Mycologia. 88(3), 1996, pp. 484-491.

Armillaria nabsnona, a new from western North America

Thomas J. Volk1 pleiomorphic species with a wide host range and dis- Harold H. Burdsall, Jr. tribution. Herink (1973), among others, suspected Mark T. Banik that this single species might actually be a species Center for Forest Mycology Research, Forest Products complex. However, since basidioma morphology is Laboratory, Forest Service, United States Department of difficult to study because of overlapping and appar- Agriculture, One Gifford Pinchot Dr., Madison, 2 ently inconsistent traditional characters, other ave- Wisconsin 53705-2398 nues of research were pursued. Hintikka (1973) de- veloped a technique that allowed determination of Abstract: Armillaria nabsnona is characterized mor- mating types in Armillaria. Using a modification of phologically and described as a new species. It is com- this method, Korhonen (1978) was able to distin- patible with tester strains of North American biolog- guish five European biological species (EBS). The ical species (NABS) IX. Using restriction fragment technique depended on growing single spore isolates length polymorphism (RFLP) of the polymerase together in a petri dish and observing the change or chain reaction (PCR) products amplified from the lack of change in colony morphology. Single spore intergenic spacer (IGS) region of the ribosomal isolates of Armillaria species are generally white and DNA, we also demonstrate an association of cultures fluffy, but when fusion of compatible mating types of specimens with tester strains of NABS IX. Armil- occurs, the coalesced colonies become dark brown, laria nabsnona appears to be restricted to several spe- appressed, crustose, and sometimes produce rhizo- cies of hardwood trees in the states of , morphs, depending on nutritional and cultural con- , , and , and in the Ca- ditions. If the single spore isolates are from different nadian province of . Basidiomata species, the colonies will not grow together and will have been found in both the fall and spring. remain white and fluffy (Anderson and Ullrich, Key Words: , Basidiomycotina, NABS 1982). The five EBS have been characterized mor- IX, biological species, systematic, Tricholomataceae. phologically, and appropriate names have been given (see Watling, Kile, and Burdsall, 1991). Anderson and Ullrich (1979) applied the tech- INTRODUCTION niques used by Korhonen to isolates collected from widely distributed locations in North America and The genus Armillaria (Fr.: Fr.) Staude has undergone demonstrated that what had been considered as Ar- significant revision in the past twenty years. The ge- millaria mellea in North America was actually 10 dis- nus once accommodated any white-spored agaric tinct biological species (North American biological with broadly attached gills and an annulus (Volk & species or NABS), some of which are compatible with Burdsall, 1995). With Agaricus melleus Vahl: Fr. now EBS (Anderson et al., 1980; Anderson, 1986). Anoth- accepted as the species (Donk, 1962; Watling, er, apparently rare, biological species, NABS XI, was Kile, and Gregory, 1982), Armillaria has a much nar- recognized by Morrison et al. (1985). Three of the rower circumscription, including only those white- northeastern-occurring NABS were described and spored wood-inhabiting agarics with broadly attached given names by Bérubé and Dessureault (1988, to decurrent gills and forming macroscopic black to 1989). A tentative key to macromorphological iden- reddish-brown rhizomorphs. tification of Armillaria species has been published Until the late 1970s Armillaria mellea (Vahl: Fr.) (Burdsall & Volk, 1993). For additional nomenclatur- Kummer was considered by most researchers to be a al details see Volk and Burdsall (1995). Accepted for publication January 2, 1996. The cumbersome nature of the mating type meth- 1 Current email addresses [email protected], burdsall@ od of species identification has prompted a search facstaff.wisc.edu,/s=m.banik/[email protected] for other techniques for identifying collections. Sev- 2 The Forest Products Lab is maintained in cooperation with the eral methods have been employed with varying de- University of Wisconsin-Madison. This publication was written and prepared by U.S. government employees on official time, and it is grees of success including isozyme analysis (Morrison therefore in the public domain and not subject to copyright. et al., 1985; Bérubé, 1994), cultural characteristics

