Comparison of Colletotrichum orbiculare and Several Allied Colletotrichum spp. for mtDNA RFLPs, Intron RFLP and Sequence Variation, Vegetative Compatibility, and Host Specificity

Abstract

ABSTRACT Based on spore morphology, appressorium development, sequence similarities of the rDNA, and similarities in amplified restriction fragment length polymorphism (AFLP), it has been proposed that Colletotrichum orbiculare, C. trifolii, C. lindemuthianum, and C. malvarum represent a single phylogenetic species, C. orbiculare. In the current study, the phylogenetic relationship among isolates in the C. orbiculare species complex was reassessed. In all, 72 isolates of C. orbiculare from cultivated cucurbit or weed hosts, C. trifolii from alfalfa, C. lindemuthianum from green bean, and C. malvarum from prickly sida (Sida spinosa) were examined for mitochondrial DNA (mtDNA) restriction fragment length polymorphisms (RFLPs), RFLPs and sequence variation of a 900-bp intron of the glutamine synthetase gene and a 200-bp intron of the glyceraldehyde-3-phosphate dehydrogenase gene, and vegetative compatibility. In addition, host specificity was examined in foliar inoculations on cucurbit, bean, and alfalfa hosts. Inoculations also were conducted on cucumber fruit. Genetically distinct isolates, based on vegetative compatibility, within the species complex (C. orbiculare, C. trifolii, and C. malvarum) had an identical mtDNA haplotype (haplotype A) when examined with each of three different restriction enzymes. Isolates of C. lindemuthianum had a very similar mtDNA haplotype to haplotype A, with a single polymorphism detected with the enzyme HaeIII. The four species represent a phylogenetically closely related group based on a statistical analysis of the 900- and 200-bp intron sequences. However, distinct RFLPs in the 900-bp intron were consistently associated with each species and could be used to qualitatively and quantitatively distinguish each species. Furthermore, each of the species showed distinct host specificity, with isolates of C. orbiculare (from cucurbits), C. lindemuthianum, and C. trifolii being pathogenic only on cucurbits, green bean, and alfalfa, respectively. Consequently, distinct and fixed nucleotide, or genotypic (intron sequences and RFLPs) and phenotypic (host specificity) characteristics can be used to distinguish C. orbiculare, C. lindemuthianum, and C. trifolii from one another; therefore, they should be recognized as distinct species. This species delineation is consistent with the most current species concepts in fungi. More isolates and further characterization is needed to determine whether C. orbiculare from cocklebur and C. malvarum represent distinct species. RFLPs of the 900-bp intron may represent a relatively inexpensive, reliable, and useful diagnostic tool for general species differentiation in the genus Colletotrichum.

Full text

Vol. 97, No. 10, 2007 1305
Population Biology
Comparison of Colletotrichum orbiculare and Several
Allied Colletotrichum spp. for mtDNA RFLPs, Intron RFLP
and Sequence Variation, Vegetative Compatibility, and Host Specificity
B. Liu, L. A. Wasilwa, T. E. Morelock, N. R. O’Neill, and J. C. Correll
First, second, and fifth authors: Department of Plant Pathology, and third author, Horticulture Department, University of Arkansas,
Fayetteville 72701; and fourth author: U.S. Department of Agriculture–Agricultural Research Service, Beltsville, MD 20705.
Accepted for publication 16 May 2007.
ABSTRACT
Liu, B., Wasilwa, L. A., Morelock, T. E., O’Neill, N. R., and Correll, J. C.
2007. Comparison of Colletotrichum orbiculare and several allied
Colletotrichum spp. for mtDNA RFLPs, intron RFLP and sequence
variation, vegetative compatibility, and host specificity. Phytopathology
97:1305-1314.
Based on spore morphology, appressorium development, sequence
similarities of the rDNA, and similarities in amplified restriction fragment
length polymorphism (AFLP), it has been proposed that Colletotrichum
orbiculare, C. trifolii, C. lindemuthianum, and C. malvarum represent a
single phylogenetic species, C. orbiculare. In the current study, the phylo-
genetic relationship among isolates in the C. orbiculare species complex
was reassessed. In all, 72 isolates of C. orbiculare from cultivated
cucurbit or weed hosts, C. trifolii from alfalfa, C. lindemuthianum from
green bean, and C. malvarum from prickly sida (Sida spinosa) were ex-
amined for mitochondrial DNA (mtDNA) restriction fragment length
polymorphisms (RFLPs), RFLPs and sequence variation of a 900-bp
intron of the glutamine synthetase gene and a 200-bp intron of the glycer-
aldehyde-3-phosphate dehydrogenase gene, and vegetative compatibility.
In addition, host specificity was examined in foliar inoculations on
cucurbit, bean, and alfalfa hosts. Inoculations also were conducted on
cucumber fruit. Genetically distinct isolates, based on vegetative com-
patibility, within the species complex (C. orbiculare, C. trifolii, and
C. malvarum) had an identical mtDNA haplotype (haplotype A) when
examined with each of three different restriction enzymes. Isolates of
C. lindemuthianum had a very similar mtDNA haplotype to haplotype A,
with a single polymorphism detected with the enzyme HaeIII. The four
species represent a phylogenetically closely related group based on a
statistical analysis of the 900- and 200-bp intron sequences. However,
distinct RFLPs in the 900-bp intron were consistently associated with
each species and could be used to qualitatively and quantitatively
distinguish each species. Furthermore, each of the species showed distinct
host specificity, with isolates of C. orbiculare (from cucurbits), C. linde-
muthianum, and C. trifolii being pathogenic only on cucurbits, green
bean, and alfalfa, respectively. Consequently, distinct and fixed nucleo-
tide, or genotypic (intron sequences and RFLPs) and phenotypic (host
specificity) characteristics can be used to distinguish C. orbiculare,
C. lindemuthianum, and C. trifolii from one another; therefore, they should
be recognized as distinct species. This species delineation is consistent
with the most current species concepts in fungi. More isolates and further
characterization is needed to determine whether C. orbiculare from cockle-
bur and C. malvarum represent distinct species. RFLPs of the 900-bp
intron may represent a relatively inexpensive, reliable, and useful diag-
nostic tool for general species differentiation in the genus Colletotrichum.
Additional keywords: C. lagenarium.
Colletotrichum spp. cause anthracnose diseases of many eco-
nomically important crops worldwide (40). The classification of
species within this genus traditionally has been based on conidial
shape and size, presence of sclerotia, appressoria production, and,
often, on host origin. Therefore, the species concept for the genus
Colletotrichum is based on both morphological characteristics and
host specificity (40). Von Arx (47,48) recognized 13 species in the
genus Colletotrichum based on conidial shape and size, host range,
and type of appressorium produced. Within the species Colletot-
richum gloeosporioides, nine specialized forms were recognized
(47). Sutton (40) later elevated each of these nine forms to a
distinct species, which included C. orbiculare (Berk. & Mont.)
Arx (= C. lagenarium (Pass.) Ellis & Halst.) from cucurbits,
C. trifolii Bain from alfalfa (Medicago sativa), C. lindemuthia-
num (Sacc. & Magnus) Briosi & Cavara from green bean (Pha-
seolus vulgaris L.), C. malvarum (A. Braun & Casp.) Southw.
from prickly sida (Sida spinosa L.), C. gnaphalii Syd. (no host
specified), C. helichrysi (G. Winter) Arx (no host specified),
C. miyabeana (no host specified), C. musae (Berk. & M. A.
Curtis) Arx from banana (Musa spp. L.), and C. psoraleae (Peck)
Arx (no host specified). In 1992, 37 species of Colletotrichum
were recognized, many of which were distinguished based on
their host origin (41).
Although initially recognized as “special forms” of C. gloeo-
sporioides, C. orbiculare, C. trifolii, C. lindemuthianum, and
C. malvarum are distinct from C. gloeosporioides in that they
generally have ovoid conidia (38), no septum in germinated conidia
(2,3,27,38), and produce appressoria of similar shapes and
dimension (38). Thus, although these species can be distinguished
from C. gloeosporioides based on morphology, their distinction
from one another has been based largely on host origin.
