cryptic speciation in fusarium subglutinans
TRANSCRIPT
1032
Mycologia, 94(6), 2002, pp. 1032–1043.q 2002 by The Mycological Society of America, Lawrence, KS 66044-8897
Cryptic speciation in Fusarium subglutinans
Emma T. Steenkamp1
Brenda D. WingfieldDepartment of Genetics, Forestry and AgriculturalBiotechnology Institute, University of Pretoria, Pretoria0002, South Africa
Anne E. DesjardinsMycotoxin Research Unit, National Center forAgricultural Utilization of Research, USDA,Agricultural Research Services, 1815 N UniversityStreet, Preoria, Illinois 61604
Walter F.O. MarasasPROMEC, Medical Research Council, P.O. Box19070, Tygerberg 7505, South Africa
Michael J. WingfieldForestry and Agricultural Biotechnology Institute,University of Pretoria, Pretoria 0002, South Africa
Abstract: Fusarium isolates that form part of theGibberella fujikuroi species complex have been classi-fied using either a morphological, biological, or phy-logenetic species concept. Problems with the taxon-omy of Fusarium species in this complex are mostlyexperienced when the morphological and biologicalspecies concepts are applied. The most consistentidentifications are obtained with the phylogeneticspecies concept. Results from recent studies have pre-sented an example of discordance between the bio-logical and phylogenetic species concepts, where agroup of F. subglutinans sensu stricto isolates, i.e., iso-lates belonging to mating population E of the G. fu-jikuroi complex, could be sub-divided into more thanone phylogenetic lineage. The aim of this study wasto determine whether this sub-division representedspecies divergence or intraspecific diversity in F.subglutinans. For this purpose, we included 29 F.subglutinans isolates belonging to the E-mating pop-ulation that were collected from either maize or te-osinte, from a wide geographic range. DNA sequencedata for six nuclear regions in each of these isolateswere obtained and used in phylogenetic concordanceanalyses. These analyses revealed the presence of twomajor groups representing cryptic species in F. sub-
Accepted for publication April 28, 2002.1 Corresponding author, Email: [email protected]; current address:Department of Biology, University of York, PO Box 373, York YO105YW, UK.
glutinans. These cryptic species were further sub-di-vided into a number of smaller groups that appearto be reproductively isolated in nature. This suggestsnot only that the existing F. subglutinans populationsare in the process of divergence, but also that eachof the resulting lineages are undergoing separationinto distinct taxa. These divergences did not appearto be linked to geographic origin, host, or phenotyp-ic characters such as morphology.
Key Words: Fusarium subglutinans, maize, repro-ductive isolation, teosinte
INTRODUCTION
Gibberella fujikuroi (Sawada) Wollenw. is a speciescomplex that encompasses many Fusarium species(Nirenberg and O’Donnell 1998, O’Donnell and Cig-elnik 1997, O’Donnell et al 1998a, 2000a, Steenkampet al 1999, 2000a, 2001). In the global environment,species in this complex are important, because oftheir association with diseases of agronomically im-portant plants (Correll et al 1991, Leslie 1995, Leslieet al 1990, Sun and Snyder 1981, Varma et al 1974,Ventura et al 1993). These fungi also affect humanand animal health, since many species produce toxicsecondary metabolites such as moniliformin, beau-vericin, fumonisin, fusaproliferin, and fusaric acid(Leslie et al 1992, Logrieco et al 1993, Marasas et al1983, Sewram et al 1999a, 1999b, Shephard et al1999, Vesonder et al 1995).
The taxonomy of Fusarium species in the G. fujiku-roi complex has been subject to much controversy(Leslie 1995). This is mainly due to a lack of consen-sus among researchers on how to define a morpho-logical species for the fungi in this complex (Gerlachand Nirenberg 1982, Nirenberg and O’Donnell 1998,Snyder and Hansen 1945). In an attempt to resolvethis problem, a biological species concept (Dobzhan-sky 1937, Mayr 1940) was introduced, whereby eightbiological species have been identified (Britz et al1999, Hsieh et al 1977, Klittich and Leslie 1992, Kuhl-man 1982, Leslie 1991, 1995, 1996). These biologicalspecies were designated as mating populations A toH, where mating population E, for example, repre-sents F. subglutinans (Wollenw. & Reinking) Nelson,Toussoun & Marasas sensu stricto (Britz et al 1999,Hsieh et al 1977, Klittich and Leslie 1992, Kuhlman
1033STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
1982, Leslie 1991, 1995, 1996). The eight biologicalspecies, however, exclude the more than 80% of spe-cies in this complex with no apparent sexual repro-ductive cycle. Currently, the only method for classi-fying all the fungal strains in the G. fujikuroi speciescomplex is through the application of the phyloge-netic species concept (Cracraft 1983, Taylor et al2000). With this method, fungi in the G. fujikuroicomplex are classified into more than 40 differentphylogenetic species (O’Donnell et al 1998a, Steen-kamp et al 1999, 2000a).
In a recent study, ten additional phylogeneticallydistinct species in the G. fujikuroi complex were re-ported (O’Donnell et al 2000a). Among these was aF. subglutinans strain associated with maize in SouthAfrica, that had previously been classified in matingpopulation E of the G. fujikuroi species complex us-ing the biological species concept (Steenkamp et al1999). Application of the phylogenetic species con-cept has, however, indicated that this biological spe-cies is subdivided into more than one phylogeneticlineage (O’Donnell et al 2000a, Steenkamp et al1999, 2001). These lineages may either reflect speciesdivergence within F. subglutinans or are simply theresult of intraspecific diversity. However, all of thestudies dealing with the molecular classification ofstrains representing F. subglutinans were based onsmall sets of different isolates (O’Donnell et al 2000a,Steenkamp et al 1999, 2001) and no definite conclu-sions could be drawn on the taxonomic status ofthese fungi. Clarification of the relationships amongdifferent F. subglutinans isolates would undoubtedlyshed light on how the phylogenetic species conceptmight influence our interpretation of the biologicalspecies concept in the G. fujikuroi species complex.The aim of this study was, therefore, to address thephylogenetic status of members of F. subglutinans by(i) including strains isolated from maize and its wildteosinte relatives from a wide geographic range; (ii)obtaining DNA sequence data from six nuclear re-gions for these strains; and (iii) using phylogeneticconcordance analysis (Avise and Ball 1990, Taylor etal 2000) to determine whether these isolates are in-terbreeding in nature and, if not, to define sub-groups within F. subglutinans.
