Trichoderma brevicompactum sp. nov

Mycologia, 96(5), 2004, pp. 1059–1073. q 2004 by The Mycological Society of America, Lawrence, KS 66044-8897

Trichoderma brevicompactum sp. nov.

Gu¨nther F. Kraus1

Abstract: Trichoderma brevicompactum, a new species, was isolated from soil or tree bark in North, Central and South America, including the Caribbean Islands, and southwestern and southeastern Asia. Morphological and physiological characters, the internal transcribed spacer regions of the rDNA cluster (ITS15.8SrDNA-ITS2) and partial sequences of translation elongation factor 1-alpha (tef1) are described. Trichoderma brevicompactum is characterized by a pachybasium-type morphology, morphologically resembling other small-spored species referable to Trichoderma section Pachybasium but with essentially subglobose conidia. It is most closely related phylogenetically to Hypocrea lutea, from which it differs in morphological and physiological characters. Key words: Biolog, Hypocrea, molecular phylogeny, Pachybasium, soil mycoflora, taxonomy, Trichoderma

Austrian Center of Biological Resources and Applied Mycology (ACBR), Institute of Applied Microbiology (IAM), University of Agricultural Sciences, Nußdorfer La¨nde 11, A-1190 Wien, Austria

Irina Druzhinina1 Research Area of Gene Technology and Applied Biochemistry (DGTAB), Institute of Chemical Engineering, University of Technology, Getreidemarkt 9/1665, A-1060 Wien, Austria

Walter Gams Centraalbureau voor Schimmelcultures, P.O. Box 85167, 3506 AD Utrecht, The Netherlands

John Bissett Agriculture and Agri-Food Canada, Eastern Cereal and Oilseed Research Center, Central Experimental Farm, Ottawa, Ontario, K1A 0C6

Doustmorad Zafari INTRODUCTION

Department of Plant Protection, Bu Ali Sina University, Hamadan, Iran

Species of the deuteromyceteous genus Trichoderma are cosmopolitan and typically soil-borne or wooddecaying fungi (Klein and Eveleigh 1998). About 40 taxa of Trichoderma have been described to date. However, based on the number of described taxa in the Hypocreales with Trichoderma anamorph morphology, the actual number has been estimated to exceed 200 (Samuels 1996). Taking into account the fact that many Trichoderma spp. may lack a teleomorph, this estimate indicates that most Trichoderma spp. have not yet been described. Most Trichoderma spp. have been described from North America and Europe. The investigation of other geographic areas likely will lead to the recognition of new species. We have undertaken a study recently of the global biodiversity of Trichoderma. From an initial investigation in central and southeastern Asia (Kullnig et al 2000, Kubicek et al 2003), seven new species in Trichoderma were differentiated by molecular and physiological methods and subsequently described by Bissett et al (2003). However, many more putative new species were collected from other geographic ar-

George Szakacs Department of Agricultural Chemical Technology, Technical University of Budapest, 1111 Budapest, Gellert ter 4, Hungary

Alexei Koptchinski Research Area of Gene Technology and Applied Biochemistry (DGTAB), Institute of Chemical Engineering, University of Technology, Getreidemarkt 9/1665, A-1060 Wien, Austria

Hansjo¨rg Prillinger Austrian Center of Biological Resources and Applied Mycology (ACBR), Institute of Applied Microbiology (IAM), University of Agricultural Sciences, Nußdorfer La¨nde 11, A-1190 Wien, Austria

Rasoul Zare Department of Botany, Plant Pests Diseases Research Institute, PO Box 1454, Tehran, Iran

Christian P. Kubicek2 Research Area of Gene Technology and Applied Biochemistry (DGTAB), Institute of Chemical Engineering, University of Technology, Getreidemarkt 9/1665, A-1060 Wien, Austria

Accepted for publication March 30, 2004. 1 These authors contributed equally to this paper. 2 Corresponding author. E-mail: [email protected]

1059

1060

MYCOLOGIA

eas and are being studied in detail. One of these undescribed species, ‘‘Trichoderma sp. 1’’ (Kullnig-Gradinger et al 2002), was encountered from diverse regions of North, Central and South America and southern Asia and shown to occupy a unique phylogenetic position within Trichoderma (Kullnig-Gradinger et al 2002). This taxon is described here as Trichoderma brevicompactum, using a polyphasic (morphological characters, sequence analysis and substrate assimilation patterns) approach. MATERIALS AND METHODS

Strains.—Those investigated in this study are given in TABLE I. They are maintained in the culture collections of the Austrian Center of Biological Resources and Applied Mycology, Vienna, Austria; at the Division of Gene Technology and Applied Biochemistry, Vienna, Austria; and at DAOM (Eastern Cereal and Oilseed Research Centre, Ottawa, Canada). Representative cultures also have been deposited at the Centraalbureau voor Schimmelcultures (CBS), Utrecht, The Netherlands. Morphological examination.—Cultures were grown on oatmeal agar (OA) (Gams et al 1998), cornmeal agar (CMA) and 3% malt-extract agar (MA), at 20–22 C under ambient daylight or in a 12 /12 h light/dark cycle under fluorescent light or near-UV lamps. Descriptions are based on observations on OA because this enabled heavier sporulation and more regular development of conidiophores (G. Kraus and W. Gams unpubl data). Color codes and terminology are based on Kornerup and Wanscher (1978). Microscopic observations and measurements were made from slides mounted in 90% (w/v) lactic acid (Bissett 1991a) plus 0.1% (w/v) cotton blue to aid in the contrast for measurements and retard conidia movement. Conidiophore structure and morphology were described from conidiophores taken from the edge of conidiogenous pustules or fascicles, as conidia were maturing, usually after 4–7 d incubation. Conidial morphology and measurements were recorded after 14 d. For linear growth measurement, MA plates point-inoculated with spores were incubated in the dark at 20 6 1 C. Where possible, 30 measurements and at least three replicas of each parameter for each isolate were made. Mean and range were determined for each character and each isolate. Scanning electron microscopy was performed on specimens that were shock-frozen in liquid nitrogen and coated with gold in argon gas (2170 C) as described by Becket and Read (1986). DNA sequencing and phylogenetic analyses.—DNA was isolated from fresh mycelium as described previously (Turner et al 1997). A region of nuclear rDNA, containing the internal transcribed spacer regions 1 and 2 and the 5.8S rDNA gene region, was amplified by PCR using the primer combinations SR6R and LR1 in 50 mL volumes (White et al 1990) in an automated temperature-cycling device (Biotron, Biometra, Go¨ttingen, FRG), using these parameters: 1 min initial denaturation at 94 C, followed by 30 cycles of 1 min denaturation at 94 C, 1 min primer annealing at 50 C,