Molecular technologies based on the analysis of DNA have
been used to examine the relationship of Colletotrichum spp. (2,9,
15,20,26,32,38). Based on molecular and morphological data, a
close relationship between C. orbiculare from cucumber, C. tri-
folii from alfalfa, C. malvarum from prickly sida, and C. lindemu-
thianum from bean has been observed (2,27,34,38). Based on
spore morphology, appressorium development, and sequence
similarities of the rDNA, it was proposed that C. orbiculare,
C. trifolii, C. lindemuthianum, and C. malvarum should be con-
sidered a single species (38). It also was suggested that, within the
broader species concept of C. orbiculare, isolates which showed
host specialization should be classified as distinct formae speciales
Corresponding author: J. C. Correll; E-mail address: jcorrell@uark.edu
doi:10.1094/ PHYTO-97-10-1305
© 2007 The American Phytopathological Society

1306 PHYTO PAT HO LOGY
of C. orbiculare. Current species concepts in fungi, however,
suggest that host specificity is useful at the species rank rather
than the informal rank of formae speciales (18). Thus, host speci-
ficity, along with other fixed genotypic characters, could provide
justification for species delineation within the genus Colletot-
richum more specifically within the C. orbiculare species complex.
Analysis of DNA sequences continues to be a valuable tool to
help resolve relationships among and within species and species
complexes of Colletotrichum (5,12,13,15–17,20,24–26). DNA
sequence analysis, along with morphology and host range, were
used to delineate species within the broader C. graminicola com-
plex on sorghum, maize, wheat, oat, forage, turf, and amenity
grasses (12). The phylogenetic analysis demonstrated the rela-
tively low level of phylogenetic resolution of the internal tran-
scribed spacer (ITS) region compared with the MAT1-2 mating
type (HMG) and manganese-type superoxide dismutase gene
(SOD2) sequences (12,13). Based on the sequences of HMG and
SOD2 deposited in GenBank, it was found that the most phylo-
genetically informative characters were present in the intron por-
tion of the genes examined (13). In addition, earlier work to
define the phylogenetic origin of the panama disease pathogen
Fusarium oxysporum f. sp. cubense also showed that intron se-
quences within the EF-1α gene were highly informative (28,29).
Single-copy nuclear genes are often favorable for use in phylo-
genetic analyses because they are biparentally inherited and fol-
low patterns of concerted evolution, and the sequence alignment
based on single-copy genes limits the alignment ambiguity (1). In
contrast, ITS data typically yield lower consistency index (CI)
and retention index (RI) values compared with other loci, indi-
cating a higher level of homoplasy (1). Moreover, examination of
the ITS gene in Colletotrichum spp. typically has resulted in low
bootstrap value support, indicating that the ITS gene is less phylo-
genetically informative (13). A recent study of inter- and intra-
specific variation within the species complex C. acutatum con-
firmed that sequence variation in two introns from the single-copy
genes (glutamine synthase and glyceraldehyde-3-phosphate dehy-
drogenase) were phylogenetically informative, whereby the concor-
dance of both intron phylogenetic tree topologies revealed several
distinct clades that were supported by very high bootstrap values
and consistency indices (17). Based on these data, it was hypothe-
sized that a similar approach could be used for a robust analysis
of the relationship between C. orbiculare and several allied species.
The objective of this study were to test the hypothesis that
distinct and fixed phenotypic (host specificity) and genotypic
(sequence) characteristics exist among the various species in the
C. orbiculare complex. To test this hypothesis, representative
isolates of C. orbiculare, C. lindemuthianum, C. trifolii, and
C. malvarum were compared for mitochondrial (mt)DNA restric-
tion fragment length polymorphisms (RFLPs), intron RFLPs and
sequence variation, vegetative compatibility, and host specificity
on cucurbits, green bean, and alfalfa.
MATERIALS AND METHODS
Isolates. The isolates used in this study were either recovered
from symptomatic tissue by the authors, received from other
TABLE 1. Isolates of Colletotrichum orbiculare, C. lindemuthianum, C. trifolii, and C. malvarum, their host specificity, and their 900-bp intron restriction frag-
ment length polymorphism (RFLP) haplotype
Host reactionx Intron haplotypez
Species Isolate VCGw Host A B C mtDNAy HHH HHM PstI MspI HhaI HaeIII P+M
C. orbiculare JC1 1001 Cucumber + – – A C1 C1 C C C D C
C. orbiculare NC3 1001 Cucumber + – – A C1 C1 C C C D C
C. orbiculare AK9 1001 Pumpkin + – – A C1 C1 C C C D C
C. orbiculare CP12 1002 Cucumber + – – A C1 C1 C C C D C
C. orbiculare LB4 1001 Cantaloupe + – – A C1 C1 C C C D C
C. orbiculare MMB 1001 Cucumber + – – A C1 C1 C C C D C
C. orbiculare JH3 1001 Cucumber + – … … C1 C1 C C C D C
C. orbiculare LD1 1001 Cucumber + – – … C1 C1 C C C D C
C. orbiculare AK6 1002 Cantaloupe + – … A C1 C1 C C C D C
C. orbiculare BB17 1002 Cantaloupe + – – A C1 C1 C C C D C
C. orbiculare CP6 1002 Watermelon + – – A C1 C1 C C C D C
C. orbiculare CP7 1002 Watermelon + – – A C1 C1 C C C D C
C. orbiculare JX13 1002 Watermelon + – – A C1 C1 C C C D C
C. orbiculare LB2 1002 Cucuzzi gourd + – – A C1 C1 C C C D C
C. orbiculare MME 1002 Watermelon + – – … C1 C1 C C C D C
C. orbiculare RL1 1002 Watermelon + – – … C1 C1 C C C D C
C. orbiculare RL2 1002 Watermelon + … … … C1 C1 C C C D C
C. orbiculare CR3 1002 Watermelon + – – A C1 C1 C C C D C
C. orbiculare JX11 1002 Watermelon + – – A C1 C1 C C C D C
C. orbiculare MH2 1003 Cucumber + – – A C1 C1 C C C D C
C. orbiculare MH5 1003 Cucumber + – – A C1 C1 C C C D C
C. orbiculare DAR61396 1003 Cucumber + – … A C1 C1 C C C D C
C. orbiculare JX10A 1004 Melon + – – A C1 C1 C C C D C
C. orbiculare JX10B 1004 Melon + – – A C1 C1 C C C D C
C. lindemuthianum 14-2-39 CL-1 Bean – … – A2 D D C D C D D
C. lindemuthianum 14-2-45 CL-1 Bean – + – A2 D D C D C D D
C. lindemuthianum JZ1 CL-1 Bean – … … … D D C D C D D
C. lindemuthianum JZ2 CL-1 Bean – + – … D D C D C D D
C. lindemuthianum JK1 Bean – + – A2 D D C D C D D
C. lindemuthianum JK2 Bean – + – A2 D D C D C D D
C. lindemuthianum JK3 Bean – + – A2 D D C D C D D
C. lindemuthianum JK4 Bean – + – A2 D D C D C D D
(continued on next page)
w VCG = vegetative compatibility group.
x A = cucurbits, B = green bean, C = alfalfa, and … = missing data. + = disease symptoms and – = no disease symptoms.
y Mitochondrial (mt)DNA RFLP haplotype of total DNA cut with PvuII and probed with mtDNA clones from C. orbiculare isolate JC1.
z Haplotypes were determined based on the RFLP pattern of the 900-bp intron cut with individual enzymes or in combination; HHH = HindIII + HinfI + HaeIII;
HHM = HindIII + HinfI + MspI; P+M = PstI + MspI. The same letter, within a column, denotes that the isolates have an identical RFLP haplotype.

Vol. 97, No. 10, 2007 1307
researchers, obtained from the American Type Tissue Culture
Collection (ATCC, Rockville, MD), or obtained from the Col-
letotrichum collection, University of Arkansas (Table 1). Single-
spore isolates were stored on desiccated filter paper at 4°C.