MATERIAL AND METHODS
Fungal isolates. Twenty-nine F. subglutinans sensu strictoisolates belonging to the E-mating population of the G. fu-jikuroi species complex were included in this study (TABLE
I). The six South African, ten United States and six Mexicanstrains were isolated from maize. The remaining five Mex-ican strains and the two Guatemalan strains were isolatedfrom teosinte. For outgroup purposes we also included two
isolates of F. circinatum Nirenberg et O’Donnell that wereisolated from pines.
DNA isolation, PCR amplification and sequencing. DNAwas isolated using a CTAB (N-cetyl-N,N,N-trimethyl-ammo-nium bromide) extraction method (Steenkamp et al 1999).A portion of three nuclear genes, histone H3 (Glass andDonaldson 1995), calmodulin (Carbone and Kohn 1999)and b-tubulin (Glass and Donaldson 1995), were amplifiedfrom all 31 isolates. We also used an additional set of prim-ers that amplify three unlinked nuclear regions of unknownfunction (H. Britz unpubl). The first primer set is HB9-a(59-tcaatacccctcgcctagaa-39) and HB9-b (59-gaccacagcctcga-gaacat-39), the second is HB14-a (59-ttccaccatgagaggaaaccc-39) and HB14-b (59-ccattgccaatcttgatcct-39), and the thirdHB26-a (59-gacttgagtatctgcactgc-39) and HB26-b (59-gaatgt-actactcgacgtcg-39).
For amplification of all these loci, the PCR mixture con-tained 1 mM deoxynucleotide triphosphates (0.25 mMeach), 2.5 mM MgCl2, 0.2 mM of each primer, 0.25 ng/mLDNA, 0.05 U/mL of Super-Therm DNA polymerase [South-ern Cross biotechnology (Pty.) Ltd., Cape Town, South Af-rica] and 1 3 Super-Therm reaction buffer. PCR-cyclingconditions were as follows: denaturation at 92 C for 20 s,annealing for 20 s at 55 C (calmodulin, tubulin, and his-tone) or 47 C (HB9, HB14, and HB26), and elongation for20 s at 72 C. This was repeated 30 times and was precededby an initial denaturation at 92 C for 1 min and followedby a final elongation step at 72 C for 5 min.
After PCR, the products were purified with a QIAquickPCR Purification Kit (Qiagen GmbH, Hilden, Germany)and sequenced in both directions with the respective prim-ers. Reactions were performed on an ABI PRISMy 377 au-tomated DNA sequencer, with an ABI PRISMy Dye Ter-minator Cycle Sequencing Ready Reaction Kit (Perkin El-mer, Warrington, United Kingdom). Sequences were ana-lyzed with Sequence Navigator version 1.0.1.y (PerkinElmer Applied BioSystems, Inc., Foster City, California).
Phylogenetic analyses. The data sets obtained for eachprimer set were aligned manually by inserting gaps(TreeBase accession numbers S789 and M1249). Phyloge-netic analyses were performed with PAUP (PhylogeneticAnalysis Using Parsimony) version 4.0b1 (Swofford 1998),with gaps treated as fifth characters in heuristic parsimonysearches, and with tree-bisection-reconnection (TBR)branch swapping and MULTREES (saving of all optimaltrees) effective. For bootstrap analyses 1000 replicationswere performed.
RESULTS
PCR amplification and sequencing. With the primersused, we were able to amplify and sequence 480 basepairs (bp), 456 bp, and 332 bp of the calmodulin, b-tubulin, and histone H3 genes, respectively. For thethree regions of unknown function, 250 bp, 235 bp,and 236 bp were sequenced for HB9, HB14, andHB26, respectively. Of the 1989 nucleotides (nc) se-quenced, 17 nc (0.9%) were polymorphic in the dif-
1034 MYCOLOGIAT
AB
LE
I.H
osts
,geo
grap
hic
orig
ins
and
sour
ces
ofth
eFu
sari
um
subg
luti
nan
san
dF.
circ
inat
um
isol
ates
used
inth
isst
udy
Isol
atea
Hos
tb
Geo
grap
hic
orig
inc
Sour
ced
Gen
Ban
kac
cess
ion
No.
e
MR
C11
5Z
.m
ays
Eas
tern
Cap
e,So
uth
Afr
ica
W.F
.O.
Mar
asas
(Ste
enka
mp
etal
1999
)A
F150
843,
AF3
7401
5,A
F374
036,
AF3
7408
4,A
F374
116,
AF3
7414
8M
RC
714
Z.
may
sN
orth
ern
Prov
ince
,Sou
thA
f-ri
caW
.F.O
.M
aras
as(S
teen
kam
pet
al19
99)
AF1
5084
1,A
F366
548,
AF3
6652
0,A
F374
085,
AF3
7411
7,A
F374
149
MR
C75
6Z
.m
ays
Mpu
mal
anga
,Sou
thA
fric
aW
.F.O
.M
aras
as(S
teen
kam
pet
al20
01)
AF1
5083
9,A
F366
549,
AF3
6652
7,A
F374
086,
AF3
7411
8,A
F374
150
MR
C83
7Z
.m
ays
Eas
tern
Cap
e,So
uth
Afr
ica
W.F
.O.
Mar
asas
(Ste
enka
mp
etal
1999
)A
F150
840,
AF3
7401
6,A
F374
037,
AF3
7408
7,A
F374
119,
AF3
7415
1M
RC
1077
Z.
may
sE
aste
rnC
ape,
Sout
hA
fric
aW
.F.O
.M
aras
as(S
teen
kam
pet
al19
99)
AF1
5083
7,A
F374
017,
AF3
7403
8,A
F374
088,
AF3
7412
0,A
F374
152
MR
C10
84Z
.m
ays
Eas
tern
Cap
e,So
uth
Afr
ica
W.F
.O.
Mar
asas
(Ste
enka
mp
etal
2001
)A
F150
838,
AF3
6655
0,A
F366
521,
AF3
7408
9,A
F374
121,
AF3
7415
3M
RC
6483
,M36
96,K
SU99
0Z
.m
ays
Illin
ois,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F150
845,
AF3
6655
2,A
F366
522,
AF3
7410
0,A
F374
132,
AF3
7416
4M
RC
6512
,M36
93,K
SU21
92Z
.m
ays
Illin
ois,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F15
0844
,AF3
6655
1,A
F366
525,
AF3
7409
9,A
F374
131,
AF3
7416
3K
SU43
4,M
6496
Z.
may
sK
ansa
s,U
nit
edSt
ates
J.F.
