage at Île-d’Orléans. The storage temperatures ranged from1°C to 6°C, with the initial temperature about 5°C during the first two weeks, then gradually decreasing. The relative hu-midity was 90–95% except near the door and evaporator coils, where some root desiccation occurred. No gasses wereadded to the storage atmosphere. The crop was produced in mineral soil. The previous crop was rutabaga preceded by al-falfa.

NOTES: The fungus is thoroughly described by Årsvoll (1965. Acta Agric. Scand. 15: 101–114) with additional informationon environmental influences published later (Årsvoll. 1971. Acta Agric. Scand. 21: 3–10). From the works by Årsvoll, itis known that conidium length and septation is influenced by temperature. Mean length is greatest near 3°C (40 µm) anddecreases linearly with temperature to 25°C (17 µm). Conidia are equally 3- and 5-septate from 0°C to 6°C; predomi-nantly 3-septate at 18°C and equally 0- and 3-septate at 25°C. Conidium width was independent of changes in tempera-ture, 5–6.5 µm. Growth on 2% malt agar at the optimum temperature (18°C) measured about 24 mm in 11 days or about60 mm in 28 days. At 25°C, nearly the maximum temperature, colony diameter was reduced to 20 mm in 28 days orabout 8 mm in 11 days, which is more or less in agreement with our results given above. In the Norwegian study the dis-ease was rarely found on roots at harvest time (October) but was observed on a late-harvested crop (December). Rot de-velopment in storage was more rapid at temperatures above 0°C. Leaf spot and petiole lesions were observed and so wasdamping-off of seedlings. The host range seems limited to Daucus carota. Seven umbellifers tested with foliar spore ap-plications did not develop leaf spot symptoms. More intense application of inoculum to hypocotyls induced symptoms ina number of umbellifers, especially Daucus muricatus L.

The fungus grows from –3°C to 25°C, with an optimum near 18°C. Conidium length is greatest at –1°C to 3°C. Thefungus sporulates on root, leaf, and petiole lesions. In vitro sporulation occurs from pH 3.1 to 7.6 with the optimum from4.7 to 5.2. Continuous light is best for in vitro sporulation. Spore germination in vitro requires free water. The pH forgrowth, sporulation, and spore germination ranges from 3.4 to 7.1 (optimum 4.2–4.7). This affinity for acidic conditionsmay explain the concentration of this disease in the rather acidic soils of southwestern Norway. Germination occurredfrom –3°C to 28°C, was most rapid (4 h) at 12–28°C, but the percent germination decreased beyond 18°C. Spore germi-nation was inhibited in nonsterile soil but not in similar soil that had been steam sterilized. Spore germination innonsterile soil was proportional to the proximity to carrot rootlets. Conidia on leaflets produced a germ tube, anappressorium, an infection peg into an epidermal cell, and then intracellular and intercellular mycelium. Conidia buriedin nonsterile soil remained viable at 0°C and 3°C for 12 months, at 6°C for 6 months, and at 12°C for 2 months, indicat-ing a capacity to overwinter in cold regions.

A study by de Hoog (1985. Stud. Mycol. 26: 1–122) based on the original culture deposited at CBS by Årsvoll (CBS418.65) revealed the presence of a frill on the conidium scar and the conidiogenous cell. These authors suggested thespecies be accommodated in Pseudocercosporidium Deighton. After another study of the type culture, Braun (1994.Mycotaxon, 51: 37–68) presented arguments against that choice of genus and, instead, proposed a new genus,Rhexocercosporium, based on this species. He emphasized the rhexolytic secession of the conidia and illustrated the frillon conidia and, with less certain definition, on the conidiogenous cell.

The genus Acrothecium (Corda) G.T. Preuss in which the carrot fungus was originally placed was considered by Braunto be a nomen dubium. Braun (1994. Mycotaxon, 51: 37–68) cited Hughes (1958. Can. J. Bot. 36: 727–836) to supportthis contention. Hughes (1958) did not list Acrothecium as a nomen dubium but questioned the identity of what he tookto be the type species, A. multisporum Preuss 1851, for which he had not seen type material. Hughes (1958) did refer, aswell, to Trichothecium Link subg. Acrothecium Corda 1838 typified by Trichothecium parasitans Corda 1838. This cer-tainly brings into question the typification of Acrothecium. Index Nominum Genericorum lists T. parasitans as the typeof Acrothecium. This suggests that Preuss raised subgenus Acrothecium Corda to generic rank, where it would be cited asAcrothecium (Corda) G.T. Preuss 1851. Acrothecium multisporum Preuss 1851 was the only species treated. Later, T.parasitans was transferred to Acrothecium as A. parasitans (Corda) Corda ex Streinz, Nomencl. Fungorum, p. 3. 1862.

The fact that the carrot pathogen now has its own genus, Rhexocercosporidium, is a temporary solution that avoids thequestion of the affinities of this fungus. The fungus has such delicate, simple morphological features that it is not easy toplace it at the generic level and may have been reported under more than one generic name.