Isolates of C. orbiculare from cucumber (Cucumis sativus L.),
cantaloupe (C. melo L. (Cantaloupensis group)), watermelon (Cit-
rullus lanatus (Thunb.) Matsumura & Nakai), melon [Cucumis
melo L. (Reticulatus group), gooseberry gourd (C. myriocarpus),
and cocklebur (Xanthium spinosum L.) were used in this study.
The isolates of Colletotrichum orbiculare used were representa-
tive of the genetic diversity previously identified within a world-
wide collection of C. orbiculare (51,52). The isolates belonged to
four different vegetative compatibility groups (VCGs) (52) (Table
1). Two of the isolates of C. orbiculare (LW1 and LW6) were from
cocklebur (49,51). The isolates were collected from throughout
the United States or from Australia and Taiwan from 1990 to 1997.
In all, 23 isolates of C. lindemuthianum recovered from green
bean (P. vulgaris), 20 isolates of C. trifolii from alfalfa (M. sativa)
(4,23), and three isolates of C. malvarum from prickly sida
(S. spinosa) (22) were examined (Table 1). The isolates were ob-
tained from the Colletotrichum culture collection of the Univer-
sity of Arkansas between 1996 and 1997, or from other re-
searchers. The isolates of C. trifolii originated from Canada or the
United States. The isolates of C. lindemuthianum originated from
the United States, Canada, Mexico, Costa Rica, the Dominican
Republic, Honduras, Ecuador, Argentina, and Columbia.
DNA isolation. The isolates of C. orbiculare, C. lindemuthia-
num, C. trifolii, and C. malvarum were examined for mtDNA
RFLPs (Table 1). All isolates were grown in 0.5-liter flasks
containing 200 ml of complete medium broth (7). The flasks were
incubated at room temperature on an orbital shaker at 120 rpm for
4 to 5 days. Mycelium was harvested by vacuum filtration
through miracloth, rinsed with 300 ml of deionized water, and
frozen. The mycelium was lyophilized and ground into a powder
using liquid nitrogen. Total DNA was extracted using a modified
minipreparation protocol (7).
RFLP analysis. Three restriction enzymes were used indi-
vidually to digest 1.5 µg of DNA by incubating with 10 units of
endonucleases (EcoRI, HaeIII, and PvuII) for 16 h at 37°C fol-
lowing the manufacturer’s recommendations (New England Bio-
labs, Inc., MA). Electrophoresis was conducted in 0.8% agarose
gels (25 by 20 cm) for 2 h at 20 V and 18 h at 54 V. Each gel was
photographed on a UV transilluminator. DNA was blotted onto a
nylon membrane (Hybond N+; Amersham, Arlington Heights, IL)
by Southern blotting. The membrane was then heated 80°C for 2 h.
Two large non-overlapping mtDNA clones, 4U40 (13.7 kb) and
2U18 (10.1 kb), of mtDNA of C. orbiculare (representing 65% of
the mitochondrial genome) were combined in equimolar concen-
trations and used for hybridization (11). The enhanced chemi-
luminescent (ECL) nonradioactive labeling and detection kit was
used for hybridization following the manufacturer’s instructions
(Amersham). Prehybridization (1 h) and hybridization (12 to 16 h)
were done at 42°C in a shaker according to the manufacturer’s
instructions.
VCGs. All isolates of C. orbiculare were characterized for
vegetative compatibility using nitrate-nonutilizing (nit) mutants as
TAB LE 1 . (continued from preceding page)
Host reactionx Intron haplotypez
Species Isolate VCGw Host A B C mtDNAy HHH HHM PstI MspI HhaI HaeIII P+M
C. lindemuthianum JK5 Bean – + – A2 D D C D C D D
C. lindemuthianum JK6 Bean – + – A2 D D C D C D D
C. lindemuthianum JK7 Bean – + – A2 D D C D C D D
C. lindemuthianum JK8 Bean – + – A2 D D C D C D D
C. lindemuthianum JK9 Bean – + – A2 D D C D C D D
C. lindemuthianum JK10 Bean – + – A2 D D C D C D D
C. lindemuthianum JK11 Bean – + – A2 D D C D C D D
C. lindemuthianum JK12 Bean – + – A2 D D C D C D D
C. lindemuthianum JK14 Bean – + – A2 D D C D C D D
C. lindemuthianum JK15 Bean – + – A2 D D C D C D D
C. lindemuthianum JK16 Bean – + – A2 D D C D C D D
C. lindemuthianum JK17 Bean – + – A2 D D C D C D D
C. lindemuthianum JK18 Bean – + – A2 D D C D C D D
C. lindemuthianum JK19 Bean – + – A2 D D C D C D D
C. lindemuthianum JK20 Bean – + – A2 D D C D C D D
C. trifolii 14-2-63 CT-1 Alfalfa … … + A E E C C C D C
C. trifolii ON2 Alfalfa – – + A E E C C C D C
C. trifolii ON3 Alfalfa – – – A E E C C C D C
C. trifolii ON7 Alfalfa – – + A E E C C C D C
C. trifolii ON9 Alfalfa – – … A E E C C C D C
C. trifolii ON10 Alfalfa – – + A E E C C C D C
C. trifolii ON12 Alfalfa – – + A E E C C C D C
C. trifolii ON15 Alfalfa – – + A E E C C C D C
C. trifolii ON16 Alfalfa – – + A E E C C C D C
C. trifolii ON1 Alfalfa – – + A E E C C C D C
C. trifolii ON4 Alfalfa – – + A E E C C C D C
C. trifolii ON5 Alfalfa – – + A E E C C C D C
C. trifolii ON6 Red clover – – + A E E C C C D C
C. trifolii ON8 Alfalfa – – + A E E C C C D C
C. trifolii ON11 Alfalfa – – + A E E C C C D C
C. trifolii ON13 Alfalfa – – + A E E C C C D C
C. trifolii ON14 Alfalfa – – + A E E C C C D C
C. trifolii ON17 Alfalfa – – + A E E C C C D C
C. trifolii ON18 Alfalfa – – + A E E C C C D C
C. trifolii ON19 Alfalfa – – + A E E C C C D C
C. orbiculare LW1 1051 Cocklebur … … – A1 G G D E D D C
C. orbiculare LW6 1051 Cocklebur … … – A1 G G D E D D C
C. malvarum 3/7/2011 CM-1 Prickly sida … – – A F F D E D E C
C. malvarum 4/3/2012 CM-1 Prickly sida … – – A F F D E D E C
C. malvarum 4/3/2018 CM-1 Prickly sida … – – A F F D E D E C

1308 PHYTO PAT HO LOGY
described in Puhalla (36) and Correll et al. (10). Nitrate non-
utilizing mutants were recovered on minimal medium containing
1.5% potassium chlorate (10) as previously described (52). The
mutants were characterized into two phenotypes (nit1 and NitM),
whereby nit1 mutants grew as thin colonies on MM amended
with nitrate as the sole nitrogen source and, NitM mutants grew as
thin colonies on MM with hypoxanthine as the sole nitrogen
source (10). Vegetative compatibility tests were performed by
pairing NitM mutants with nit1 mutants in all possible combina-
tions. Vegetative compatibility tests were conducted at least twice.
Pairing among isolates where a distinct heterokaryon developed
as a result of complementation were vegetatively compatible and
were placed into the same VCG.
All isolates of C. lindemuthianum, C. trifolii, and C. malvarum
were tested for their vegetative compatibility with the reference
VCGs of C. orbiculare (51,52). In addition, a subset of isolates of
each of these species were characterized for vegetative compati-
bility with one another.
Pathogenicity tests. All isolates were tested for pathogenicity
on several cucurbit hosts, green bean, and alfalfa. Emerson
medium (YPSS) was used for producing inoculums (46). Cultures
were grown on YPSS and incubated at room temperature (≈23°C)
under a 16-and-8-h light-and-dark cycle, where light was pro-
vided by six fluorescent lights (34 W). Inoculum was prepared
from 7- to 10-day-old cultures by washing conidia off the agar
surface using cold, sterile, deionized water and straining the sus-
pensions through two layers of cheese cloth. The conidial suspen-
sions were adjusted to a spore concentration of 8 × 104 conidia/ml
using a hemacytometer.