484 VoLK ET AL.: ARMILLARIA NABSNONA 485

(Rishbeth, 1986), immunology (Burdsall and Banik, from various areas on the Olympic Peninsula of 1990) and molecular biological techniques (Harring- Washington, USA, in October 1993. Compatibility ton and Wingfield, 1995). Of these, molecular bio- with tester strains was determined according to the logical techniques seem to hold the most promise in methods of Hintikka (1973), Korhonen (1978) and terms of quick, accurate diagnosis from a variety of Anderson and Ullrich (1979) using second genera- test materials. Barrington and Wingfield (1995) re- tion tester strains that we developed by testing against cently reported a technique involving restriction frag- the original bank of tester strains of Anderson and ment length polymorphisms (RFLPs) of the riboso- Ullrich (1979). Specific strain information is avail- mal DNA intergenic spacer (IGS) region amplified able from the authors upon request. To determine by the polymerase chain reaction (PCR). They were sexuality of this species, pairwise crosses of ten single able to separate all North American species of Ar- isolates from TJV-93-188 were done on millaria except for A. calvescens and A. gallica, which 1.5 % malt extract, 2 % agar medium (MEA). All cul- are apparently very closely related (Anderson and tures were incubated at 24 C. Stasovski, 1992). In Barrington & Wingfield (1995), Basidiomata were examined microscopically with digestion of the IGS region of NABS IX with the re- phloxine in 3 % KOH or in Melzer’s solution. Colors striction endonuclease Alu I yielded a unique band- in quotation marks are from Ridgway (1912). All ing pattern that distinguished it from the other spe- specimens and cultures (single spore and tissue iso- cies. This technique may prove invaluable for iden- lates) are deposited in the Center for Forest Mycol- tifying collections of this and other species in the fu- ogy Research (CFMR) Madison, Wisconsin, USA, ex- ture, and thus we examined it in conjunction with cept where noted. Other herbarium abbreviations the formal description of this species. are taken from Holmgren et al. (1990). In spite of being a known entity, NABS IX has re- mained morphologically undescribed and not validly Molecular studies.—Eighteen tissue isolates of NABS named. The species has been reported from Idaho IX and 2 single spore isolates of the original NABS (Anderson and Ullrich, 1979), California (Jacobs et IX testers (Anderson & Ullrich, 1979; TABLE I) were al., 1994), Alaska (Shaw & Loopstra, 1988), and Brit- analyzed for RFLPs in the IGS region using the tech- ish Columbia (Morrison et al., 1985). The species has nique of Barrington and Wingfield (1995). Fungal also been reported from Connecticut (Wargo, 1988), material for PCR was obtained from mycelial mats Newfoundland (Warren, 1994) and Japan (Moham- grown on cellophane overlaying enriched medium 2 med et al., 1994), but these identifications are incor- (EM, Larsen et al., 1992). Approximately 25 mm of rect, as noted later in this work. mycelial mat was ground with 500µL TE (10 mM The major difficulty in naming and describing Tris, pH 8, 1 mM EDTA) in a sterile 15 mL ground NABS IX has been finding specimens, and particu- glass tissue homogenizer and centrifuged at 14000g larly, an appropriate type collection, partially because for 10 min. The supernatant was diluted 1 to 100 in of the long lag period (up to two months) between TE and 5µL of this was used as template DNA in collection of the and the