Les
lie(Y
anet
al19
93)
AF3
7406
1,A
F374
018,
AF3
7403
9,A
F374
090,
AF3
7412
2,A
F374
154
KSU
507,
M51
19Z
.m
ays
Kan
sas,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F374
062,
AF3
7401
9,A
F374
040,
AF3
7409
1,A
F374
123,
AF3
7415
5K
SU73
1,M
5126
Z.
may
sK
ansa
s,U
nit
edSt
ates
J.F.
Les
lie(Y
anet
al19
93)
AF3
7406
3,A
F374
020,
AF3
7404
1,A
F374
092,
AF3
7412
4,A
F374
156
KSU
993,
M36
98Z
.m
ays
Illin
ois,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F374
064,
AF3
7402
1,A
F374
042,
AF3
7409
3,A
F374
125,
AF3
7415
7K
SU12
57Z
.m
ays
Kan
sas,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F374
065,
AF3
7402
2,A
F374
043,
AF3
7409
4,A
F374
126,
AF3
7415
8K
SU14
17Z
.m
ays
Kan
sas,
Un
ited
Stat
esJ.
F.L
eslie
(Yan
etal
1993
)A
F374
066,
AF3
7402
3,A
F374
044,
AF3
7409
5,A
F374
127,
AF3
7415
9K
SU29
21,M
3763
Z.
may
sO
hio
,U
nit
edSt
ates
J.F.
Les
lie(Y
anet
al19
93)
AF3
7406
8,A
F374
025,
AF3
7404
6,A
F374
097,
AF3
7412
9,A
F374
161
KSU
3815
,M85
1N
/aN
/aJ.
F.L
eslie
(Yan
etal
1993
)A
F374
069,
AF3
7402
6,A
F374
047,
AF3
7409
8,A
F374
130,
AF3
7416
2M
RC
7786
,Fst
9Z
.di
plop
eren
nis
Jalis
co,M
exic
oA
.E.
Des
jard
ins
AF3
7405
0,A
F374
008,
AF3
7402
9,A
F374
073,
AF3
7410
5,A
F374
137
MR
C77
87,F
st10
Z.
dipl
oper
enn
isJa
lisco
,Mex
ico
A.E
.D
esja
rdin
s(S
teen
kam
pet
al20
01)
AF3
7405
1,A
F366
544,
AF3
6652
3,A
F374
074,
AF3
7410
6,A
F374
138
MR
C77
90,F
st13
,M77
94Z
.lu
xuri
ans
Ch
iqui
mul
a,G
uate
mal
aA
.E.
Des
jard
ins
(Des
jard
ins
etal
2000
)A
F374
052,
AF3
7400
9,A
F374
030,
AF3
7407
5,A
F374
107,
AF3
7413
9M
RC
7794
,Fst
17,M
7799
Z.
may
sss
p.hu
ehu
eten
-an
gen
sis
Hue
hue
ten
en,
Gua
tem
ala
A.E
.D
esja
rdin
s(D
esja
rdin
set
al20
00)
AF3
7405
3,A
F374
010,
AF3
7403
1,A
F374
076,
AF3
7410
8,A
F374
140
1035STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
TA
BL
EI.
Con
tin
ued
Isol
atea
Hos
tb
Geo
grap
hic
orig
inc
Sour
ced
Gen
Ban
kac
cess
ion
No.
e
MR
C77
99,F
st22
Z.
may
sss
p.m
exic
ana
Dur
ango
,M
exic
oA
.E.
Des
jard
ins
AF3
7405
4,A
F374
011,
AF3
7403
2,A
F374
077,
AF3
7410
9,A
F374
141
MR
C78
03,F
st26
Z.
may
sss
p.m
exic
ana
Mic
hoa
can
,Mex
ico
A.E
.D
esja
rdin
s(S
teen
kam
pet
al20
01)
AF3
7405
5,A
F366
545,
AF3
6652
6,A
F374
078,
AF3
7411
0,A
F374
142
MR
C78
17,F
st40
Z.
may
sss
p.pa
rvig
lum
isG
uerr
ero,
Mex
ico
A.E
.D
esja
rdin
sA
F374
056,
AF3
6654
6,A
F366
524,
AF3
7407
9,A
F374
111,
AF3
7414
3M
RC
7828
,Fst
51,M
8372
Z.
may
sss
p.m
exic
ana
Texc
oco,
Mex
ico
A.E
.D
esja
rdin
s(D
esja
rdin
set
al20
00;
Stee
nka
mp
etal
2001
)
AF3
7405
7,A
F366
547,
AF3
6652
9,A
F374
080,
AF3
7411
2,A
F374
144
MR
C78
33,F
st54
,M83
75Z
.m
ays
ssp.
may
sTe
xcoc
o,M
exic
oA
.E.
Des
jard
ins
AF3
7405
8,A
F374
012,
AF3
7403
3,A
F374
081,
AF3
7411
3,A
F374
145
MR
C78
38,F
st58
,M83
77Z
.m
ays
ssp.
may
sTe
xcoc
o,M
exic
oA
.E.
Des
jard
ins
AF3
7405
9,A
F374
013,
AF3
7403
4,A
F374
082,
AF3
7411
4,A
F374
146
MR
C78
49,F
st69
,M83
80Z
.m
ays
ssp.
may
sTe
xcoc
o,M
exic
oA
.E.
Des
jard
ins
(Des
jard
ins
etal
2000
)A
F345
060,
AF3
7401
4,A
F374
035,
AF3
7408
3,A
F374
115,
AF3
7414
7M
RC
7861
,M39
35Z
.m
ays
ssp.
may
sN
/a,
Mex
ico
A.E
.D
esja
rdin
sA
F374
049,
AF3
7400
7,A
F374
028,
AF3
7407
2,A
F374
104,
AF3
7413
6M
RC
7862
,M38
69Z
.m
ays
ssp.
may
sN
/a,
Mex
ico
A.E
.D
esja
rdin
sA
F374
048,
AF3
7400
6,A
F374
027,
AF3
7407
1,A
F374
103,
AF3
7413
5M
RC
7870
,Fsp
34P
inu
ssp
p.C
alif
orn
ia,U
nit
edSt
ates
T.R
.G
ordo
n(D
esja
rdin
set
al20
00)
AF3
7407
0,A
F366
555,
AF3
6652
8,A
F374
101,
AF3
7413
3,A
F374
165
MR
C62
13P
inu
ssp
p.M
pum
alan
ga,S
outh
Afr
ica
A.