Pathogenicity tests on cucurbits were conducted as previously
described (52). Four differential cultivars—cucumber cvs.
Marketer (Harris Moran Co.) and Arkansas Little Leaf (H19; Peto
Seed) and watermelon cvs. Black Diamond (Northrup King Co.)
and Charleston Gray (Northrup King Co.)—that have been shown
to distinguish races 1, 2, and 2B in cotyledon inoculations were
used (51,52). To ensure that all plants were of uniform age and
size at the time of inoculation, seed were pregerminated on moist
tissue paper in the dark at room temperature (23°C). Germinated
seed were planted in a soilless mixture of 1:1 peat and perlite
(Sunshine mix no. 1). Fully expanded cotyledons (4-day-old seed-
lings) were sprayed with the conidial suspension to run-off using
an air brush. Inoculated plants were incubated in a dew chamber
at a relative humidity (RH) of 100% (23 to 28°C) for 24 h. After
incubation, the plants were returned to the greenhouse and
maintained at 23 to 35°C.
Evidence of disease symptoms and severity ratings were ob-
tained by visually assessing the cotyledon area showing symp-
toms (chlorosis and necrosis) of infection as previously described
(52). A disease rating scale of 0 to 7 was used, where 0 = no
infection, 1 = 1 to 10, 2 = 11 to 25, 3 = 26 to 50, 4 = 51 to 75, 5 =
76 to 89, 6 = >90 to 98, and 7 = 100% necrosis or chlorosis. The
mean disease ratings on the eighth day were used to evaluate
pathogenicity. Isolates with mean disease ratings of 0.5 on the
susceptible cvs. Marketer and Black Diamond were considered
nonpathogenic. All cotyledon inoculation tests consisted of three
replicates (pot) and three plants per replicate. Disease ratings per
replicate were averaged from six cotyledons. The inoculation
experiments were conducted three times.
A subset of isolates representing the various taxa was used in
an inoculation test on cucumber fruit (Table 2). Cucumber fruit
from an unknown slicing cucumber type cultivar were purchased
at a grocery store. Fruit were washed with a commercial dish-
washing detergent and surface sterilized with 10% (vol/vol) solu-
tion of commercial bleach for 2 min. The fruit then were rinsed in
deionized water and air dried. Fruit were wounded with a cork
borer, making a hole 8 mm in diameter and ≈5 mm deep. The
holes then were inoculated by placing a colonized green bean
agar plug (8 mm in diameter) into the wounds. Uncolonized plugs
were used as controls. Fruit inoculation sites were a minimum of
4 cm apart. Following inoculation, fruit were incubated for 7 days
TABLE 2. Virulence of Colletotrichum orbiculare and allied Colletotrichum spp. on fruit of cucumber, Cucumis sativus
Fruit lesion
Isolate VCGw Racex mtDNA haplotypey Diameter (mm)z Vol u m e ( c m 3) Host origin
Colletotrichum orbiculare
JC1 1001 1 A 30.1 a 4.06 ab Cucumis sativus
BB18 1002 2 A 31.4 a 4.78 a C. melo
CR2 1002 2B A 31.6 a 4.36 ab Citrullus lanatus
MH2 1003 1 A 25.7 ab 2.47 bc Cucumis sativus
JX10 1004 2B A 29.3 a 3.74 ab C. melo
LW1 1050 av A1 11.3 d 0.32 d Xanthium spinosum
LW6 1050 av A1 11.0 d 0.29 d X. spinosum
Colletotrichum trifolii
4-1-32 CT-1 1 A 11.2 d 0.26 d Medicago sativa
14-2-63 CT-1 2 A 11.5 d 0.49 d M. sativa
14-3-1 CT-1 2 A 10.1 d 0.20 d M. sativa
14-3-3 CT-1 1 A 14.4 cd 0.52 d M. sativa
C. malvarum
3-7-11 CM-1 – A 10.4 d 0.26 d Sida spinosa
4-3-12 CM-1 – A 11.0 d 0.27 d S. spinosa
4-3-18 CM-1 – A 11.1 d 0.29 d S. spinosa
C. lindemuthianum
14-2-37 CL-1 Beta A2 10.7 d 0.25 d Phaseolus spp.
14-2-39 CL-1 Sigma A2 12.8 d 0.39 d Phaseolus spp.
14-2-45 CL-1 Lambda A2 19.6 bc 2.67 c Phaseolus spp.
JZ1 CL-1 Alpha – 11.6 d 0.26 d Phaseolus spp.
JZ2 CL-1 Beta – 13.3 cd 0.51 d Phaseolus spp.
14-2-41 CL-X Episilon A2 11.7 d 0.31 d Phaseolus spp.
w VCG = vegetative compatibility group.
x Races of C. orbiculare identified by Wasilwa et al. (51,52); races of C. trifolii identified by Ostazeski and Elgin (33) and Welty and Mueller (53); and races o
f
C. lindemuthiamum described by Zaumeyer and Meiners (55).
y Mitochondrial DNA (mtDNA) restriction fragment length polymorphism (RFLP) haplotypes based on patterns observed with HaeIII. The same letter within a
column indicates a common mtDNA RFLP haplotype; – = information not available.
z Each disease parameter is a mean of three replications from each of three independent experiments. Numbers followed by the same letter within a column are
not significantly different from one another (least significant difference = 0.05%).

Vol. 97, No. 10, 2007 1309
at 23°C at 100% RH. After 7 days, lesion diameters were re-
corded by measuring two perpendicular axes of the lesion surface.
A cross section through the center of the lesion was made and
lesion depth and width were measured. The lesions formed in a
cone shape from the point of inoculation and, as a result, the
lesion volume was estimated using the mathematical formula for
the volume of a cone. Visible sporulation on the fruit surface also
was recorded as either present or absent. All fruit inoculation tests
were conducted three times and each inoculation was treated as a
block. Data were analyzed as a randomized complete block by
analysis of variance with block–treatment used as the error term.
The least significant difference at the 5% probability level was
used for mean separation.
Green bean inoculation tests were performed similarly to those
previously described (21). Seedlings (10 to 14 days old) of the cv.
Cardinal were spray inoculated until run-off with a conidial
suspension (1 × 106 conidia/ml). Plants were incubated at ≈24°C
for 24 h at 100% RH before being moved to the greenhouse.
Seven days after inoculation, plants were scored for the severity
of anthracnose symptoms on leaves and stems. For the purpose of
this study, isolates were evaluated simply as pathogenic, causing
severe anthracnose symptoms; weakly pathogenic, showing only
a few anthracnose lesions; or nonpathogenic, showing no symp-
toms of infection. The inoculations were performed in two inde-
pendent inoculation tests.
Alfalfa inoculation tests were performed similarly to those previ-
ously described (4,30,31,33,53). Twenty-five seed of cv. Saranac
were grown per pot. Fourteen-day-old seedlings were inoculated
by spraying until run-off with a conidial suspension of each iso-
late (2 × 106 conidia/ml in sterile distilled water containing two
drops of Tween 20 per liter). Plants were placed in a mist cham-
ber maintained at 23°C for 48 h and then moved to a growth
chamber maintained at 23°C and a 16-h photoperiod. Eight days
after inoculation, the plants were evaluated for disease reactions
using a Horsfall-Barret scale based on the degree of necrosis,
where 0 = no disease and 9 = a dead plant. Isolates were con-
sidered nonpathogenic if plants were scored as 0 or 1. Isolates
were considered pathogenic if plants scored as 7 to 9. No plants
received intermediate scores of 2 to 6. Four replicate pots were
used per experiment per isolate and the entire experiment was
conducted twice.