90 s extension at 74 C and a final extension period of 7 min at 74 C. A 0.2 kb fragment of tef1 was amplified by the primer pair tef1fw (59-GTGAGCGTGGTATCACCATCG-39) and tef1rev (59- GCCATCCTTGGAGACCAGC-39), with this amplification protocol: 1 min initial denaturation (94 C), 30 cycles each of 1 min at 94 C, 1 min at 59 C, and 50 s at 74 C, and a final extension period of 7 min at 74 C. Template DNA (100 mL) was prepared from PCR products by purifying it with a commercial kit (Cleanmix, Fa., Talent, Italy) and sequenced by cycle-sequencing (Robocycler 40 Stratagene, La Jolla, California) with the ThermoSequenase-kit (Amersham Life Science, Piscataway, New Jersey) by the aid of a LI-COR 4000L automatic sequencing system (LI-COR Inc., Lincoln, Nebraska) as described previously (Kindermann et al 1998). The NCBI GenBank accession numbers for all sequences included in the analysis are given in TABLE I. DNA sequences were aligned first with Clustal X 1.81 (Thompson at al 1997) and manually adjusted using Genedoc 2.6 (Nicholas and Nicholas 1997). Single gaps were treated either as missing data or as the fifth base, as indicated, and sequence areas with ambiguous alignment were excluded from the analysis. Phylogenetic analyses were performed in PAUP* 4.0b10 using Hypocrea aureoviridis Plowr. & Cooke CBS 245.63 as outgroup for the sequence data from both loci, and T. harzianum Rifai CBS 226.95 as outgroup for ITS1-5.8SrDNA-ITS2 sequence data. A parsimony analysis was performed using a heuristic search, with a starting tree obtained via stepwise addition, with random addition of sequences with 1000 replicates, tree-bisection-reconnection as the branch-swapping algorithm, MulTrees in effect. Stability of clades was assessed with 500 bootstrap replications. Unique sequences obtained in this study have been submitted to GenBank (see TABLE I). The MSA file and phylogenetic trees have been deposited in the Treebase (http://www.treebase.org/treebase/submit.html) database under the submission code SN1503. Physiological studies.—Trichoderma strains were inoculated on 2% malt agar plates and allowed to incubate under ambient daylight for 7 d or until sufficient conidiation developed for inoculation of the Biolog microplates. Conidia were collected by rolling a sterile cotton swab over areas of conidiation and dispersing them in a sterile inoculating fluid comprising a gelling agent (0.25% phytagel) and surfactant (0.03% Tween 40y) in distilled water. A uniform spore concentration was achieved by adjusting the absorbance of the suspension to 75 6 2% at 490 nm using a turbidimeter. One hundred mL of suspension was inoculated into each of the 96 wells of the Biolog FF MicroPlatey (Biolog Inc., Hayward, California). Microplates were incubated at 26 C in the dark and absorbance readings at 490 nm (mitochondrial activity) and 750 nm (growth) recorded using a microplate reader after 24, 48, 72 and 96 h. Statistical analyses.—Absorbance data at 490 nm and at 750 nm were analyzed separately. Absorbance readings at 750 nm were used as a measure of mycelial density, interpreted as assimilation and growth on the test substrate. The absor-

AY324176 AY324176

AY324173/AY324183 AY324173/AY324183

MA 3296 (CBS 109720, DAOM 231232) MA 4103

AY324172/AY324183 AY324173/AY324183

(RMF 38.01) (WSF 433.6)

DAOM 229887 (WSF 3011)

CIB T37

—

soil (W. Gams) floor of mushroom farm (G. Alm) soil (P. Stotland) grey birch wood ( J. Bissett) balsam poplar wood ( J. Bissett) soil (M. Christensen) soil (M. Christensen) mud (G. Szakacs) soil in grape field (G. Szakacs) soil (M. Christensen)

soil (D. Zafari) soil (D. Zafari) soil (D. Zafari) tropical plants soil (M. Christensen) A1 horizon soil, conifer hardwood forest (M. Christensen) Al horizon soil, willow cottonwood forest (M. Christensen) ant colony soil, Caribbean Coastline (S. Orduz and L. Hoyos)

Sheffield, UK Leamington, Ontario, Canada Cape Province, South Africa Tincap, Ontario, Canada Almonte, Ontario, Canada Sidney, Nebraska, USA South Dakota, USA Danube R., Budapest, Hungary Goereme, Turkey Wisconsin, USA

Magdalena, Colombia

Wisconsin, USA

Gazvin, Iran Khoramabad, Iran Alshter Road, Khoramabad, Iran Medical Plants Trail, ACEER, Peru Wisconsin, USA

Mexico City, Mexico

Costa Rica Trivandrum, India

Geneva, NY, USA

Geneva, NY, USA

Union Island, Saint Vincent and the Grenadines Geneva, NY, USA Geneva, NY, USA

Kuriva forest, Papua New Guinea

Geographic location

TRICHODERMA BREVICOMPACTUM

Trichoderma harzianum CBS 226.95 DAOM 222183 JB RSA-7-2 JB T1380 JB T1256 RMF 7844 RMF 4460 TUB F-1020 TUB F-444 WSF 2994