Vilj
oen
(Bri
tzet
al19
99)
AF1
5084
9,A
F366
553,
AF3
6653
1,A
F374
102,
AF3
7413
4,A
F374
166
aM
RC
-isol
ates
are
mai
nta
ined
atPR
OM
EC
,M
edic
alR
esea
rch
Cou
nci
l,T
yger
berg
,So
uth
Afr
ica;
KSU
-isol
ates
atth
eD
epar
tmen
tof
Plan
tPa
thol
ogy,
Kan
sas
Stat
eU
niv
ersi
ty,M
anh
atta
n,K
ansa
s;Fs
t-iso
late
sat
NC
AU
R,U
SDA
,Peo
ria,
Illin
ois;
Fsp-
isol
ates
atth
eD
epar
tmen
tofP
lan
tPat
hol
ogy,
Un
iver
sity
ofC
alif
orn
ia,D
avis
,Cal
ifor
nia
;an
dM
-isol
ates
atth
eFu
sari
umR
esea
rch
Cen
ter,
Pen
nsy
lvan
iaSt
ate
Un
iver
sity
,Un
iver
sity
Park
,Pe
nn
sylv
ania
.b
N/a
,in
form
atio
non
the
plan
th
ost
from
wh
ich
the
stra
inw
asis
olat
edis
not
avai
labl
e.c
N/a
,ex
act
loca
tion
ofis
olat
ion
ofth
est
rain
isn
otav
aila
ble.
dR
efer
ence
sof
stud
ies
wh
ere
isol
ates
wer
eus
edin
sexu
alco
mpa
tibi
lity
test
sar
ein
dica
ted
inpa
ren
thes
es.
eG
enB
ank
Acc
essi
onn
umbe
rsfo
rth
ese
quen
ces
for
each
ofsi
xlo
cian
alyz
edar
elis
ted
inth
efo
llow
ing
orde
r:h
isto
ne
H3,
b-tu
bulin
,ca
lmod
ulin
,HB
14,H
B9
and
HB
26.
1036 MYCOLOGIA
TABLE II. Summary of polymorphic nucleotides (nc) in the six nuclear regions sequenced among the Fusarium subglutinansisolates from maize and teosinte
Isolates
Polymorphisms (nc)a
CAL
33 377
TUB
144 204 285
HIS
27 99 275 286 351 426
HB14
99 230 231 233
HB26
69
HB9
101 204Geno-type
MRC7828, MRC7833,MRC837, MRC1077,MRC1084 T C A T T C C T G C T T 2 2 2 A C C 1-1
MRC7786MRC7803
TT
CC
AA
TT
TC
CC
TT
CC
GG
CC
CC
TT
22
22
22
AA
CC
CC
1-21-3
MRC7862, MRC7861,MRC7838, MRC7849,MRC756 T C G T T C C T G C T T 2 2 2 A C C 1-4
MRC7787, MRC7790,MRC7794, MRC7799,MRC115, KSU1257,KSU1417, KSU3815 C C A C T C C T A T T C T A A G T G 2-1
KSU434, KSU731 C T A C T T C T A T T C T A A G T G 2-2MRC7817, MRC714,
MRC6512, MRC6483,KSU507, KSU993,KSU2921 C T A C T C C T A T T C T A A G T G 2-3
a Numbering of polymorphic nucleotides corresponds with their nucleotide positions in the respective DNA sequences.Horizontal lines (2) indicate nucleotide deletions.
TABLE III. Number of polymorphic and parsimony infor-mative characters, as well as the actual length of trees, gen-erated from the individual and combined sequence datasets
LocusInformativecharacters
Polymorphiccharacters
Treelength
CalmodulinHistone H3b-tubulinHB9HB14HB26Combined
262241
17
263241
18
262241
17
ferent F. subglutinans strains and none of the siteshad more than two different nucleotide characterstates. Among these strains, between one and sixpolymorphic nucleotides in each of the six regionswere identified (TABLE II). Upon combination of thepolymorphisms for each individual, we recognizedseven different genotypes within the set of 29 F. sub-glutinans isolates (TABLE II). The most frequentlysampled genotype was 2-1, which was represented byeight strains associated with maize in South Africaand the United States and teosinte in Mexico andGuatemala. Seven strains displayed genotype 2-3, andwere associated with maize in the United States and
South Africa, as well as teosinte in Mexico. Five iso-lates displayed genotype 1-1 and were collected frommaize in South Africa and Mexico, as well as Mexicanteosinte. Genotype 1-4 was also represented by fivestrains, all of which were isolated from maize inSouth Africa and Mexico. Genotype 2-2 was repre-sented by two strains associated with maize in theUnited States and both genotypes 1-2 and 1-3 wererepresented by single strains that were isolated fromteosinte in Mexico.
Phylogenetic analyses. The number of parsimony in-formative characters in the six data sets ranged fromone for HB26 to six for histone H3 (TABLE III). Inall the data sets, only the b-tubulin sequence har-bored a parsimony uninformative character (TABLE
III, FIG. 1). This variable character was present onlyin isolate MRC7803. Phylogenetic analyses generatedunique gene genealogies for each of the individualdata sets (FIG. 1). In each case, a single tree was gen-erated and the consistency (CI) and retention (RI)indices for each were 1.00 and 1.00, respectively, in-dicating no homoplastic characters in any of the sixindividual data sets. As a result, all of the single-genegenealogies were of minimal length, i.e., equal to thenumber of parsimony informative sites. A single mostparsimonious tree was also obtained from the com-bined data sets (FIG. 2). The length of this tree was
1037STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
FIG. 1. Single-gene genealogies generated from sequence data sets for six different loci studied in 29 F. subglutinansstrains associated with maize and teosinte. In each case only one most parsimonious tree was obtained. The branch associatedwith the single parsimony uninformative character in the b-tubulin data set is indicated with an asterisk (*). The consistency(CI) and retention (RI) indices for each were 1.00 and 1.00, respectively. The different geographic origins are indicated incolor (red 5 South Africa; black 5 United States; blue 5 Mexico and Guatemala). A: b-tubulin gene genealogy consistingof 2 parsimony informative characters and 2 steps. B: Histone H3 genealogy consisting of 6 parsimony informative charactersand 6 steps. C: The single-gene genealogy for each of the HB9, HB14, and HB26 nuclear regions. Because the clusteringfor each of these regions was identical, they are represented by a single tree with lengths 2, 4, and 1, respectively. D:Calmodulin gene genealogy consisting of 2 steps.
equal to the number of parsimony informative char-acters (TABLE III), since homoplastic characters werealso absent in the combined data set (CI 5 1.00, RI5 1.00). The length of this tree was equal to thesummed lengths of the individual gene trees (TABLE
III), which is a distinctive feature of absolute congru-
ence among individual gene genealogies (Dykhuizenand Green 1991, Maynard Smith and Smith 1998,Taylor et al 1999b).