determination PCR. The remaining PCR reaction mixture is as fol- of the NABS by mating studies. To resolve this prob- lows; 1.25 units Taq polymerase (Promega, Madison, lem, methodical collection, culture, and molecular WI), reaction buffer supplied with the enzyme (final studies were initiated in the fall of 1993 using large concentrations 50 mM KCl, 12.5 mM Tris-HCl, pH-9, numbers of Armillaria basidiomata from the Olympic 0.1 % Triton X-100), 1.5 mM MgCl, 200 uM each Peninsula of Washington. Identification was con- dNTP, and 0.2 uM of each primer in a volume of firmed by compatibility with tester strains, and the 25µL. The primers used were LR-12R (5’ CTGA- morphological features of the basidiomata studied ACGCCTCTAAGTCAGAA 3’) and O-1 (5’ AGTCCT- and characterized. The literature was also searched ATGGCCGTGGAT 3’) (Operon Technologies, Ala- (Volk and Burdsall, 1995) to determine if there was meda, CA). The reaction mixture was overlain with an existing name for this . Type specimens of a drop of mineral oil. Thermocyler (Perkin-Elmer) northern temperate Armillaria species were also ex- parameters were as follows: One cycle of 93 C for 3 amined. No appropriate name was found, so the fun- min, 53 C for 2 min and 72 C for 3 min followed by gus is described here as a new species. 29 cycles of 93 C for 1.5 min, 53 C for 2 min and 72 C for 3 min. This was followed by elongation at 72 C for 10 min. Following amplification, 3 µL of the un- MATERIALS AND METHODS purified reaction product was digested for 16 h at 37 Cultural and morphological studies. — Isolation of sin- C with 5 U Alu I (GIBCO BRL, Gaithersburg, Mary- gle was made according to Darmono land) mixed with the appropriate amount of buffer and Burdsall (1992) from fresh basidiomata collected supplied with the enzyme in a total volume of 20 µL. 486 MYCOLOGIA

The digestion products were separated on a 4 % aga- rose gel in tris-acetic acid-EDTA (TAE) buffer (pH 8) at 2.4 V/cm for 3-4 h. The products were visualized with UV illumination after being stained in an ethid- ium bromide solution for 10 min. The sizes of the testers. Pattern-b consisted of 3 bands approximately digestion products were determined by comparing 306, 230 and 196 bp and was found in 3 isolates. their migration distances to those of a 50-2000 bp Pattern-c consisted of bands approximately 560, 321, standard (Biorad, Hercules, California) using a semi- 237 and 203 bp and was present in 7 isolates. Bands logarithmic scale. smaller than 100 bp were not sized. To determine if Pattern-c was a heterozygous con- RESULTS dition resulting from the mating of Pattern-a and -b haplotypes, two Pattern-a and two Pattern-b single Mating studies.—Single spore isolates from several spore isolates were paired in all combinations. Each collections of Armillaria from the Olympic Peninsula, nonself pairing resulted in the development of a which were “fluffy” in appearance, became ap- crustose type colony from which hyphal tips were mi- pressed and dark when paired with tester strains of croscopically excised from the last of three successive NABS IX. These new single-basidiospore isolates were subcultures. Hyphal tip cultures that came from pair- paired with one another and exhibited a similar ings of parents of like haplotypes exhibited the same change in colony morphology, thus indicating they Alu I digestion pattern as the parents. However, hy- belong to a single compatibility group or biological phal tip cultures that were obtained from the four species (Korhonen, 1978; Anderson & Ullrich 1979). pairings of parents with different haplotypes exhib- Intra-basidioma matings between single spore isolates ited Alu I digest Pattern-c. Thus Pattern-c appears to within three basidiomata indicated that the sexuality be a heterozygous condition following a mating be- of this fungus is determined by two genetic loci; i.e. tween haploids with Patterns-a and-b. In this respect the fungus exhibits a tetrapolar mating system (data the occurrence of Pattern-c