AY324181 AY324180 AY324179 — —

AY324173/AY324183 AY324173/AY324183 AY324173/AY324183 AY324172/AY324183 AY324172/AY324183

soil, artificial rain forest (G. Szakacs)

home garden (K. Onelik and G. Harman) soil in corn field (S. Petzolt and G. Harman) soil in sunflower field (S. Petzolt and G. Harman) soil in pumpkin field (K. Onelik and G. Harman) maize field soil (M. Lu¨beck) soil in backyard (G. Szakacs)

ex rhizosphere of Glycosmis sapindoides (G. Kraus) soil (G. Kraus)

Habitat

ET AL:

—

AY324176

AY324173/AY324183

112447; TUB F-

AY324178 AY324176

AY324173/AY324183 AY324173/AY324183

112445) 112446; TUB F-

MA 4105 (CBS MA 4106 (CBS 1029) MA 4107 (CBS 1076) D.Z. 01 D.Z. 02 D.Z. 47 DAOM 229938 DAOM 229977

AY324176 AY324177

AY324173/AY324183 AY324173/AY324183

MA 3294 (CBS 109722) MA 3295 (CBS 109721)

—

AY324173/AY324183

AY324176

tef1

MA 1374 (CBS 112444)

ITS1/ITS2 AY324174/AY324183

Strain

Trichoderma and Hypocrea strains used in this study*

Trichoderma brevicompactum MA 761 (CBS 112443)

TABLE I.

KRAUS 1061

Strain

Continued ITS1/ITS2 tef1

log (G.J. Samuels, GJS 96-280) (K. Dumont) decorticated wood (G.J. Samuels, GJS 89129) Hapalopilus rutilans (K. Poldmaa) culture contaminant ( J. Bissett) peat moss (R. Otis) air ( J. Bissett) beech wood ( J. Bissett) fallen branch of oak ( J. Bissett)

branch (A. Aptroot) soil ( J.H. Miller) potting soil ( J. Bissett) mushroom farm, casing soil ( J.H. Miller) soil ( J. Bissett) soil (M. Christensen) park soil (G. Szakacs)

Habitat

Artu, Estonia Ottawa, Ontario, Canada Riviere-du-Loup, Quebec, Canada Ottawa, Ontario, Canada Wellington, Ontario, Canada San Lorenzo, Guatemala

Luquillo Mts., Puerto Rico Brazil Estatoe South, North Carolina, USA

Madang, Papua New Guinea Georgia, USA Ottawa, Ontario, Canada Airdrie, Alberta, Canada Georgia, USA Cancun, Mexico Colorado, USA Singapore

Geographic location

* ex-type strain is given in bold; each haplotype has been deposited only once. Abbreviations: CBS, Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands; DAOM, Eastern Cereal and Oilseed Research Centre, Ottawa, Canada; MA, Austrian Center of Applied Mycology and Bioresources, Vienna, Austria; RMF and WSF, collection of M. Christensen in DAOM; D.Z., collection of D. Zafari; TUB, collection of G. Szakacs; JB, from the working collection of J. Bissett, Eastern Cereal and Oilseed Research Centre (DAOM), Ottawa; RMF, WSF, from the collection of Martha Christensen, University of Wyoming, Laramie. Collectors of the various isolates are given in brackets after the habitat.

CBS 597.95 DAOM 226754 JB 00M-23 JB ID306 JB T1353 JB GS1-1

Hypocrea lutea/Gliocladium viride CBS 355.97 CBS 238.81 CBS 102037

Trichoderma virens CBS 350.96 CBS 249.59 DAOM 226904 DAOM 222102 DAOM 167651 JB M1-10 RMF 5054 TUB F-899

TABLE I.

1062 MYCOLOGIA

KRAUS

ET AL:

TRICHODERMA BREVICOMPACTUM

bance spectrum of hyaline mycelium is essentially level over the range of 490–750 nm (Kubicek et al 2003). Therefore, a corrected value for mitochondrial activity, indicated by the production of a colored formazan salt resulting from the reduction of INT mediated by succinate dehydrogenase activity in the citric acid cycle, was obtained by subtracting the 750 nm reading (490–750 nm). Absorbance data were not corrected for growth in the control well, which was treated as an independent variable in the analyses. Cluster analyses were performed using NTSYS (Rolf 1997), based on a similarity matrix using the product-moment correlation coefficient and employing the UPGMA clustering method. SAS was used for analyses of variance (ANOVA) and canonical variate analyses (SAS Institute Inc. 1989). There were insufficient degrees of freedom (37 samples, 96 variables) to perform multivariate analyses on the entire dataset. Instead, univariate ANOVAs were performed on data for each of the 95 different carbon-substrates and the control. The substrates were ranked on the ANOVA F-values, and the degree of significance of the among-species variation in the ANOVAs. An arbitrary number of variables (n 5 18, less than half the number of isolates) were selected to perform canonical variate analysis, choosing the highest ranked variables. Wilk’s Lambda and Pillai’s trace were employed to test the significance of the canonical variate analysis. Variable loadings from the ‘‘between canonical structure’’ were used to interpret the two eigenvectors obtained from the analysis.

TAXONOMY

Trichoderma brevicompactum G.F. Kraus, C.P. Kubicek & W. Gams sp. nov. (FIGS. 1 and 2; TABLE II) Coloniae in OA rapide crescentes, plus quam 100 mm diam post 5 dies 20 6 1 C, copiose sporulantes in zonis latis compactis. Conidiophora in pustulis viridi-olivaceis vel griseo-olivaceis dense aggregata. Conidiophora modo pyramidae verticillata (typo pachybasii), sed apicibus sterilibus denique carentia. Axis conidiophororum et ramorum primariorum vulgo 4–5.5 mm lata, rami ultimi 3–7.5 3 3–4 mm. Phialides 2–5 verticillatae, late ampulliformes, 5–6.5 3 (3–)3.5–4 mm, collulo brevi; phialides solitariae vel terminales longiores, usque ad 12 3 5 mm. Phialides et rami lati et breves, conidiophoro aspectum ‘‘brevi-compactum’’ praebentes. Conidia leves globosa vel ellipsoidea vulgo 2.0–3.0 mm diam. Holotypus cultura sicca ex DAOM 231232 ex solo subter Helianthibus isolatus, Geneva, NY, USA (S. Petzolt and G.E. Harman).