Intron amplification. The forward primer GSF1 (5′-AT-
GGCCGAGTACATCTGG-3′) and the reverse primer GSR1 (5′-
GAACCGTCGAAGTTCCAC-3′) were used to amplify an ≈900-
bp intron region of the glutamine synthetase (GS) gene (17,24,
25,39,45). The forward primer GDF1 (5′-GCCGTCAACGAC-
CCCTTCATTGA-3′) and the reverse primer GDR1 (5′-GGGTG-
GAGTCGTACTTGAGCATGT-3′) were used to amplify a 200-bp
intron region of the glyceraldehyde-3-phosphate dehydrogenase
(GPDH) gene (17,24,25,44). A Hybaid DNA thermocycler was
used to perform the polymerase chain reaction amplification of
the introns using 35 cycles of denaturation at 94°C and annealing
at 60°C for 1 min, with final extension at 72°C for 3 min.
Amplified DNA was digested with the following enzyme com-
binations: A, HindIII + HinfI + HaeIII; and B, HindIII + HinfI +
MspI. The restriction fragments were electrophoretically sepa-
rated in a 3.0% Agarose 1000 (Invitrogen Corp., Carlsbad, CA)
gel in 0.5 Tris-borate-EDTA buffer for 3.5 h at 140 V. Band
fragments between 0.05 and 1.0 kb (Table 3) were scored for their
presence or absence for a given enzyme combination and the data
converted into a binary character matrix. The data then were ana-
lyzed using a cluster analysis of the similarity coefficients with
the unweighted pair-grouping method with arithmetic averages
(UPGMA) in NTSYS-pc software (Department of Ecology and
Evolution, State University of New York, Stony Brook, NY), to
determine the relative relatedness.
In addition, the software Quantitative One (Bio-Rad, Rich-
mond, CA) was used to determine the approximate size of the
various fragments based on the relative distance from the top
comparing with a 100-bp molecular marker (Invitrogen).
DNA sequencing. The 900- and the 200-bp intron fragments
were purified with the Qiagen QIAquick Gel Extraction Kit
(Qiagen, Inc., Valencia, CA), and used as templates for sequenc-
ing reactions using the ABI Prism Dye Terminator cycle sequenc-
ing system (Applied Biosystems Inc., Foster City, CA). Sequenc-
ing reactions were performed as previously described (17).
Sequence alignment and phylogenetic analysis. The se-
quence of the 900- and 200-bp introns of 19 isolates used for
phylogenetic analyses were performed using four isolates as an
outgroup; the four isolates, representing other Colletotrichum
spp., included C. dematium (JG13) from spinach, C. acutatum
(A38) from apple, C. gloeosporioides (NC329) from apple, and
C. magna (AK7) from pumpkin (17).
Intron sequences were entered into the Seqpup DNA sequence
editor as previously described (14). The combined data were
aligned using ClustalX (43), and the phylogenetic analyses and
partition homogeneity test were performed using PAUP (42).
Three methods of tree building were used: maximum-parsimony
(MP), maximum-likelihood (ML), and neighbor-joining (NJ). For
each method, alignment gaps were treated as missing data in the
TABLE 3. Approximate base-pair fragment size of the intron polymerase chain reaction products digested by the restriction enzyme combination HindIII + HinfI +
HaeIII (HHH) or HindIII + HinfI + MspI (HHM)
Colletotrichum
orbiculare
C.
lindemuthianum
C. trifolii
C. orbiculare
(Xanthium sp.)
C. malvarum
C. dematium
C. acutatum
C.
gloeosporioides
C. magna
HHH RFLPsy
240 260 240 × 2z 240 260 190 230 230 250
210 220 140 190 170 140 × 2z 190 190 135
140 140 110 140 130 135 140 150 120
120 120 95 110 105 110 120 135 90
95 95 70 95 90 90 105 110 80
80 90 60 70 80 40 × 2z 90 90 60
70 60 45 60 60 60 60 50
45 45 45 55 50 55 30
HHM RFLPs
430 500 430 430 500 270 230 270 200
250 230 250 250 250 180 190 250 180
145 170 170 130 × 2z 130 130 170 220 170
75 70 75 45 45 120 130 190 140
70 45 70 85 65 150 130
45 45 30 140 90
130
y RFLPs = restriction fragment length polymorphisms.
z Two fragments with equal molecular weights existed in the digested band patterns.

1310 PHYTO PAT HO LOGY
phylogenetic analyses. The tree topology was evaluated by statis-
tical confidence using bootstrap analysis; 1,000 replicates were
performed to examine the relative bootstrap support for each
group in the resultant topologies (14).
For MP and ML analyses, a heuristic search was employed and
starting trees always were obtained by random sequence addition.
Tree visualization was done using TreeView (Win32) version
1.5.2.
The Hasegawa-Kishino-Yano model (HKY85) was used for NJ
tree construction (19). Each indel (insertion or deletion), regard-
less of its length, was treated as a single nucleotide substitution.
For the MP analysis, the heuristic search procedure was used with
the following parameters: transition/transversion ratio = 2; start-
ing branch length was obtained using the Rogers-Swofford ap-
proximation method; the substitution rates were set to conform to
a gamma distribution; and the molecular clock was not enforced.
For the ML analysis, the heuristic algorithm with TBR of PAUP
was used because the data set was too large to be used with the
exhaustive or branch-and-bound algorithms. ML settings were as
follows: number of substitution types = 2; transition/transversion
ration = 2; kappa = 4.027. The assumed nucleotide frequency
(empirical frequencies) A = 0.25788, C = 0.31081, G = 0.20218,
and T = 0.22914. The assumed proportion of invariable sites =
none; distribution rates at variable sites = equal.
The tree length CI, the CI excluding uninformative characters,
the homoplasy index (HI), the HI excluding uninformative
characters, the RI, and the rescaled consistency index (RC) for all
of the MP trees were recorded. The Kishino-Hasegawa tests were
performed to determine whether the trees stored in memory were
significantly different. The NJ tree distance matrix was calculated
based on the HKY85 model.
GenBank accession numbers for two intron sequences. The
900-bp intron sequences were deposited as follows: DQ792871
(isolate A38), DQ792872 (NC329), DQ792873 (AK7),
DQ792874 (JG13), DQ792875 (ON12), DQ792876 (ON7),
DQ792877 (14-2-63), DQ792878 (JK10), Q792879 (14-2-39),
DQ792880 (JK7), DQ792881 (JK12), DQ792882 (LW6),
DQ792883 (LW1), DQ792884 (AK9), DQ792885 (DAR61396),
DQ792886 (MH2), DQ792887 (JX10), DQ792888 (JX13),
DQ792889 (JC1), DQ792890 (CP6), DQ792891 (RL1),
DQ792892 (3-7-11), and DQ792893 (4-3-12); the 200 bp intron
sequences deposited were as follows: DQ792848 (A38),
DQ792849 (NC329), DQ792850 (AK7), DQ792851 (JG13),
DQ792852 (ON12), DQ792853 (ON7), DQ792854 (14-2-63),
DQ792855 (JK10), DQ792856 (14-2-39), DQ792857 (JK7),
DQ792858 (JK12), DQ792859 (LW6), DQ792860 (LW1),
DQ792861 (AK9), DQ792862 (DAR61396), DQ792863 (MH2),
DQ792864 (JX10), DQ792865 (JX13), DQ792866 (JC1),
DQ792867 (CP6), DQ792868 (RL1), DQ792869 (3-7-11), and
DQ792870 (4-3-12).
RESULTS
mtDNA RFLP. The geographically diverse isolates of C. or-
biculare from various cucurbit hosts representing the four known
VCGs of C. orbiculare (VCGs 1001, 1002, 1003, and 1004) and
including races 1, 2, and 2B, had a common mtDNA RFLP haplo-
type (haplotype A) when examined with three restriction enzymes
(EcoRI, PvuII, and HaeIII) and probed with two mtDNA clones
of C. orbiculare (404U [13.7 kb] and 2U18 [10.1 kb]) (Table 1;
Fig. 1). Also, two representative isolates of C. orbiculare from
cocklebur from Australia had a similar mtDNA haplotype A, with
one polymorphism detected with the enzyme PvuII (Table 1).
This polymorphism previously was shown to be due to the occur-
rence of one additional restriction site in the mitochondrial ge-
nome among isolates from cocklebur (11).