indicates a compatible not shown ). mating. Molecular biology studies. — The PCR product of the To determine the heritability of Pattern-c, the pat- IGS region for each isolate was approximately 900 bp. tern types of 10 single spore isolates from a basidi- Alu I digests of this product exhibited three different ome (TJV-93-188) with an associated Pattern-c tissue banding patterns (FIG. 1; TABLE I), two distinct pat- culture were determined. Six of the single spore iso- terns and a third that appeared to be a combination lates exhibited Pattern-a, two Pattern-b, and two Pat- of the first two. Pattern-a consisted of 2 bands ap- tern-c (Fig. 1). proximately 563 and 200 bp (similar in size to the 534 and 200 bp found for this species by actual DISCUSSION counting of base pairs in the DNA sequences by Har- rington and Wingfield, 1995) and was shared by 8 The size of the amplification product and Alu I isolates and the single spore isolates of Anderson’s digestion Pattern-a of the IGS region were similar to V OLK ET AL.: A RMILLARIA NABSNONA 487

that observed by Harrington and Wingfield (1995) it appears that the Alu I digestion pattern of the IGS for NABS IX. However, they did not observe Pat- region is a reliable diagnostic tool. terns-b and -c reported here. This is not surprising Thus, two major lines of evidence point to the cul- since they only examined haploid isolates from An- tures and specimens we collected being contaxic with derson’s original testers, which came from only two NABS IX. The single basidiospore isolates of the col- different fruiting bodies. None of the patterns ob- lected basidiocarps were compatible with tester served here for the NABS IX isolates could be con- strains of NABS IX and should be considered the fused with the patterns observed for the other species same biological species. Three unique patterns of di- of Armillaria reported by Harrington and Wingfield gest with Alu I also link the specimens and cultures (1995). Pattern-b appears to be the result of an ad- with tester strains of NABS IX. With basidiomata in ditional restriction site not present in Pattern-a. Thus hand, we are able to describe a new species of Ar- the 563 bp fragment of Pattern-a is restricted once millaria with distinctive characteristics. to yield the 306 and 230 bp fragments of Pattern-b. The occurrence of Pattern-c in single spore isolates DESCRIPTION OF SPECIES raises questions as to the true nature of the pattern. Since Pattern-b seems to result from the addition of Armillaria nabsnona Volk & Burdsall sp. nov. a restriction site not present in Pattern-a isolates, it F IGS. 2-9 is conceivable that Pattern-c is actually an artifact of Fungiformi, pileo 4-5 cm diam, aurantio-brunneo; la- an incomplete Pattern-b digest. If this were the case, mellis adnatis vel decurrentibus; stipite 8-10 cm × 2-5 mm, repeats of the digestions would yield inconsistent re- brunneo; velo supero, affixo 2-3 cm ab apice; contextu 0.5- sults with regard to the expression of either Pattern-c 1 mm late; basidiis 25-35 × 5.5-6 µm, nodose-septatis; bas- or-b. However, each isolate tested consistently exhib- idiosporis ovoideis vel subglobosis, (6-) 8-10 × 5.5-6.5 µm, ited only one pattern type when the Alu I digests inamyloideis; habitatio riparum, gregario ad lignum angios- were repeated. permarum arborum. Harrington (pers. comm.) has observed two pat- HOLOTYPE. TJV93-188 on fallen trunk of Acer macro- tern types in the IGS region of NABS IX isolates phyllum Pursh, Hoh River Rain Forest Trail, Olympic Na- when digested with Hind II. Upon digestion of our tional Park, Jefferson County, Washington, USA, leg. Tho- isolates with Hinc II, an isoschizomer of Hind II, we mas J. Volk, 19 October 1993; deposited CFMR. Type cul- ture: tissue and single spore isolates, deposited CFMR. also found two restriction patterns that occurred in isolates with either Alu I Pattern-a or Pattern-b. The Etymology. Named for NABS IX: nabs = acronym two Hinc II patterns also combined to make a third