Colonies.—On oatmeal agar fast growing, exceeding 100 mm diam in 5 d at 20 6 1 C, richly sporulating, usually in broad compacted zones. On MA moderately fast growing, 22–33 mm diam after 3 d and 44– 59 mm after 4 d at 20 6 1 C. Colonies up to 85 mm diam in 3 d on MA at the optimum temperature of 30–32 C. Maximum temperature for growth 35–36 C. However, conidiation was absent at this temperature. Colonies characterized by densely aggregated conid-

1063

iophores in coalescent pustules, which become greenish olivaceous to glaucous blue-green (28 E–F 7) or grey olivaceous. Conidiophores.—Hyaline, smooth walled, pyramidally verticillately branched in the Pachybasium-type pattern. Conspicuous short sterile appendages visible in young conidiogenous pustules but inconspicuous or absent in older cultures. Conidiophore main axes and first-order branches short and usually 425.5 mm wide, terminal branches 3–7.5 3 3–4 mm. Phialides.—In whorls of 2–5, mostly broadly ampulliform with a short slender neck, 5–6.5 3 (3–)3.5–4 mm in the broadest part, solitary terminal phialides or phialides from less complexly branched conidiophores more elongate, up to 8–12 3 5 mm. The broad and short phialides and branches giving a compact, compressed appearance to the conidiogenous structures. Conidia.— Subglobose or short ellipsoidal in some strains, mostly 2.0–3.0 mm diam, with a minimally protruding basal hilum, smooth-walled, appearing pale grey-green microscopically. Chlamydospores.—Developing in older cultures in the submerged mycelium, subhyaline, intercalary or terminal, solitary, oblong to ellipsoidal or pyriform, (6–)–10 3 4–6 mm. HOLOTY PE: UNITED STATES. NEW YORK: Geneva, New York State Agricultural Experimental Station, isolated from soil in a sunflower field 20 Jun 2000 (S. Petzolt and G. E. Harman). DAOM 231232 (dried culture ex MA3296). Also deposited in MA 3296, CBS 109720. For additional material examined see TABLE I. Phylogeny.—Kullnig-Gradinger et al (2002) found that T. brevicompactum (as ‘‘Trichoderma sp. 1’’) formed a clade together with Hypocrea lutea (Tode: Fr.) Petch and Hypocrea spp. of section Hypocreanum (Bissett 1991a). To test this hypothesis, we amplified and sequenced a 0.2 kb fragment of tef1 from 11 isolates and subjected it together with the ITS1-5.8SrDNA-ITS2 sequences to a combined parsimony analysis with PAUP, using taxa from additional defined phylogenetic clades of Trichoderma as landmarks. The result from this combined analysis (FIG. 3) shows that T. brevicompactum forms a bifurcating clade, of which H. lutea was a sister clade. Bifurcation was due to the two strains from Iran and a strain from Costa Rica, which exhibited the most divergent tef1 haplotypes. Parsimony analysis of ITS1-5.8SrDNA-ITS2 sequences from all 16 T. brevicompactum isolates showed that they formed a homogeneous clade for which the isolates from Peru and Wisconsin formed

1064

MYCOLOGIA

FIG. 1. Morphology of Trichoderma brevicompactum. (A) Colony appearance on malt-extract agar, strain MA3296 (ex-type); (B) postural appearance on oatmeal agar, strain MA 3296 (ex-type); (C) conidiophore, strain MA 4105; (D, E) SEM conidiophore strain D.Z.01; (F) conidia strain MA 3296 (ex-type).

KRAUS

ET AL:

TRICHODERMA BREVICOMPACTUM

1065

FIG. 2. Trichoderma brevicompactum. Conidia and conidiophores drawn from oatmeal agar (A, C) and malt agar (B). A, B strain MA3296 (ex-type), C strain MA4105.

a moderately supported basal branch (FIG. 4). The data provide clear molecular genetic support to treat T. brevicompactum as a new taxon. Physiology.—Statistical analyses on the metabolic data from Biolog microplates were performed to differentiate T. brevicompactum from the most closely related species in the phylogenetic analysis, H. lutea, T. virens and T. harzianum. The results of the cluster analysis based on the 750 nm absorbance readings after 96 h are presented in FIG. 5. The 750 nm readings represent mycelial production resulting from assimilation of the respective C-substrates. Analyses based on the 490 nm readings (mitochondrial activity) gave essentially the same results and are not pre-

sented. The 10 strains of Trichoderma brevicompactum formed a ‘‘monophenetic’’ grouping in the cluster analysis, closer to T. virens and H. lutea than to T. harzianum. The T. harzianum isolates also clustered together. T. virens and H. lutea were not as clearly distinguished in the cluster analysis, and one isolate of H. lutea (CBS 102037) clustered with T. virens strains. Univariate ANOVAs were performed for each of the 95 carbon substrates and the control well that did not contain a carbon source and provided a reference absorbance reading indicating an inability to assimilate substrates. Significant differences among the means for the four species were observed for 84 sub-

2.38 [60.20] 5.39 [60.35] 3.52 [60.23] 2.54 [60.18] 5.50 [60.39] 3.56 [60.24] 2.59 [60.20] 5.70 [60.52] 3.61 [60.24] 2.50 [60.21] 5.18 [60.35] 3.59 [60.25] 2.60 [60.19] 5.67 [60.49] 3.65 [60.20] 2.51 [60.15] 5.99 [60.35] 3.69 [60.20] 2.51 [60.25] 5.80 [60.39] 3.32 [60.21] Conidia diameter [mm] 2.66 [60.18] Phialide length [mm] 5.15 [60.32] Phialide width [mm] 3.41 [60.22]

MA4107 MA3296 (ex-type) MA4103 MA4105 MA3295 MA3294 MA761 MA1374 Strain

Biometric data of T. brevicompactum morphology. Mean value and standard deviation (n 5 30) TABLE II.