DNA sequences of portions of the nuclear rRNA gene were determined for DAOM 226960 and CBS 418.65 afterDNA extraction from pure cultures (FastDNA® Kit, BIO 101, Inc., Carlsbad, Calif.) and using PCR amplification and au-tomated sequencing methods described elsewhere (Hambleton et al. 1998. Mycologia, 90(5): 854–869). Sequence datafor both DNA strands were obtained for the complete ITS region, 560 base pairs (bp) spanning the ITS1, 5.8S, and ITS2,as well as 1375 bp of the small subunit (18S) starting from the conserved NS1 priming site (White et al. 1990. In PCRProtocols: a Guide to Methods and Applications. Academic Press, N.Y. pp. 315–322). The presence of at least one largeintron in the amplicons of each strain terminated efforts to obtain more complete 18S sequence data. GenBank accession

360 Can. J. Plant Pathol. Vol. 24, 2002

Shoemaker et al.: Rhexocercosporidium carotae 361

Fig. 10. One of 18 single most parsimonious trees resulting from a maximum parsimony analysis of the small subunit rDNA datamatix. Bootstrap support values over 50% are indicated adjacent to the relevant node. All branches were present in the strict consensusof all 18 trees except for three within the Leotiomycetes clade, represented by narrow lines. Classification follows O.E. Eriksson et al.(2001. http://www.umu.se/myconet/Myconet.html).

numbers for the sequences are as follows: AF487894 (DAOM 226960, ITS); AF487895 (CBS 418.65, ITS); AF487896(DAOM 226960, 18S); AF487897 (CBS 418.65, 18S). The GenBank BLAST Search function was used to retrieve theclosest 18S and ITS sequence matches.

To evaluate the phylogenetic placement of R. carotae, small subunit sequences were aligned in a data matrix with 48sequences from GenBank representative of a diversity of Ascomycetes, then subjected to parsimony analysis usingPAUP* version 4.0b8 (Swofford. 2001. Sinauer Associates, Sunderland, Mass.). An analysis using the heuristic searchalgorithm (default settings) resulted in 18 equally parsimonious trees (MPTs) of 1756 steps and with a consistency indexof 0.506. A bootstrap analysis (1000 replicates) using the “fast” stepwise-addition search option was performed to deter-mine the level of support for the topology of the inferred phylogeny.

ITS sequences were identical for both strains, as were the 18S sequences, confirming the identification of DAOM226960 as R. carotae and providing molecular support for the taxonomic discussion presented here. The closest ITS se-quence matches in GenBank were for a number of fungal isolates from oat roots (AJ246140–AJ246144) and severalstrains of Leptodontidium orchidicola isolated from roots of Ericaceae (AF214576–AF214578). These matches were notoverly informative; the taxa are either unidentified or of uncertain phylogenetic affinity. Also, alignment of the first 45bp was ambiguous and sequence identities were no more than 92% for the ITS1 spacer (235 bp) alone and 97% for ITS2(149 bp).

Parsimony analysis of small subunit sequences indicated that R. carotae is related to the inoperculate discomycetes(Fig. 10), clustering in the same clade as Sclerotinia, Rhytisma, Leotia, Spathularia, and Microglossum (Leotiomycetes).While there was no bootstrap support for this outcome, this clade was present in the strict consensus of all 18 MPTs. Aswell, R. carotae was clearly excluded from the other ascomycete groups sampled, based on the strong bootstrap supportfor other clades in the phylogram. These results provide a phylogenetic framework for any further studies of the carrotpathogen, and increase the likelihood that the teleomorph and closer relatives of R. carotae may be discovered. Some un-expected fungi, such as members of the Erysiphales (Saenz et al. 1994. Mycologia, 86(2): 212–216), thePseudeurotiaceae (Suh and Blackwell. 1999. Mycologia, 91(5): 836–848), and the Myxotrichaceae (Sugiyama et al.1999. Mycoscience, 40: 251–258), have been shown previously to group in this clade. These taxa were among those re-trieved by the BLAST search of GenBank sequences. Species representing two of these groups cluster together in thestrict consensus with Sclerotinia sclerotiorum, R. carotae, and two unidentified fungi isolated from roots (AF168167 andAB016175). In general, relationships among discomycetous taxa are not well resolved above the family level in phylo-genetic hypotheses derived from 18S sequence data. This is due not only to the low level of phylogenetic signal in thesequence data for the deeper branches of the tree but also to the paucity of available sequences for the apothecia-formingfungi and our limited understanding of the diversity of morphologically dissimilar species that are phylogenetically re-lated to them.

The fungus was first described from Norway. It apparently occurs in the Netherlands because in the CBS catalogue(cbs.knaw.nl) there are two cultures (additional to the type), 331.86 and 415.88, originating from stored roots of carrot atWageningen. As far as we can tell, this is the first report of this fungus in North America. However, the symptomsstrongly resemble the colour illustrations of an undescribed brown side rot published by W.E. Rader (1952. Diseases ofstored carrots in New York State. Cornell Univ. Agric. Exp. Stn. Bull. 889. Plate II, Bottom right). It is hoped that aredescription with illustrations from fresh material may aid in future identifications.

R.A. Shoemaker and Sarah HambletonEastern Cereal and Oilseed Research Centre

Agriculture and Agri-Food Canada, Research Branch303 Saunders Building (49)

Central Experimental Farm, 930 Carling AvenueOttawa, ON K1A 0C6, Canada

(e-mail: shoemakerb@agr.gc.ca and hambletons@agr.gc.ca)

and

Michel Lacroix, Mario Tesolin, and Jean CoulombeQuébec ministère de l’Agriculture, des Pêcheries et de l’Alimentation

Direction des services technologiquesLaboratoire de diagnostic en phytoprotection

2700, rue Einstein, D.1.200hSainte-Foy, QC Q1P 3W8, Canada

(courriel : michel.lacroix@agr.gouv.qc.ca)

Accepted for publication 4 March 2002Contribution No. 011719

362 Can. J. Plant Pathol. Vol. 24, 2002