The isolates of C. trifolii, C. lindemuthianum, and C. malvarum
had an mtDNA haplotype identical to haplotype A of C. orbicu-
lare when examined with two (EcoRI and PvuII) different restric-
tion enzymes. The isolates of C. orbiculare, C. trifolii, and
C. malvarum also had an identical mtDNA haplotype (haplotype
A) when examined with a third restriction enzyme (HaeIII). How-
ever, a single mtDNA RFLP polymorphism (1.7 kb) was detected
among all of the isolates of C. lindemuthianum with the enzyme
HaeIII (Fig. 1).
Vegetative compatibility. nit Mutants were recovered from
isolates of all four taxa (C. orbiculare, C. trifolii, C. malvarum,
and C. lindemuthianum) on minimal medium amended with po-
tassium chlorate after 10 to 40 days (47,52). All isolates of C. or-
biculare that originated from cucurbit hosts belonged to one of
four distinct VCGs (VCGs 1001, 1002, 1003, and 1004) (Table
1). Isolates of C. orbiculare from cocklebur (LW1 and LW6) were
vegetatively compatible with each other but vegetatively incom-
patible with isolates in VCGs 1001, 1002, 1003, and 1004. In
addition, all isolates of C. trifolii, C. lindemuthianum, and C. mal-
varum were vegetatively incompatible with isolates of all four
VCGs of C. orbiculare and with each other (Table 1). Four
isolates of C. trifolii tested were in the same VCG and designated
VCG CT-1. The three isolates of C. malvarum, also in mtDNA
haplotype A, belonged to a single VCG designated VCG CM-1.
Five of the seven isolates of C. lindemuthianum examined
belonged to a single VCG (CL-1).
Foliar pathogenicity tests. Overall, the cucurbit isolates of
C. orbiculare, the green bean isolates of C. lindemuthianum, and
the alfalfa isolates of C. trifolii were pathogenic on cucurbits,
green bean, and alfalfa, respectively (Table 1). The cocklebur iso-
lates of C. orbiculare and the C. malvarum isolates from prickly
sida were not pathogenic on any of these three hosts.
All isolates of C. orbiculare that originated from cucurbit hosts
were pathogenic on the susceptible cucumber (Marketer) and
watermelon (Black Diamond) differentials (disease ratings 5.0)
(Table 1). H19 was resistant (disease ratings 2.5) to races 2 and
2B and Charleston Gray was resistant to races 1 and 2B (52). The
mean disease ratings of isolates of C. orbiculare from cocklebur
and isolates of C. trifoii, C. lindemuthianum, and C. malvarum all
were <0.5 and no evidence of infection was observed for the
isolates examined; thus, were all considered nonpathogenic on the
susceptible cucurbit hosts (Table 1). Furthermore, several isolates
of all of the species were tested in additional pathogenicity tests
whereby inoculum concentrations were increased to 4 × 106 co-
nidia/ml and, again, no infections were observed on the cotyle-
dons of any of the cucurbit differentials (data not shown). All
Fig. 1. Mitochondrial (mt)DNA restriction fragment length polymorphisms o
f
the various Colletotrichum spp. Total DNA was digested with HaeIII and probe
d
with two mtDNA clones (4u40 and 2u18) of Colletotrichum orbiculare (11).

Vol. 97, No. 10, 2007 1311
isolates of C. lindemuthianum tested were pathogenic on green
bean whereas no other isolates were observed to cause disease on
green bean. Similarly, only isolates of C. trifolii caused disease on
alfalfa; one isolate from alfalfa (ON3) did not cause disease on
alfalfa.
Fruit pathogenicity tests. Although all taxa were able to
colonize wounded cucumber fruit, only isolates of C. orbiculare
from cucurbits (except MH2) had significantly larger lesion diam-
eters and lesion volumes than isolates of C. orbiculare from
cocklebur, C. trifolii, C. lindemuthianum, and C. malvarum (Table
2). There were no significant differences in lesion diameter or
lesion volume among isolates representing VCGs 1001, 1002,
1003, and 1004 of C. orbiculare (Table 2). Conidial production
was evident on the cucumber fruit surface after 7 days for all
cucurbit isolates of C. orbiculare. No sporulation was visible on
the lesion surface for isolates of C. trifolii, C. lindemuthianum,
C. malvarum, or C. orbiculare from cocklebur.
Intron RFLPs and sequence. An intron of the GS gene and an
intron of the GPDH gene were amplified successfully from 19
representative isolates of C. orbiculare, C. trifolii, C. lindemuthia-
num, and C. malvarum (Table 1). Although some variation in size
was observed, the fragment from the various species was ≈900 bp
(864 to 912 bp) for GS intron and ≈200 bp (97 to 193) for GPDH
intron. The lower size range for the GPDH intron among the iso-
lates of C. lindemuthianum was due to a series of relatively large
deletions near the 3′ end of the intron.
Various restriction enzymes were able to cut the 900-bp GS in-
tron, but a combination of enzymes, particularly HindIII + HinfI +
HaeIII and HindIII + HinfI + MspI, produced highly polymorphic
profiles with seven to nine polymorphic bands and provided the
highest level of resolution for qualitatively distinguishing each
species (Fig. 2; Table 3). In addition, isolates of C. orbiculare
from cucurbit hosts could be distinguished from those from
cocklebur (Fig. 2). UPGMA analysis based on the RFLP data
from the 900-bp GS intron clearly distinguished the various taxa
(Fig. 3).
The 900-bp intron of the GS gene and the 200-bp GPDH intron
of the GPDH gene from 19 isolates of C. orbiculare, C. trifolii,
C. lindemuthianum, and C. malvarum also were sequenced and
compared. The overall sequence similarity among isolates of
C. orbiculare, C. trifolii, C. lindemuthianum, and C. malvarum
was >90% for the 900-bp intron and >86% for the 200-bp intron
(except for C. lindemuthianum, which was 50 to 53% due to large
deletions near the 3′ end of the intron sequence). The overall
sequence similarity of the 900- and 200-bp intron among isolates
within each species was >97%.
Interestingly, isolates of C. orbiculare described from Xanthium
spp. in Australia (41) had a 900-bp intron sequence (>99.9%) very
similar to isolates of C. malvarum from the United States. In
addition, the 200-bp intron sequence of Xanthium isolates was
95% similar to those of C. malvarum and 98% similar to isolates
of C. orbiculare.
Various isolates representing C. dematium, C. acutatum,
C. gloeosporioides, and C. magna as outgroups showed little
sequence similarity to isolates in the C. orbiculare complex and
were <64 and <30% for the 900- and 200-bp introns, respectively.
Overall, the statistical clustering of isolates based on all three
tree topologies were very similar for both the 900- and 200-bp
intron as well as the combined data set of both introns (only MP
trees shown) (Figs. 4, 5, and 6). For the 900-bp intron, all three
tree methods grouped all of the taxa examined (excluding the out-
groups) into a single clade with 100% bootstrap values for both
Fig. 2. Restriction fragment length polymorphisms of the 900-bp intron of the
glutamine synthetase gene of various Colletotrichum spp. The amplified
product (≈900 bp) was digested with a combination of three enzymes (A,
HindIII + HinfI + HaeIII and B, HindIII + HinfI + MspI) and the fragments
then separated in a 3% agarose 1000 gel.
Fig. 3. Cluster analysis of the 900-bp intron based on the presence or absence
of the bands of the glutamine synthetase intron restriction fragment length
polymorphisms digested with enzyme combination HindIII + HinfI + HaeIII.
The dendrogram was generated using the unweighted pair-grouping method
with arithmetic averages of NTSYS-PC.

1312 PHYTO PAT HO LOGY
introns. The high CI and low HI values clearly indicated the
phylogenetic relationship with a high degree of confidence with
intron sequence data.
For the 900-bp intron, tree topologies of the sequence data for
all three methods (MP, ML, and NJ) consistently grouped
C. orbiculare, C. trifolii, and C. malvarum together and separate
from C. lindemuthianum (Fig. 4, MP tree shown). However, with
the 200-bp intron, C. lindemuthianum was nested with C. trifolii
(Fig. 5). This may be due to a smaller number of bases compared
(only 97) due to the deletions in the 200-bp intron among the
isolates of C. lindemuthianum, and these data may not be as ro-
bust as that from the 900-bp intron.