2.44 [60.16] 5.39 [60.35] 3.52 [60.23]

MYCOLOGIA

MA4106

1066

strates (probability of type I error P , 0.05). ANOVA results for the statistically significant substrates are shown in TABLE III. Trichoderma brevicompactum had the lowest growth rate (26 C) on 51 (61%) of the statistically significant substrates. This could indicate overall slower growth in this species, or significantly higher optimum temperature (30–32 C) for growth. However, among the readily utilized substrates, T. brevicompactum assimilated glucose at a much higher rate than T. harzianum, which had the highest rate of growth on 34 (40%) of the substrates. T. brevicompactum had the highest assimilation rates for the disaccharides maltose and turanose, D-glucose-1-phosphate, and the purine nucleotide adenosine-59-monophosphate. Absorbance values close to those of the control well indicated an inability of T. brevicompactum to assimilate some C-sources used by one or more of the other species, notably sucrose, and also D-gluconic acid. All four species grew poorly on, or were unable to assimilate, L-alaninamide, putrescine, the methyl ester of D-lactic acid, 2-aminoethanol, uridine, D-melezitose, glucuronamide, N-acetyl-D-mannosamine, N-acetyl-Dgalactosamine, sedoheptulosan, palatinose, a-cyclodextrin and maltitol. A canonical variate analysis was performed on the 750 nm absorbance readings after 96 h incubation for the 18 most significant substrates. All three canonical variates were significant (P , 0.01, TABLE IV). However, 94% of total variation was represented on the first two canonical variates. A plot of the 37 Trichoderma strains representing the four related species on the first two canonical variates is provided in FIG. 6. The plot clearly distinguishes T. brevicompactum from the other species. The first canonical variate distinguishes T. harzianum from the other three species. Examination of the variable loadings from the ‘‘between canonical structure’’ for the first two canonical variables (TABLE V) indicates that T. harzianum is distinguished by faster growth on the organic acids sebacic acid, quinic acid, malic acid, and the mono-methyl ester of succinic acid, the amino acids L-phenylalanine, L-serine and L-alanine, the readily assimilated amino sugar N -acetyl-D-glucosamine, and the amide succinamic acid. Conversely, T. harzianum has slower growth on a wide range of sugars including sucrose, D-raffinose, D-trehalose, D-xylose, D-galactose, stachyose, gentobiose and the glycoside a-methyl-D-glucoside. T. brevicompactum is distinguished from the other three species on the second canonical variate. The variable loadings on the second canonical variate indicate poor assimilation of all of the most highly significant variables in the analysis, with the exception of succinamic acid, which was not assimilated by H. lutea or T. virens. H. lutea and

KRAUS

ET AL:

TRICHODERMA BREVICOMPACTUM

1067

FIG. 3. One of 23 most parsimonious trees obtained by phylogenetic analysis of ITS1 and 2 and tef1 sequences. For GeneBank numbers of T. brevicompactum strains see TABLE I, and for ex-type strains see Kullnig-Gradinger et al 2002. The gene bank accession numbers for H. lutea CBS 102037 ITS1, ITS2 and tef1 gene sequences are AY324175, AY324184 and AY324182, respectively. The numbers given over and below selected branches indicate bootstrap coefficients when gaps were treated as missing data or fifth base, respectively.

T. virens cluster closely in the canonical analysis, indicating similar patterns of substrate utilization in these two species. DISCUSSION

Compared with other new taxa of Trichoderma and Hypocrea described recently (Samuels et al 2000, Bissett et al 2003, Lu et al 2003), T. brevicompactum was found to have a relatively broad distribution. Morphological analysis clearly allies T. brevicompactum as a member of Trichoderma section Pachybasium Sacc., as described by Bissett (1991b), which currently includes most of the described species of the genus

Trichoderma. Species belonging to this section are characterized by relatively short, broad phialides clustered on broad conidiophore branches and often produce sterile elongations of their conidiophores. Within Trichoderma section Pachybasium, T. brevicompactum resembles several other small-spored species, such as T. minutisporum Bissett, with compact, highly branched conidiophores and small ampulliform phialides arising in crowded verticils. It can be distinguished by the presence of subglobose conidia and the formation of mostly fertile conidiophore apices in older cultures. The nearly globose conidia distinguish T. brevicompactum from all other known taxa of Trichoderma section Pachybasium except for T. har-

1068

MYCOLOGIA

FIG. 4. One of four most parsimonious trees obtained by phylogenetic analysis of ITS1 and 2 sequences. Sequences obtained during this study are listed by their GenBank numbers in TABLE I. For ITS1 and 2 sequences of H. lutea see legend to FIG. 3. The numbers given over and below selected branches indicate bootstrap coefficients when gaps were treated as missing data or fifth base, respectively.

zianum. However, the latter species has less compact conidiophores and slightly paler conidia of roughly the same size. Within the variation seen in T. harzianum, the ex-type strain of T. inhamatum Veerkamp & W. Gams (which has been synonymized with T. harzianum; Bissett 1991b, Chaverri et al 2003), has the shortest and most compact phialides and are com-

parable to those of T. brevicompactum but differs from it in having narrower (rarely exceeding 3 mm) and more divergent phialides. The new species therefore broadens the morphological variation seen within this large and heterogeneous section of Trichoderma. Phylogenetically, T. brevicompactum is related closely to H. lutea, which corroborates previous findings

KRAUS

FIG. 5.

ET AL:

TRICHODERMA BREVICOMPACTUM

1069

Cluster analysis of Trichoderma strains based on 750 nm absorbance readings (growth) after 96 h incubation.