Clustering within the different genealogies wasvery similar. For the data sets HB9, HB14, and HB26the isolates were separated into two groups that al-
1038 MYCOLOGIA
FIG. 2. A: Genealogy generated from the combined data sets (color coding same as in FIG. 1). The branch associatedwith the single parsimony uninformative character in the b-tubulin data set is indicated with an asterisk (*). One single mostparsimonious tree with a length of 17 steps was obtained (CI 5 1.00; RI 5 1.00). B: The individuals included in each of theseven clusters correspond with the individuals displaying each of seven genotypes. Bootstrap values based on 1000 replicationsare indicated in parentheses.
ways included the same isolates (FIG. 1). The b-tu-bulin and calmodulin genealogies consisted of fourand three clusters of isolates, respectively. Althoughthere was some overlap, the clustering patterns forthe b-tubulin and calmodulin genes were unique.The genealogy generated from the histone H3 datagenerated four groups of isolates, some of whichshowed some resemblance to those generated for theother data sets.
Phylogenetic analysis of the combined data setfrom all the isolates included in this study revealedthe presence of two distinct groups among the iso-lates associated with maize and teosinte (FIGS. 2 and3). They were designated as Group 1 and 2 (FIG. 3).The genotypes 1-1, 1-2, 1-3 and 1-4 were present inGroup 1 and the genotypes 2-1, 2-2 and 2-3 were pres-
ent in Group 2 (FIGS. 2 and 3). Although this clus-tering pattern was not immediately detectable fromthe individual gene trees, the combined gene gene-alogy was not discordant with any of them.
DISCUSSION
In this study, we set out to use phylogenetic tools toanswer what appeared to be either a population orspecies level question. Although this approach is notwidely used in fungal taxonomy, it is not without pre-cedence and a number of researchers have reportedon the value of using phylogenetic concordance anal-ysis (Carbone et al 1999, Geiser et al 1998, Koufo-panou et al 1997, O’Donnell et al 1998b, 2000a, b,Taylor et al 1999a, b, 2000). Based on previous work
1039STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
FIG. 3. A single most parsimonious phylogram generated from the combined sequence data sets. Parsimony informativeand parsimony uninformative characters (indicated with an asterisk) were included in the analysis. In parentheses are indi-cated the different hosts followed by their specific geographic origins within South Africa, the United States, or Mexico andGuatemala. The tree is rooted to F. circinatum. Branch lengths are indicated above the branches and bootstrap values basedon 1000 replications are indicated in bold. (N/a 5 not available)
(Steenkamp et al 1999, 2001), our null hypothesiswas that the set of isolates associated with maize andteosinte, from a wide geographic range, would formpart of the existing E-mating population of the G.fujikuroi species complex (i.e., F. subglutinans sensustricto), and that the observed sequence variationmerely reflect intraspecific diversity. However, consid-ering the data generated in this study, we had to re-ject this hypothesis. The results clearly showed thatF. subglutinans is separated into reproductively isolat-ed populations that probably constitute separate sib-ling species.
The basic rationale behind the use of phylogeneticconcordance analyses involves the detection of con-gruence or the lack thereof, among different genetrees (Avise and Ball 1990, Dykhuizen and Green1991, Taylor et al 1999a, 1999b, 2000). Incongruenceamong gene trees from different loci indicates inter-breeding among individuals, since sexual recombi-nation ‘reshuffles’ their genomes. In an interbreed-
ing population this ‘reshuffling’ results in uniqueevolutionary histories for the genes in a specific in-dividual, while the genes from different individualsshare numerous characteristics or polymorphisms. Inreproductively isolated populations genes do not nor-mally flow among individuals, resulting in a lack ofshared polymorphisms and consequently perfectlycongruent gene trees. Our analyses showed that theevolutionary histories of the six nuclear regions wereperfectly congruent in each of the F. subglutinans in-dividuals studied. This is because every gene tree, aswell as the one generated from the combined data,was of minimum length (FIGS. 1 and 2). The F. sub-glutinans isolates from the United States, South Af-rica, Mexico, and Guatemala thus appear to repre-sent reproductively isolated populations.
The tree generated from the combined data sep-arated the 29 F. subglutinans isolates into sevengroups (FIG. 2). These groups corresponded to theseven multilocus genotypes identified from the se-
1040 MYCOLOGIA
quence comparisons (TABLE II). The evolutionaryhistories for each of the six nuclear regions in all theisolates from genotype 1-1, for example, are identical,indicating a lack of genome ‘reshuffling’ throughsexual reproduction. All the isolates in genotype 1-1can, therefore, be considered as clones of one anoth-er, since there is no evidence of interbreeding. Thesame is also true for genotypes 1-4, 2-1, 2-2 and 2-3.Although genotypes 1-2 and 1-3 are represented bysingle isolates, analysis of additional isolates will mostlikely reveal a similar trend. In theory, the overall lackof shared polymorphisms in the six nuclear regionsamong the seven groups or genotypes, separate the29 isolates into seven clones, a notion that is probablynot entirely correct (see below).
In light of similar studies on other fungi (reviewedby Taylor et al 1999a, 1999b, 2000) our results wererather unexpected. In most of these studies, phylo-genetic concordance analysis revealed the presenceof so-called cryptic species where individuals fromone species were shown to be reproductively isolatedfrom those in the other species. However, these testsalso detected that individuals within some of thesecryptic species were interbreeding in nature, becauseof the presence of numerous shared polymorphisms.As a result, generation of consensus trees from com-bined sequence data resulted in trees that were lon-ger than the expected minimum (observed treelength . number of polymorphic characters). Incontrast, the observed tree length in our study wasequal to the number of polymorphic characters (TA-BLE III), providing no evidence for interbreedingamong any of the seven F. subglutinans genotypes. Byemploying phylogenetic concordance analysis, ourstudy is one of few, and perhaps the first to suggestthat a group of fungi capable of interbreeding in thelaboratory appears to be propagating exclusivelyasexually in nature.