for North American biological species, nona = ninth. pattern in isolates that demonstrated Pattern-c with Basidiomata tricholomatoid (FIG. 2-4). Pileus at Alu I. Thus it appears that Pattern-c is not the result first convex, later plane; 4-7 cm broad when fully of partial digestion but of heterokaryon formation expanded; surface smooth, hygrophanous, slimy-ap- involving Pattern-a and-b haplotypes. pearing when wet (FIG. 3), sometimes with short dark Interestingly, the two single spore isolates that fibrils (“hairs”) on the disc when young; disc “snuff showed Pattern-c with Alu I also showed a combined brown,” paler toward margins, often with darker ir- pattern with Hinc II. The presence of two different regular bruise-like areas on or near the surface; mar- haplotypes in a single spore isolate could have several gin slightly incurved, “clay color,” translucent striate explanations. It is possible that the single-spore iso- (striate appearance due to observation of gills lates actually contained more than one nucleus, ei- through the thin flesh) to furrowed. Context 0.5-1 ther in a compatible or incompatible relationship, as mm thick, white. Lamellae adnate to subdecurrent, has been proposed for other basidiomycetes (Kühn- subdistant, 0.75-1 mm wide, white to cream colored, er, 1977; Hallenberg, 1983). Morphologically these darkening in age to pinkish-tan, often developing cultures appeared to be “typical” haploids, although brownish patches. 8-10 cm long, 4-5 mm broad one of the isolates displayed erratic mating behavior at base, narrowing to 2–3 mm broad at apex, “mum- during the intrabasidiomata crosses. This may be an my brown” at base, paler “buckthorn brown” to indication of the presence of two nuclei with hemi- “warm buff’ at the annulus; with white cottony compatible mating types; Anderson and Ullrich patches below the annulus; context white, fibrous, (1982) found that auxotrophic haploid isolates of Ar- “peeling” away in parallel strips (as is typical for Ar- millaria could be paired to form common-A or millaria species). in buttons dense white common-B “fluffy” diploid prototrophs. Another less cottony until rupture; annulus flaring upward at first likely explanation for Pattern-c is that the IGS region (FIG. 4), soon becoming ragged as the pileus ex- may vary among its copies within a given nucleus. pands, sometimes persisting as an evanescent cortina, Despite this anomaly, for the NABS IX isolates tested, but more often washed away and difficult to observe, 488 MYCOLOGIA

sometimes completely lost upon drying. Rhizomorphs frequently lacking when present, thick (1-2 mm broad), black, and branching. Basidiospores white in mass; ovoid to subglobose; (6-)8-10 × 5.5-6.5 µm, smooth, somewhat thick- walled at maturity, hyaline, negative in Melzer’s Re- agent (FIG. 5). Basidia clavate; 25-35 × 5.5-6 µm, 3- 4 µm broad just distal to clamped basal septum; four- sterigmate; sterigmata 4.5-6 µm long at maturity, 1 µm wide at base; with a clamp connection at base (FIG. 6); a second basidium usually forms from the basal clamp connection, sometimes with a third ba- sidium developing from a clamp on the second (FIG. 7). This is most obvious in young incompletely ex- panded basidiomata. Cystidia not observed, but thin (2-3 um) wide hypha-like cells can sometimes be found among basidia. Pileipellis composed of swollen unbranched, hyaline, terminal cells, 40-60 × 14-16 urn; with an underlying layer of more or less parallel, V OLK ET AL.: ARMILLARIA NABSNONA 489