(Kullnig-Gradinger et al 2002). The inclusion of additional isolates of H. lutea and its anamorph, Gliocladium viride, in the phylogenetic analysis gave the same result (data not shown). Also, neither the inclusion of additional taxa from ‘‘clade B’’, which contains most of the species with Pachybasium-type morphology, such as T. minutisporum, T. tomentosum, T. oblongisporum and T. fertile (Kullnig-Gradinger et al 2002), nor using T. longibrachiatum instead of H. aureoviridis as outgroup changed the fact that H. lutea remained phylogenetically the sister group of T. brevicompactum (I. Druzhinina unpublished data). The anamorph of Hypocrea lutea (Gliocladium viride) has morphology referable to Gliocladium. Its close phy-

logenetic relationship with species referable to Trichoderma section Pachybasium was unexpected because H. lutea morphologically does not fit Trichoderma and especially does not fit section Pachybasium. The phylogenetic position of H. lutea suggests that the genetic traits responsible for either Gliocladium or Pachybasium anamorph morphology can be lost or gained within a short evolutionary period. The assumption that the development of a typical Pachybasium-like morphology easily can be gained during speciation and is not the result of a long evolutionary development also would explain why Trichoderma section Pachybasium (Bissett 1991b) is paraphyletic (Kindermann et al 1998, Kullnig-Gradinger et al 2002)

1070 TABLE III.

MYCOLOGIA Analysis of variance results for statistically significant substrates Mean1

Rank

Substrate

F-value

Prob.

Brevicompactum

Hazianum

Lutea

Virens

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

sebacic acid L-phenylalanine Quinic acid N-acetyl-D-glucosamine sucrose D-raffinose D-trehalose succinic acid mono-methyl ester D-xylose succinamic acid D-galactose Stachyose L-serine b-methyl-D-glucoside a-D-glucose L-malic acid Gentobiose L-alanine a–D-lactose D-mannose fumaric acid L-pyroglutamic acid Salicin g-amino-butyric acid D-saccharic acid L-erythritol D-melibiose bromosuccinic acid L-proline D-fructose Succinic acid L-lactic acid D-ribose L-alaninamide Putrescine D-lactic acid methyl ester glycyl-L-glutamic acid Tween 80 2-aminoethanol D-gluconic acid Amygdalin Xylitol D-galacturonic acid Uridine L-threonine D-glucuronic acid D-melezitose L-glutamic acid glucuronamide L-ornithine L-fucose L-sorbose N-acetyl-b-D-mannosamine L-asparagine N-acetyl-D-galactosamine

55.09 55.92 45.77 43.09 41.82 38.93 37.58 34.69 34.08 32.65 27.64 27.35 24.61 22.53 21.96 20.30 19.48 19.29 18.46 18.34 18.13 17.63 14.70 14.18 14.15 14.04 13.99 13.53 13.09 12.29 11.38 11.11 10.92 10.79 10.51 10.49 10.38 9.99 9.65 9.40 9.37 9.22 9.01 8.61 8.37 8.16 7.82 7.74 7.72 7.63 7.43 7.05 6.94 6.93 6.93

,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 ,.0001 .0001 .0001 .0001 .0001 .0002 .0002 .0003 .0003 .0004 .0005 .0005 .0005 .0006 .0009 .0010 .0010 .0010

0.314 0.334 0.615 0.869 0.141 0.363 0.662 0.245 0.731 0.374 0.761 0.495 0.277 0.366 0.718 0.293 0.801 0.578 0.442 0.861 0.368 0.564 0.444 0.572 0.205 0.611 0.585 0.281 0.251 0.761 0.382 0.194 0.536 0.166 0.157 0.177 0.276 0.391 0.148 0.157 0.762 0.231 0.535 0.156 0.233 0.312 0.137 0.512 0.126 0.455 0.448 0.372 0.122 0.463 0.127

1.654 1.432 1.381 2.332 0.251 0.271 0.425 0.770 0.324 0.539 0.521 0.333 0.659 0.389 0.443 0.475 0.500 1.013 0.458 0.538 0.560 1.083 0.700 1.300 0.268 0.601 0.551 0.360 0.510 0.503 0.569 0.297 0.447 0.255 0.240 0.290 0.481 0.448 0.239 0.326 1.110 0.595 0.994 0.239 0.345 0.399 0.199 0.694 0.196 0.652 0.587 0.551 0.195 0.632 0.181

0.844 0.574 0.566 1.276 0.879 0.578 0.804 0.494 0.754 0.201 0.798 0.721 0.472 0.821 0.822 0.505 0.908 0.758 0.645 0.878 0.566 0.827 0.596 0.596 0.207 0.631 0.643 0.342 0.306 0.818 0.464 0.318 0.642 0.265 0.248 0.297 0.455 0.546 0.208 0.289 0.685 0.347 0.596 0.251 0.278 0.529 0.210 0.438 0.176 0.363 0.845 0.503 0.167 0.541 0.167

1.005 0.564 0.629 1.227 1.219 0.938 1.124 0.514 0.985 0.172 1.213 1.115 0.483 0.894 1.023 0.384 1.079 0.621 0.865 1.138 0.453 0.574 0.613 0.884 0.457 1.285 1.054 0.249 0.307 1.094 0.377 0.242 0.996 0.266 0.207 0.192 0.377 0.528 0.195 0.302 1.196 0.585 0.457 0.214 0.235 0.468 0.200 0.536 0.206 0.558 0.846 0.653 0.168 0.535 0.174

KRAUS TABLE III.

ET AL:

1071

TRICHODERMA BREVICOMPACTUM

Continued Mean1

Rank 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 1

Substrate L-alanyl-glycine D-sorbitol Maltose L-aspartic acid D-glucosamine sedoheptulosan g-hydroxybutyric acid D-malic acid Dextrin D-cellobiose P-hydroxyphenylacetic acid b-hyroxybutyric acid Palatinose D-arabitol D-arabinose D-mannitol L-arabinose maltotriose a–D-glucose-1-phosphate D-psicose Lactulose Adonitol Arbutin Adenosine-59-monophosphate Glycogen a–ketoglutaric acid D-tagatose 2-keto-D-gluconic acid water (control) Turanose

F-value

Prob.