Of the total 1,313 characters in the 900-bp GS intron sequence,
172 were phylogenetically informative in the MP analysis. The
tree length was 491 steps, the CI was 0.8916, the CI excluding
uninformative characters was 0.8083, the HI was 0.1039, the HI
excluding uninformative characters was 0.1917, the RI was 0.948,
and the RC was 0.8019 (Fig. 4).
Of the total 303 characters in the 200-bp intron sequence, 55
were phylogenetically informative in the MP analysis. The tree
length was 152 steps, the CI was 0.8487, the CI excluding un-
informative characters was 0.7500, the HI was 0.1513, the HI ex-
cluding uninformative characters was 0.2500, the RI was 0.8631,
and the RC was 0.7325 (Fig. 5).
For the combined 900- and 200-bp intron sequences in the
MP analysis, there were 1,610 characters, of which 229 were
phylogenetically informative. The tree length was 660 steps, the
CI was 0.9045, the CI excluding uninformative characters was
0.8147, the HI was 0.0955, the HI excluding uninforma-
tive characters was 0.1826, the RI was 0.9023, and the RC was
0.8162 (Fig. 6).
Before the two intron sequences were combined, the partition
homogeneity test was used to compare the sequences. The P value
of the partition homogeneity test for the two intron sequence data
sets were calculated using PAUP. The high P value (P = 0.79)
from the two sets of intron sequences in this study indicated that
there is no significant difference between the two sequence data
sets and, thus, two intron sequences can combine to infer phylo-
genetic relationship. Moreover, the similar tree topologies based
on two intron sequences and the combined sequences indicate that
the considerable concordance existed between the two intron
sequences.
DISCUSSION
There is considerable uncertainty in the literature regarding the
demarcation of taxa of many Colletotrichum spp., including the
species complex C. orbiculare, which includes C. orbiculare,
C. trifolii, C. lindemuthianum, and C. malvarum. Spore morphol-
ogy, appressorium development, and sequence similarities of the
rDNA have been used to infer that C. orbiculare from cucurbits,
C. trifolii from alfalfa, C. lindemuthianum from bean, and C. mal-
varum from prickly sida represent a single phylogenetic species
and should be collectively recognized as C. orbiculare (2,27,
32,38). In the current study, a common mtDNA RFLP haplotype,
haplotype A, was observed among representative isolates of
C. orbiculare, C. trifolii, and C. malvarum. Isolates of C. linde-
muthianum had an mtDNA haplotype very similar to haplotype A,
with a single mtDNA RFLP detected with the enzyme HaeIII.
These data provide supporting evidence that these taxa clearly
have a relatively recent common ancestry. The fact that these taxa
have such a similar mtDNA haplotype may indicate that they may
Fig. 4. Maximum-parsimony (MP) tree based on the glutamine synthetase
intron sequence showing the relationship among species in the Colletotrichum
orbiculare complex. MP tree scores were as follows: 1,313 total characters,
tree length = 491, consistency index = 0.8961, homoplasy index = 0.1039,
retention index =0.8948, and rescaled consistency index = 0.8019. Bootstrap
values are labeled on the branch of the tree. Scale bar represents the number o
f
transformations from one character to another.
Fig. 5. Maximum-parsimony (MP) tree based on glyceraldehyde-3-phosphate
dehydrogenase intron sequence showing the relationship of Colletotrichum
orbiculare complex. MP tree scores were as follows: 303 total characters, tree
length = 152, consistency index = 0.8487, homoplasy index = 0.1513, reten-
tion index =0.8631, and rescaled consistency index = 0.7325. Bootstrap values
are labeled on the branch of the tree. Scale bar represents the number o
f
transformations from one character to another.

Vol. 97, No. 10, 2007 1313
even belong to a common ancestral mating population. Isolates of
C. lindemuthianum have been shown to be heterothallic and pro-
duce the Glomerella lindemuthiana sexual stage (37). However,
little is know about mating abilities between these different taxa.
Sequence analysis of two independent introns in the current
study were phylogenetically informative and clearly show that
isolates of C. orbiculare, C. trifolii, C. lindemuthianum, and
C. malvarum represent a single well-supported clade. Although
similar, each of the taxa, including isolates of C. orbiculare that
were pathogenic and nonpathogenic to cucurbits, could consis-
tently be demarcated based on the qualitative (RFLPs) and quanti-
tative (sequence) analyses of the two introns. Similarly, sequence
information on both of these introns also has been phylogeneti-
cally informative in C. acutatum, whereby distinct clades could
be demarcated within the broader species complex (17). Thus,
sequence information of these two introns had a very high level of
resolution relative to other target DNA sequences such as ITS
regions, was statistically robust, and, consequently, can be phylo-
genetically informative at both the inter- and intraspecific levels
for the genus Colletotrichum.
The intron sequences that have been examined appear universal
in Colletotrichum spp. in that they have been recovered from a
wide array of taxa, including C. acutatum, C. gloeosporioides,
C. orbiculare, C. lindemuthianum, C. trifolii, C. malva, C. magna,
C. dematium, C. graminicola, C. sublineolum, and several Col-
letotrichum taxa from turfgrass (8,16,17,24,25,50). Because of the
RFLP diversity observed between the taxa, it is anticipated that
the RFLP haplotypes identified from the restriction digestion of
the 900-bp intron could be a valuable diagnostic tool for species
identification, particularly where time, expertise, and expense
may preclude sequencing efforts from being conducted.
Although a limited number of isolates were examined for vege-
tative compatibility, none of the isolates representing C. orbicu-
lare, C. trifolii, C. lindemuthianum, and C. malvarum were
vegetatively compatible with each other. Among isolates of
C. orbiculare from cucurbits, the four VCGs 1001, 1002, 1003,
and 1004 were identified (9,51,52). There are a limited number of
studies that have used vegetative compatibility (nit mutants) to
study population diversity of C. trifolii (6). In the current study,
four isolates of C. trifolii representing two races (races 1 and 2)
belonged to a single VCG (CT-1). Isolates of C. malvarum from
prickly sida (S. spinosa) from Arkansas, Mississippi, and Ten-
nessee also belonged to a single VCG (CM-1), indicating that this
VCG has a wide geographical distribution in the United States.
Data from the current study support the hypothesis that C. or-
biculare, C. lindemuthianum, C. trifolii, and C. malvarum repre-
sent a closely aligned phylogenetic species (43). However, iso-
lates of each taxa showed a distinct host specificity whereby only
isolates of C. orbiculare, C. trifolii, and C. lindemuthianum were
pathogenic on cucurbits, alfalfa, and bean, respectively. These
data support the host specificity of C. lindemuthianum to legumes
(21,55), C. trifolii to alfalfa and red clover (Trifolium pratense L.)
(4,33), and C. malvarum to hosts in the Malvaceae (prickly sida
and hollyhock [Althaena rosea L.]) (20). Kirkpatrick et al. (22)
showed that isolates of C. malvarum were unable to infect 38
plant species, including cucurbits (Citrullus lanatus (Thunb.)
Matsum and Nakai (watermelon), Cucumis sativus L. (cucumber),
C. melo L. (muskmelon), Cucurbita pepo (pumpkin), or P. vulgaris
L. (bean). Because isolates within the broader species concept of
Colletotrichum orbiculare (which includes C. lindemuthianum,
C. trifolii, and C. malvarum) clearly show host specificity to
various hosts, it is proposed that this fixed phenotypic charac-
teristic can be used to distinguish the species. Recent recom-
mendations support the use of phenotypic characteristics, such as
host specialization, to be useful at the species rank rather than as
the informal rank of formae speciales (18). Consequently, because
C. orbiculare from cucurbit hosts, C. orbiculare from noncucurbit
hosts, C. lindemuthianum, C. trifolii, and C. malvarum can be dis-
tinguished based on apparently fixed genotypic (intron nucleotide
sequence differences) and phenotypic (host specificity) characters,
they should be recognized as distinct but closely related species.