The idea that isolates of F. subglutinans exist inclonal populations is not congruent with previous re-ports. Based on analyses of phenotypic characters,isolates of a specific genotype do not always displaysimilar degrees of resistance to antimicrobial agentsand demonstrate similar abilities to produce myco-toxins (Desjardins et al 2000, Marasas et al 1983,1984, Sewram et al 1999a, 1999b, Shephard et al1999, Yan et al 1993). For example, the two isolatesin genotype 2-2 are significantly different in their re-sistance to hygromycin B (Yan et al 1993) and thethree South African isolates of genotype 1-1 producemarkedly different levels of moniliformin and fusa-proliferin (Sewram et al 1999a, 1999b, Shephard etal 1999). Another trait suggesting that isolates withsimilar genotypes are not ‘true clones’ is the distri-bution of mating type. If all the isolates with a specific
genotype are clones of one another, their matingtypes determined using either laboratory crosses orPCR-based methods (Steenkamp et al 2000b) shouldbe the same. Among the F. subglutinans isolates stud-ied, this was not the case, because each of the fivemulti-isolate genotypes was represented by bothMAT-1 and MAT-2 isolates (Desjardins et al 2000,Steenkamp et al 2000a, Yan et al 1993). This variationwithin groups of isolates with the same genotype, wasundoubtedly generated by earlier sexual recombina-tion events. Even though we did not include suffi-cient isolates and polymorphic loci to address ques-tions on the reproductive mode of F. sublutinans, ourresults suggest that sexual reproduction is not a com-mon phenomenon. Among the 17 polymorphic sitesanalyzed, all were fixed within isolates associated witha specific genotype. It is possible that the inclusionof additional data, especially those associated withmycotoxin production, antibiotic resistance and mat-ing, will result in the identification of non-fixed orshared polymorphism. However, the ongoing ab-sence of sex will also bring about fixation in thesecharacters and eventually the absence of shared poly-morphisms. Our results, therefore, suggest F. subglu-tinans populations are diverging into various repro-ductively isolated lineages that will most probablyeach eventually constitute separate species.
Classifying the isolates used in this study into theirsmallest diagnosable units or redefining species limitsin the existing F. subglutinans sensu stricto is problem-atic. Given that sexual reproduction between theseisolates appears to be absent in nature, the biologicalspecies concept cannot be applied. The phylogeneticspecies concept assists to some extent, but as shownabove, it identifies ‘clones’. Determining whetherthese ‘clones’ represent separate species is an arbi-trary exercise, since it is impractical to designate eachclonal population as a distinct species. Likewise, toclassify each as the same species would be incorrectand would not reflect the natural situation. Separat-ing F. subglutinans into smaller units/taxa is furthercomplicated by the fact that some isolates differ byno more than a single nucleotide at the six nuclearregions analyzed and display no known or diagnos-able phenotypic differences. Among the isolates stud-ied, we have, however, observed two major phyloge-netic groups (groups 1 and 2) (FIG. 3), which mayrepresent a species partition where groups 1 and 2represent cryptic species. Genotypes 1-1, 1-2, 1-3 and1-4 would belong to group 1 and 2-1, 2-2 and 2-3would belong to group 2. Whether this partition willprove to be diagnosable using phenotypic charactersremains to be tested.
There are no apparent links between the genotypeto which the F. subglutinans isolates belong and geo-
1041STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
graphic origins or host (FIG. 3). South African iso-lates were present in all genotypes except 1-2, 1-3 and2-2, while Mexican/Guatemala isolates were presentin all but genotype 2-2. The United States isolatesappeared to be restricted to genotypes 2-1, 2-2 and2-3. Isolates from maize were present in all genotypesexcept 1-2 and 1-3, while those from teosinte wereonly absent from genotype 2-2. From the limited dataavailable, the separation of F. subglutinans into theseven genotypes/groups also does not appear to belinked to mycotoxin production (Desjardins et al2000, Marasas et al 1983, 1984, Sewram et al 1999a,1999b, Shephard et al 1999) or morphology (Desjar-dins et al 2000, Steenkamp et al 1999, 2001). Therewere also no apparent links between the Groups 1and 2 separation and geographic origin, host or my-cotoxin production (FIG. 3) (Desjardins et al 2000,Marasas et al 1983, 1984, Sewram et al 1999a, 1999b,Shephard et al 1999, Steenkamp et al 1999, 2001).Noticeably, maize in the United States is associatedonly with group 2, but this is probably a samplingeffect. The 17 polymorphic sites, therefore, appearto be the only diagnosable features separating theseven genotypes and Groups 1 and 2. As with otherfungi (Geiser et al 2000, Peng et al 1999, Taylor et al2000), our results may lead to the future identifica-tion of morphological, physiological, pathogenic, ortoxigenic characters supporting either the presenceof Groups 1 and 2 or their sub-lineages.
Based on the results presented here, we have at-tempted to interpret the findings of previous matingstudies (Desjardins et al 2000, Steenkamp et al 1999,2001). Again, there was no relationship between thegenotype represented and ability to sexually interactin the laboratory. The genotype 1-1 isolate MRC1084from South African maize, for example, was capableof producing fertile offspring when mated with ge-notype 1-3 and 2-1 isolates. Many other isolates werealso capable of successful sexual interaction with iso-lates displaying genotypes other than their own. Thepresence of the two proposed cryptic species helpedto clarify some of the problems that were encoun-tered in previous mating studies (Desjardins et al2000, Steenkamp et al 1999, 2001). For example,some of the F. subglutinans strains that were collectedfrom maize and teosinte in Mexico and CentralAmerica were sexually incompatible with those fromthe United States (Desjardins et al 2000). As revealedhere all the available F. subglutinans isolates from theUnited States belong to Group 2, whereas those fromMexico and Central America belong to both Groups1 and 2, which may partially explain their incompat-ibility. Our results can, however, not explain why theUnited States isolates were sexually incompatibleeven with the Mexican strains from Group 2, or why
most of the isolates collected from Mexico and Cen-tral America are capable of fertile sexual interaction,even though they represent separate cryptic species.Inclusion of all 29 isolates in a mating study mighteventually shed light on the connection between theability to reproduce sexually and genotype/crypticspecies, but conditions in the laboratory do not nec-essarily mimic those in nature. Many fungi are ableto sexually interact across the species barrier (e.g.,Vilgalys and Sun 1994), which was also demonstratedfor strains of F. subglutinans and F. circinatum (Des-jardins et al 2000, Steenkamp et al 2001).