Brevicompactum

Hazianum

Lutea

Virens

6.60 6.32 5.95 5.94 5.91 5.70 5.53 5.49 5.44 5.32 5.29 5.15 5.12 4.88 4.85 4.50 4.48 4.24 4.20 4.16 4.15 4.05 4.04 3.98 3.89 3.87 3.58 3.56 3.41 3.21

.0013 .0017 .0023 .0024 .0024 .0029 .0035 .0036 .0037 .0042 .0043 .0049 .0051 .0065 .0066 .0094 .0096 .0122 .0127 .0132 .0134 .0148 .0149 .0158 .0175 .0178 .0241 .0247 .0288 .0356

0.440 0.509 0.447 0.356 0.273 0.162 0.269 0.263 0.716 0.806 0.235 0.274 0.167 0.814 0.335 0.696 0.361 0.556 0.366 0.182 0.415 0.232 0.584 0.555 0.458 0.276 0.235 0.441 0.120 0.320

0.647 0.920 0.444 0.427 0.294 0.210 0.334 0.342 0.670 0.846 0.303 0.436 0.210 0.746 0.281 0.777 0.436 0.542 0.284 0.273 0.355 0.299 0.793 0.422 0.754 0.342 0.285 0.550 0.170 0.236

0.596 0.652 0.259 0.406 0.365 0.228 0.290 0.385 0.767 0.927 0.227 0.398 0.252 0.960 0.492 0.877 0.405 0.507 0.262 0.306 0.446 0.268 0.668 0.382 0.588 0.342 0.275 0.644 0.162 0.267

0.550 1.185 0.438 0.331 0.544 0.187 0.199 0.344 0.963 1.168 0.208 0.365 0.227 1.093 0.698 1.093 0.613 0.685 0.236 0.330 0.554 0.471 0.716 0.497 0.669 0.281 0.366 0.504 0.156 0.236

Absorbance at 750 nm. Highest value in bold font, lowest value in italics, for statistically significant substrates.

and separable into two major phylogenetic groups (which have been termed clade A and clade B; Kullnig-Gradinger et al 2002). One of these clades (A) contains most members of section Trichoderma, whereas the other (B) forms a highly diverse but monophyletic sister group. Because of its convenience, the morphological descriptor ‘‘pachybasiumlike’’ is still used to describe a characteristic type of conidiophore. However, the large group of species in clade B warrants a formal taxonomic status within Hypocrea/Trichoderma that takes evolutionary relationships into account. Of note, the phylogenetic analysis clearly separated TABLE IV.

1 2 3

H. lactea and H. citrina from T. brevicompactum and H. lutea, indicating that these two taxon pairs are phylogenetically less close than previously thought (Kullnig-Gradinger et al 2002). Isolates investigated in this study came from widely diverse geographic areas such as North, Central and South America, the Caribbean, the Persian Gulf region, southern India and Papua New Guinea (see TABLE I). These isolates came from hot and humid locales (Papua New Guinea, Mexico, India, Costa Rica, Union Island, Colombia and Peru). In contrast, the isolates from the northern United States came from areas with cooler temperatures and the isolates from

Significance tests of canonical variates Eigenvalue

Proportion

Likelihood ratio

F-value

df

Prob.

36.59 11.71 3.49

0.71 0.23 0.07

0.00047 0.0175 0.2228

10.89 6.55 3.93

54, 48.5 34, 34 16, 18

,0.0001 ,0.0001 0.0033

1072

MYCOLOGIA TABLE V. structure

FIG. 6. Canonical variate analysis of 750 nm absorbance readings after 96 h.

Iran come from areas with little humidity. Trichoderma brevicompactum therefore seems to be a cosmopolitan versatile species, capable of adapting to comparatively diverse climatic conditions. Within the 13 strains investigated, four ITS and seven tef1 haplotypes could be distinguished and all of the variations in ITS occurred in ITS1. In this regard, T. brevicompactum belongs to the genetically more variable species of Trichoderma (Kubicek et al 2003). It should be noted that the strains from Costa Rica and Iran exhibited the highest genetic distance to the other strains. Since a high genetic variability and basal clustering in phylogenetic trees is an indication for an origin of a species, the center of diversification must have been an area from which strains moved to both western Asia and Central America and from which strains have not yet been found. However, although this hypothesis is based on several strains from diverse locations, the low total number of strains might mask the true origin of this species. In spite of this genetic variability, individual isolates of T. brevicompactum display an excellent consistency in their physiological properties. Kubicek et al (2003), investigating the physiological variation of about 90 strains of Trichoderma comprising 16 taxa, showed that physiological clustering is not necessarily a consistent criterion for species identification in Trichoderma and that some taxa, such as T. harzianum, are physiologically diverse. The finding of a consistency in nutrient assimilation among geographically diverse isolates of T. brevicompactum indicates that T. brevicompactum occupies an ecological niche for which a defined range of nutritional abilities is essential. In this regard, T. brevicompactum used most carbohydrates (mono-, oligo- and polysaccharides) and carbohydrate-related compounds (polyols) at higher rates than other Trichoderma taxa examined that used amino acids and organic acids much less

Variable loadings from the between canonical

Substrate

Canonical variate 1

Canonical variate 2

N-acetyl-D-glucosamine D-galactose Gentobiose a-D-glucose b-methyl-D-glucoside D-raffinose Stachyose Sucrose D-trehalose D-xylose L-malic acid Quinic acid Sebacic acid succinamic acid succinic acid mono-methyl ester L-alanine L-phenylalanine L-serine

0.944 20.817 20.925 20.907 20.538 20.685 20.728 20.507 20.830 20.964 0.536 0.963 0.815 0.844 0.807 0.960 0.955 0.786

0.315 0.492 0.379 0.420 0.811 0.707 0.661 0.860 0.542 0.256 0.630 0.047 0.560 20.476 0.586 0.245 0.267 0.618

efficiently. Trichoderma/Hypocrea, it is speculated, originally had been primarily mycoparasitic fungi, which later acquired the ability to follow their hosts into their habitat (decaying wood) and compete for their nutrients (Klein and Eveleigh 1998). The physiological properties of T. brevicompactum, in combination with its comparably high temperature for optimal growth (30–32 C), may indicate that this fungus is especially adapted to the degradation of plant polysaccharides at or close to the soil surface.