hyaline hyphae, 7-8 µm wide (FIG. 8). Partial Veil DAVFP; DAVFP-24863, Alnus rubra, Goldstream, Van- composed of more or less parallel hyaline hyphae, 4- couver Island, leg. D. Morrison, 20 Oct. 1990. 949- 7 µm diam, rarely branching, with frequent simple 5003-01. CFMR, DAVFP. septa. Stipe Context composed of parallel and infre- Other cultures studied. TO provide additional infor- quently branching, simple septate, hyaline hyphae; mation on hosts and geographical distribution the elements 10-15 µm wide, sometimes swollen in the following additional cultures were studied. mid-region. Stipitipellis of closely appressed, parallel, Other cultures, without known specimens, collected simple-septate, brown hyphae that frequently anas- and determined by Duncan J. Morrison as NABS IX tomose with H-connections between the parallel hy- and deposited at DAVFP (Morrison, pers. comm.). CAN- phae; elements 8-10 µm broad, not swollen. Subhy- ADA BRITISH COLUMBIA: Sp8128, Parksville, Van- menium with large clamp connections present on couver Island, unknown host, Oct. 1981; Sp8130, some hyphae (FIG. 9) especially on the 2-3 septa Sp8131, Ladysmith, Vancouver Island, Alnus rubra, proximal to the basidia where hyphal branches pro- Oct. 1981; Sp8132, Colwood, Vancouver Island, Acer liferate directly from the clamps. Gill trama regular, macrophyllum, Oct. 1981; Sp8133, Colwood, Vancou- hyphae usually simple septate, with occasional clamp ver Island, Alnus rubra, Oct. 1981; Sp8135, Egmont, connections. unknown host, Oct. 1981; Sp8258, Squamish BC, Pop- Habit and habitat. Gregarious, but not caespitose ulus trichocarpa Torr. & Gray, 4 Nov. 1982; Sp8260, (FIG. 4), forming from branched black rhizomorphs Squamish, Alnus rubra, 4 Nov. 1982; Sp8261, Vancou- or from 1 mm thick pseudosclerotial plates in the ver, Univ. of BC campus, Alnus rubra, 4 Nov. 1982; wood. Basidiomata have been found in both the fall Sp8263, Vancouver, unknown host, 4 Nov. 1982; (Sept.-Nov.) and the spring (April in Oregon). Sp8352, Hope, Alnus rubra, 27 Oct. 1983; Sp8361, Lake Cowichan, Vancouver Island, Alnus rubra, 1 Specimens examined. (* indicates cultures also stud- NOV. 1983. ied and on deposit): USA, WASHINGTON: TJV-93- Other cultures at CFMR uithout known specimens. 188 (holotype) on fallen trunk of CANADA, BRITISH COLUMBIA 121-1, 121-2, Acer Pursh, Hoh River Rain Forest Trail, Olympic National macrophyllum, Anderson et al., 1980, AM-111, Mor- Park (ONP), Jefferson County, leg. T.J. Volk, 19 Oct. rison, Coquihalla BC; USA, IDAHO: 139-1, 139-2, 1993. CFMR*; TJV-93-200, on trunk of Acer circina- unknown host, Anderson et al., 1980; USA, CALI- tum Pursh, Hoh River Rain Forest, 200 m west of Rain FORNIA: Alnus rubra, Anderson and Ullrich, 1979; Forest Trail, ONP, Jefferson County, G.R. Walker Ex- USA, IDAHO: HB-19, McDonald, 1992, Priest River perimental Transect 10-4-93HO, Plot 7, leg. Experimental Forest, rotten log, leg. G. McDonald; G. Walker and M. Puccio, 20 Oct. 1993. CFMR*; TJV- HB-20, 1992, Priest River Experimental Forest, Idaho, 93-198, on down log of Alnus rubra Bong., flood leg. G. McDonald; HB-21, 1992, Priest River Experi- plain of Hoh River, ONP, Jefferson County, leg. TJ. mental Forest, Idaho, Elderberry, leg. G. McDonald; Volk, 20 Oct. 1993, CFMR*; OKM-25910, on Alnus USA, ALASKA: Arm. -Shaw C. C., 5 Nov. 1989, Juneau, sp. under very large Picea sitchensis and Tsuga heter- Alaska, leg. G.C. Shaw. ophylla. Klahanie campground, Olympic National Forest, Clallam County, leg. O.K. & H. Miller, L. & Comments.—Armillaria nabsnona can be identified M.Bailey, 14 Oct. 1993. CFMR*, VPI*; OKM-25911, most easily by its geographical distribution, primarily Alnus rubra snag, Klahanie campground, Olympic the west coast of North America, and host range, pri- National Forest, Clallam County, leg. M. & L. Bailey, marily on hardwoods in riparian areas, especially fre- O.K & H.Miller. 14 Oct. 1993. CFMR* VPI*; USA, quent on Alnus species. Macroscopic characters that CALIFORNIA: Men-39, (= Men 1, specimen 39, = UC may be used to distinguish A. nabsnona from other 1598524) soil, Mendocino County, leg. K.Jacobs, (Ja- North American species include a more orange col- cobs et al., 1994) Spring 1990, CFMR. UC.; USA, OR- oration when fresh and also a narrower stipe in com- EGON: JR-A-l, JR-A-2, JR-A-3, JR-A-4, Rivermill Park, parison to the size of the pileus. The stipe is darker Estacada, Clackamas Co., Oregon Elev. 400 ft. on Acer than other Armillaria species, especially when dried. circinatum