The inconsistency between the biological and phy-logenetic species concepts illustrated in this study in-troduces serious complications for the classificationof fungi, especially those in the G. fujikuroi speciescomplex, since classification relies heavily on bothconcepts. Problems with using the morphologicalspecies concept have been reported by many workers(O’Donnell et al 1998a, 2000a, Steenkamp et al 1999,2001), but the current study is the first to report dis-parities using the biological species concept for clas-sifying fungi in the G. fujikuroi species complex. Un-doubtedly, species recognition using phylogeneticconcordance analysis will yield the most informativeresults, but there are two major problems associatedwith the use of this approach. Firstly, even though wehave considerable knowledge on the subject of phy-logenetics, new ideas are constantly emerging and weare still in the learning-phase with respect to its meth-odologies and the interpretation of results. Secondly,we are still in the process of developing a definitionfor a fungal species and how to recognize this unit.Although we have attempted to address the taxonom-ic status and reproductive mode of isolates residingin F. subglutinans sensu stricto, many questions remainunanswered. These include an understanding of howan asexual species should be defined, how many nu-cleotide differences constitute a species divergence inFusarium and others. Our study and similar cases willthus represent an interesting topic for discussion asthe fields of fungal reproduction and phylogenetictaxonomy advance.
LITERATURE CITED
Avise JC, Ball RM. 1990. Principles of genealogical concor-dance in species concepts and biological taxonomy. In:Futuyma D, Antonovics J, eds. Oxford Surveys in Evo-lutionary Biology. Vol. 7. Oxford: Oxford UniversityPress. p 45–67.
Britz H, Coutinho TA, Wingfield MJ, Marasas WFO, GordonTR, Leslie JF. 1999. Fusarium subglutinans f. sp. pinirepresents a distinct mating population in the Gibber-
1042 MYCOLOGIA
ella fujikuroi species complex. Appl Environ Microbiol65:1198–1201.
Carbone I, Anderson JB, Kohn LM. 1999. Patterns of de-scent in clonal lineages and their multilocus finger-prints are resolved with combined gene genealogies.Evolution 53:11–21.
———, Kohn LM. 1999. A method for designing primersets for speciation studies in filamentous ascomycetes.Mycologia 91:553–556.
Correll JC, Gordon TR, McCain AH, Fox JW, Koehler CS,Wood DL, Schultz ME. 1991. Pitch canker disease inCalifornia: pathogenicity, distribution and canker de-velopment on Monterey pine (Pinus radiata). Plant Dis75:676–682.
Cracraft J. 1983. Species concepts and speciation analysis.In: Johnston RF, ed. Current Ornithology. Vol. 1. NewYork: Plenum Press. p 159–187.
Desjardins AE, Plattner RD, Gordon TR. 2000. Gibberellafujikuroi mating population A and Fusarium subglutin-ans from teosinte species and maize from Mexico andCentral America. Mycol Res 104:865–872.
Dobzhansky T. 1937. Genetics and the origin of species.New York: Columbia University Press. p 364.
Dykhuizen DE, Green L. 1991. Recombination in Escherich-ia coli and the definition of biological species. J Bac-teriol 173:7257–7268.
Geiser DM, Pitt JI, Taylor JW. 1998. Cryptic speciation andrecombination in the aflatoxin-producing fungus As-pergillus flavus. Proc Natl Acad Sci USA 95:388–393.
———, Dorner JW, Horn BW, Taylor JW. 2000. The phy-logenetics of mycotoxin and sclerotium production inAspergillus flavus and Aspergillus oryzae. Fungal GenetBiol 31:169–179.
Gerlach W, Nirenberg HI. 1982. The genus Fusarium—apictorial atlas. Mitt Biol Bundesanst Land-Forstw Ber-lin-Dahlem 209:1–406.
Glass NL, Donaldson GC. 1995. Development of primer setsfor use in PCR to amplify conserved genes from fila-mentous ascomycetes. Appl Environ Microbiol 61:1323–1330.
Hsieh WH, Smith SN, Snyder WC. 1977. Mating groups inFusarium moniliforme. Phytopathology 67:1041–1043.
Klittich CJR, Leslie JF. 1992. Identification of a second mat-ing population within the Fusarium moniliforme ana-morph of Gibberella fujikuroi. Mycologia 84:541–547.
Koufopanou V, Burt A, Taylor JW. 1997. Concordance ofgene genealogies reveals reproductive isolation in thepathogenic fungus Coccidioides immitis. Proc Natl AcadSci USA 94:5478–5482.
Kuhlman EG. 1982. Varieties of Gibberella fujikuroi with an-amorphs in Fusarium section Liseola. Mycologia 74:756–768.
Leslie JF. 1991. Mating populations in Gibberella fujikuroi(Fusarium section Liseola). Phytopathology 81:1058–1060.
———. 1995. Gibberella fujikuroi: available populations andvariable traits. Can J Bot 73:S282–S291.
———. 1996. Introductory biology of Fusarium moniliforme.In: Jackson LS, De Vries JW, Bullerman LB, eds. Ad-
vances in experimental medicine and biology. Vol. 392.New York: Plenum Press. p 153–164.
———, Pearson CAS, Nelson PE, Toussoun TA. 1990. Fu-sarium spp. from corn, sorghum and soybean fields inthe central and eastern United States. Phytopathology80:343–350.
———, Plattner RD, Desjardins AE, Klittich CJR. 1992. Fu-monisin B1 production by strains from different mat-ing populations of Gibberella fujikuroi (Fusarium sec-tion Liseola). Phytopathology 82:341–345.
Logrieco A, Moretti A, Ritieni A, Chelkowski J, Altomare C,Bottalico A, Randazzo G. 1993. Natural occurrence ofbeauvericin in preharvest Fusarium subglutinans in-fected corn ears in Poland. J Agric Food Chem 41:2149–2152.