ACKNOWLEDGMENTS

This work was supported partly by Hungarian Science Foundation Grant OTKA T032690 and Hungarian Ministry of Education Grant FKFP 516/1999 to GS and Austrian Science Fund grants P-12748-MOB and FWF P-16601 to CPK. The authors are grateful to Professor Gary Harman (Cornell University, Geneva, New York), Dr Mette Lu¨beck (The Royal Veterinary and Agricultural University, Frederiksberg, Denmark) and Dr Martha Christensen (University of Wyoming, Laramie, Wyoming) for the gift of Trichoderma isolates and to Dr Sergio Orduz and Lilliana Hoyos (Corporacio´n para Investigaciones Biolo´gicas, Medellı´n, Colombia) for permission to publish ITS 1 and 2 and tef1 sequence data for the Colombian isolate of T. brevicompactum. The authors also thank Carol Ann Nolan, Maria Voloaca and Greg Chang for their contributions to the physiological and molecular studies undertaken at ECORC. Publication No. 2003-370 of the Eastern Cereal and Oilseed Research Centre.

KRAUS

ET AL:

TRICHODERMA BREVICOMPACTUM

LITERATURE CITED

Beckett A, Read NN. 1986. Low temperature scanning electron microscopy. In: Aldrich HC, Todd WC, eds. Ultrastructure techniques for microorganisms. New York and London: Plenum Press. p 45–86. Bissett J. 1991a. A revision of the genus Trichoderma. II. Infrageneric classification. Can J Bot 69:2357–2372. . 1991b. A revision of the genus Trichoderma. III. Section Pachybasium. Can J Bot 69:2373–2417. , Szakacs G, Nolan CA, Druzhinina IS, Kullnig-Gradinger CM, Kubicek CP. 2003. Seven new taxa of Trichoderma from Asia. Can J Bot 81:570–586. Chaverri P, Castlebury LA, Samuels GJ, Geiser D. 2003. Multilocus phylogenetic structure within the Trichoderma harzianum/Hypocrea lixii complex. Mol Phylogenet Evol 27:302–313. Gams W, Hoekstra ES, Aptroot A. 1998. CBS Course of Mycology. 4th ed. Centraalbureau foor Schimmelcultures, Baarn, The Netherlands. Kindermann J, El-Ayouti Y, Samuels GJ, Kubicek CP. 1998. Phylogeny of the genus Trichoderma based on sequence analysis of the internal transcribed spacer region 1 of the rDNA clade. Fungal Genet Biol 24:298–309. Klein D, Eveleigh DE. 1998. Ecology of Trichoderma. In: Kubicek CP, Harman GE, eds. Trichoderma and Gliocladium. Basic biology, taxonomy and genetics. London, UK: p 57–74. Kornerup A, Wanscher JH. 1978. Methuen handbook of colour. 3rd ed. Revised by D. Pavey. London, UK: Methuen. 252 p. Kubicek CP, Bissett C, Druzhinina I, Kullnig-Gradinger CM, Szakacs G. 2003. Genetic and metabolic diversity of Trichoderma: a case study on South East Asian isolates. Fungal Genetics Biol 38:310–319. Kullnig C, Szakacs G, Kubicek CP. 2000. Molecular identification of Trichoderma species from Russia, Siberia and the Himalaya. Mycol Res 104:1117–1125.

1073

Kullnig-Gradinger CM, Szakacs G, Kubicek CP. 2002. Phylogeny and evolution of the fungal genus Trichoderma—a multigene approach. Mycol Res 106:757–767. Lu BS, Druzhinina I, Chaverri P, Fallah P, Gradinger C, Kubicek CP, Samuels GJ. 2003. Hypocrea/Trichoderma species with pachybasium-like conidiophores: teleomorphs for T. minutisporum and T. polysporum, and their newly discovered relatives. Mycologia 96:310–342. Nicholas KB, Nicholas HB Jr. 1997. Genedoc: a tool for editing and annotating multiple sequence alignments. http://www.psc.edu/biomed/genedoc. Rolf, FJ. 1997. NTSYSpc Numerical Taxonomy and Multivariate Analysis System, version 2.00. Exeter Software, Setauket, New York, 31 p. Samuels GJ. 1996. Trichoderma: a review of biology and systematics of the genus. Mycol Res 100:923–935. Samuels GJ, Pardo-Schultheiss R, Hebbar KP, Lumsden RD, Bastos CN, Costa JC, Bezerra JL. 2000. Trichoderma stromaticum sp. nov., a parasite of the cacao witches broom pathogen. Mycol Res 104:760–764. SAS Institute Inc. 1989. SAS/STAT user’s guide, version 6. 4th ed. Vol. 1. Cay, North Carolina: SAS Institute Inc. 943 p. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. 1997. The Clustal X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 24:4876–4882. Turner D, Kovacs W, Kuhls K, Lieckfeldt E, Peter B, ArisanAtac I, Strauss J, Samuels, GJ, Bo¨rner T, Kubicek CP. 1997. Biogeography and phenotypic variation in Trichoderma section Longibrachiatum and associated Hypocrea species. Mycol Res 101:449–459. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ eds. PCR protocols. A guide to methods and applications. New York: Academic Press. p 315–322.