roots and buried limbs, leg. J.A. Roger 14- There are no scales, but small black hairs may be 21 April 1995, CFMR*; CANADA, BRITISH COLUM- present on the surface of the pileus, a similar situa- BIA: HHB-13535, Alnus sp. log, in mixed conifer tion to that found in A. mellea. Microscopically, Ar- area, One Mile Creek. near Callaghan Lake, Garibal- millaria nabsnona can be distinguished from other di National Forest, near Whistler BC. leg. H.H. Burd- western species of Armillaria by the pattern of sall, Jr., 6 Oct. 1990. CFMR; DAVFP-24864, Acer ma- branching of the basidia; the second basidium crophyllum, Goldstream, Vancouver Island, leg. emerges from the clamp of the first, the third basid- D.Morrison, 20 Oct. 1990. 94-9-5003-01. CFMR, ium emerging from the clamp of the second, and so 490 MYCOLOGIA

on. This is particularly obvious in immature speci- tures; G. Walker and M. Puccio (WTU), H. Miller, M. Bailey, mens, but often becomes obscured in more mature and L. Bailey for assistance in collecting P. M. Wargo, specimens as the basidia become larger and more (Hamden, CT), G. McDonald (Moscow, ID), G. C. Shaw closely packed together. Armillaria nabsnona has un- (Fort Collins, CO), and G. Kile (CSIRO, Australia) for in- formation on distribution; I. Tavarres (UC) for loan of spec- branched terminal cells in the cuticle of the pileus, imens; M. J. Larsen (Moscow, ID) for helpful comments: distinguishing it from the similar A. mellea, which has and Judy Roger (Estacada, OR) for collection and gift of frequently-branching terminal cells in the pileus cu- spring-fruiting specimens. We also thank the Olympic Na- ticle. Armillaria mellea also lacks clamps at the bases tional Park and Olympic National Forest for permission to of the basidia. Although it fruits much more com- collect within their boundaries. monly in the fall, Armillaria nabsnona is also the only Armillaria species so far found fruiting in the spring in the , although the frequency of this phenomenon is not known. The Armillaria cultures from Connecticut on co- nifer (Wargo, 1988) and on black oak (Strain AW-20; Wargo, pers. comm.), indicated as NABS IX by War- go and mentioned by Burdsall and Volk (1993) have been retested and are not compatible with tester strains of NABS IX or with any other biological spe- cies (Wargo, pers. comm. ), although Harrington (pers. comm.) reports that strain AW-20 shows the RFLP pattern of Armillaria ostoyae (Romagn.) Herink (NABS I). Warren (1994) stated that NABS IX is the most prevalent species in Newfoundland, but Bérubé (1994) found only A. ostoyae in his isozyme study of 39 Newfoundland isolates, many of which were War- ren’s (Warren and Bérubé, pers. comm.). Thus, Ar- millaria nabsnona has not been confirmed in North America east of the Rocky Mountains. Guillaumin et al. (1988), Kile et al. (1994) and Mo- hammed et al. (1994) have stated that NABS IX has been isolated from the Japanese island of Hokkaido. J. J. Guillaumin generously provided these isolates to us; we retested them in “dip-hap” crosses and found no evidence for compatibility of the Japanese isolates with our fresh NABS IX testers. In their study of the Armillaria species on Hokkaido, Cha et al. (1994; Cha, pers. comm.) have also reported finding none of their Armillaria isolates compatible with testers of NABS IX. There is no good evidence, in our opinion, that NABS IX exists in Japan or anywhere outside Northwestern North America.

ACKNOWLEDGMENTS

We thank O. K. Miller (VPI) for gift of specimens, slides, and descriptions; W. J. Sundberg (SIU), D. M. Rizzo (CFMR, UCD), and T. C. Harrington (ISU) for review of manuscript; J. Ammirati (WTU) for support of collecting; J. Paul (CFMR) for technical assistance; K. K. Nakasone (CFMR) for help with the Latin diagnosis; D. J. Morrison and B. Callan (DAVFP) for generous gift of specimens and information on cultures; K. Jacobs (Morton Arboretum, Il- linois) and J. J. Guillaumin (INRA, France) for gift of cul- V OLK ET AL.: ARMILLARIA NABSNONA 491