Marasas WFO, Nelson PE, Toussoun TA. 1984. ToxigenicFusarium species: identity and mycotoxicology. Univer-sity Park: Pennsylvania State University Press. p 328.
———, Thiel PG, Rabie CJ, Nelson PE, Toussoun TA. 1983.Moniliformin production in Fusarium section Liseola.Mycologia 78:242–247.
Maynard Smith J, Smith NH. 1998. Detecting recombina-tion from gene trees. Mol Biol Evol 15:590–599.
Mayr E. 1940. Speciation phenomena in birds. Am Natu-ralist 74:249–278.
Nirenberg HI, O’Donnell K. 1998. New Fusarium speciesand combinations within the Gibberella fujikuroi speciescomplex. Mycologia 90:434–458.
O’Donnell K, Cigelnik E. 1997. Two divergent intragenomicrDNA ITS2 types within a monophyletic lineage of thefungus Fusarium are nonorthologous. Mol PhylogenetEvol 7:103–116.
———, ———, Nirenberg HI. 1998a. Molecular systematicsand phylogeography of the Gibberella fujikuroi speciescomplex. Mycologia 90:465–493.
———, ———, Aoki T, Cigelnik E. 2000a. A multigene phy-logeny of the Gibberella fujikuroi species complex: de-tection of additional phylogenetically distinct species.Mycoscience 41:61–78.
———, Kistler HC, Cigelnik E, Ploetz RC. 1998b. Multipleevolutionary origins of the fungus causing Panama dis-ease of banana: concordant evidence from nuclear, mi-tochondrial gene sequences. Proc Natl Acad Sci USA95:2044–2049.
———, ———, Tacke BK, Casper HH. 2000b. Gene gene-alogies reveal global phylogenetic structure and repro-ductive isolation among lineages of Fusarium grami-nearum, the fungus causing wheat scab. Proc Natl AcadSci USA 97:7905–7910.
Peng T, Orsborn KI, Orbach MJ, Galgiani JN. 1999. Proline-rich vaccine candidate antigen of Coccidioides immitis:conservation among isolates and differential expres-sion with spherule maturation. J Infect Dis 179:518–521.
Sewram V, Niewoudt TW, Marasas WFO, Shephard GS, Ri-tieni A. 1999a. Determination of the Fusarium myco-toxins, fusaproliferin and beauvericin by high-perfor-mance liquid chromatography-electrospray ionizationmass spectromertry. J Chromatogr A 858:175–185.
———, ———, ———, ———, ———. 1999b. Determi-
1043STEENKAMP ET AL: CRYPTIC SPECIATION IN FUSARIUM
nation of the mycotoxin moniliformin in cultures ofFusarium subglutinans and in naturally contaminatedmaize by high-performance liquid chromatography-at-mospheric pressure chemical ionization mass spec-trometry. J Chromatogr A 848:185–191.
Shephard GS, Sewram V, Niewoudt TW, Marasas WFO, Ri-tieni A. 1999. Production of the mycotoxins fusaproli-ferin and beauvericin by South African isolates in theFusarium Section Liseola. J Agric Food Chem 47:5111–5115.
Snyder WC, Hansen HN. 1945. The species concept in Fu-sarium with reference to Discolor and other sections.Am J Bot 32:657–666.
Steenkamp ET, Coutinho TA, Desjardins AE, Wingfield BD,Marasas WFO, Wingfield MJ. 2001. Gibberella fujikuroimating population E is associated with maize and teo-sinte. Mol Plant Pathol 2:215–221.
———, Britz H, Coutinho TA, Wingfield BD, Marasas WFO,Wingfield MJ. 2000a. Molecular characterization of Fu-sarium subglutinans associated with mango malforma-tion. Mol Plant Pathol 1:187–193.
———, Wingfield BD, Coutiho TA, Zeller KA, WingfieldMJ, Marasas WFO, Leslie JF. 2000b. PCR-based identi-fication of MAT-1 and MAT-2 in the Gibberella fujikuroispecies complex. Appl Environ Microbiol 66:4378–4382.
———, Wingfield BD, Coutinho TA, Wingfield MJ, MarasasWFO. 1999. Differentiation of Fusarium subglutinans f.sp. pini by histone gene sequence data. Appl EnvironMicrobiol 65:3401–3406.
Sun S-K, Snyder WC. 1981. The bakanae disease of the riceplant. In: Nelson PE, Toussoun TA, Cook RJ, eds. Fu-sarium: diseases, biology and taxonomy. UniversityPark: Pennsylvania State University Press. p 104–113.
Swofford DL. 1998. PAUP*. Phylogenetic Analysis UsingParsimony (*and Other Methods). Version 4.0*. Sun-derland, Massachusetts: Sinauer Associates.
Taylor JW, Geiser DM, Burt A, Koufopanou V. 1999a. Theevolutionary biology and population genetics underly-ing fungal strain typing. Clin Microbiol Rev 12:126–146.
———, Jacobson DJ, Kroken S, Kasuga T, Geiser DM, Hib-bett DS, Fisher MC. 2000. Phylogenetic species recog-nition and species concepts in fungi. Fungal Genet Biol31:21–32.
———, ———, Fisher MC. 1999b. The evolution of asexualfungi: reproduction, speciation and classification.Annu Rev Phytopathol 37:197–246.
Varma A, Lele VC, Raychaudhuri SP, Ram A, Sang A. 1974.Mango malformation: a fungal disease. Phytopath Z 79:254–257.
Ventura JA, Zambolim L, Gilbertson RL. 1993. Proposicaode nowa forma speciales em Fusarium subglutinans noabacaxizeiro. Fitopatol Bras Supl 18:280.
Vesonder RF, Logrieco A, Bottalico A, Altomore C, PetersonSW. 1995. Fusarium species associated with bananafruit rot and their potential toxigenicity. Mycotoxin Res11:93–98.
Vilgalys R, Sun BL. 1994. Ancient and recent patterns ofgeographic speciation in the oyster mushroom Pleuro-tus ostreatus revealed by phylogenetic analysis of ribo-somal DNA sequences. Proc Natl Acad Sci USA 91:4599–4603.
Yan K, Dickman MD, Xu JR, Leslie JF. 1993. Sensitivity offield strains of Gibberella fujikuroi (Fusarium sectionLiseola) to benomyl and hygromycin B. Mycologia 85:206–213.