Molecular studies to examine species boundaries in Colletot-
richum will continue to provide a better understanding of inter-
and intraspecific variation within the genus, particularly within
the broader species complexes such as C. acutatum sensu lato, C.
gloeosproioides sensu lato, and C. graminicola sensu lato. How-
ever, it is important, as we continue to demarcate individual taxa
within the genus, that we have a concomitant effort on developing
a better understanding of the lifestyles (35,54), the ecology, and
overall biology of these individual taxa and their impact and
control in agricultural ecosystems.
LITERATURE CITED
1. Alvarez, I., and Wendel, J. F. 2003. Ribosomal ITS sequences and plant
phylogenetics inference. Mol. Phylogenet. Evol. 29:417-434.
2. Bailey, J. A., Nash, C., Morgan, L. W., O’Connel, R. J., and TeBeest, D.
O. 1996. Molecular taxonomy of Colletotrichum species causing
anthracnose of Malvaceae. Phytopathology 86:1076-1083.
3. Bailey, J. A., Sherriff, C., and O’Connell, R. J. 1995. Identification of
specific and intraspecific diversity in Colletotrichum. Pages 197-211 in:
Disease Analysis Through Genetics and Biotechnology—Interdisciplinary
Bridges to Improved Sorghum and Millet Crops. J. H. Leslie and R. A.
Fredericksen, eds. Iowa State University Press, Ames.
4. Bain, S. M., and Essary, S. H. 1906. A new anthracnose of alfalfa and red
clover. J. Mycol. 12:192-193.
5. Balardin, R. S., Smith, J. J., and Kelly, J. D. 1999. Ribosomal DNA poly-
morphism in Colletotrichum lindemuthianum. Mycol. Res. 103:841-848.
6. Brooker, N. L., Leslie, J. F., and Dickman, M. B. 1991. Nitrate non-
utilizing mutants of Colletotrichum and their use in studies of vegetative
compatibility and genetic relatedness. Phytopathology 81:672-677.
Fig. 6. Maximum-parsimony (MP) tree based on the combined glutamine
synthetase and glyceraldehyde-3-phosphate dehydrogenase intron sequences
showing the relationship of Colletotrichum orbiculare complex. MP tree
scores were as follow: 1,610 total characters, tree length = 660, consistency
index = 0.9045, homoplasy index = 0.0955, retention index =0.9023, and re-
scaled consistency index = 0.8162. Bootstrap values are labeled on the branch
of the tree. Scale bar represents the number of transformations from one
character to another. P = 0.79 (partition homogeneity test) indicating, con-
siderable concordance between the evolutionary ratio of the two intron
sequences.

1314 PHYTO PAT HO LOGY
7. Correll, J. C., Gordon, T. R., and McCain, A. H. 1992. Genetic diversity
in California and Florida populations of pitch canker fungus Fusarium
subglutinans f. sp. pini. Phytopathology 82:415-420.
8. Correll, J. C., Guerber, J. C., Bentley, T., and Uddin, W. 2003. Charac-
terization of Colletotrichum graminicola from turfgrass using mtDNA and
intron RFLPs, vegetative compatibility, colony and spore morphology,
and intron sequence analysis. (Abstr.) Phytopathology 93:S112.
9. Correll, J. C., Guerber, J. C., Wasilwa, L. A., Sherrill, J. F., and Morelock,
T. E. 2000. Inter- and intra-species variation in Colletotrichum and
mechanisms which affect population structure. Pages 145-179 in:
Colletotrichum: Host Specificity, Pathogenicity, and Host–Pathogen
Interactions. D. Prusky, M. B. Dickman, and S. Freeman, eds. American
Phytopathological Society, St. Paul, MN.
10. Correll, J. C., Klittich, C. J. R., and Leslie, J. F. 1987. Nitrate nonutilizing
mutants of Fusarium oxysporum and their use in vegetative compatibility.
Phytopathology 77:1640-1645.
11. Correll, J. C., Rhoads, D. D., and Guerber, J. C. 1993. Examination of
mitochondrial DNA restriction fragment length polymorphisms, DNA
fingerprints, and randomly amplified polymorphic DNA of Colletot-
richum orbiculare. Phytopathology 83:1199-1204.
12. Crouch, J. A., Clarke, B. B., and Hillman, B. I. 2006. Unraveling evolu-
tionary relationships among the divergent lineages of Colletotrichum
causing anthracnose disease in turfgrass and corn. Phytopathology 96:46-
60.
13. Du, M. Z. Schardl, C. L., Nuckles, E. M., and Vaillancourt, E. M. 2005.
Using mating-type gene sequences for improved phylogenetic resolution
of Colletotrichum species complexes. Mycologia 97:641-658.
14. Felsenstein, J. 1985. Confidence limits on phylogenies. An approach
using the bootstrap. Evolution 39:783-791.
15. Freeman, S., Katan, T., and Shabi, E. 1998. Characterization of Colletot-
richum species responsible for anthracnose diseases of various fruits.
Plant Dis. 82:596-605.
16. Gonzalez, E., Sutton, T. B., and Correll, J. C. 2006. Clarification of the
etiology of Glomerella leaf spot and bitter rot of apple caused by
Colletotrichum spp. based on morphology and genetic, molecular, and
pathogenicity tests. Phytopathology 96:982-992
17. Guerber, J. C., Liu, B., Correll, J. C., and Johnston, P. R. 2003.
Characterization of diversity in Colletotrichum acutatum sensu lato by
sequence analysis of two gene introns, mtDNA and intron RFLPs, and
mating compatibility. Mycologia 95:872-895.
18. Harrington, T. C., and Rizzo, D. M., 1999. Defining species in the fungi.
Pages 43-71 in: Structure and Dynamics of Fungal Populations. J. J.
Worrall, ed. Kluwer Press, Dordrecht, The Netherlands.
19. Hasegawa, M., Kishino, J., and Yano, T. 1985. Dating of the human - ape
splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol.
21:160-174.
20. Johnston, P. R., and Jones, D. 1997. Relationships among Colletotrichum
isolates from fruits-rots assessed using rDNA sequences. Mycologia
89:420-430.
21. Kelly, J. D., Afanador, L., and Cameron, L. S. 1994. New races of
Colletotrichum lindemuthianum in Michigan and implication in dry bean
resistance breeding. Plant Dis. 78:892-894.
22. Kirkpatrick, T. L., Templeton, G. E., and TeBeest, D. O. 1982. Potential
of Colletotrichum malvarum for biological control of prickly sida. Plant
Dis. 66:323-325.
23. Lenné, J. M. 1992. Colletotrichum diseases of legumes. Pages 134-166 in:
Colletotrichum: Biology, Pathology and Control. J. A. Bailey and M. J.
Jeger, eds. CAB International, Wallingford, UK.
24. Liu, B., and Correll, J. C. 2000. Characterization of Colletotrichum
species by analysis of a 1 kb intron of the glutamine synthetase gene.
(Abstr.) Phytopathology 90:S8.
25. Liu, B., Guerber, J. C., and Correll, J. C. 2001. Genetic diversity within
and between subgroups of Colletotrichum acutatum based on RFLP and
sequence analysis of introns of the glutamine synthetase and glyceralde-
hyde phosphate dehydrogenase genes. (Abstr.) Phytopathology 91:S55.
26. Mesquita, A. G. , Paula, T. J., Moreira, M. A., and de Barros, E. G. 1998.
Identification of races of Colletotrichum lindemuthianum with aid of
PCR-based molecular markers. Plant Dis. 82:1084-1087.
27. O’Connell, R. J., Nash, C., and Bailey, J. A. 1992. Lectin cytochemistry:
A new approach to understanding cell differentiation, pathogenesis and
taxonomy in Colletotrichum. Pages 67-87 in: Colletotrichum: Biology,
Pathology and Control. J. A. Bailey and M. J. Jeger, eds. CAB
International, Wallingford, UK.
28. O’Donnell, K., Kistler, H. C., Cigelnik, E., and Ploetz, R. 1998. Multiple
evolutionary origins of the fungus causing Panama disease of banana:
Concordant evidence from nuclear and mitochondrial gene genealogies.
Proc. Natl. Acad. Sci. USA 95:2044-2049.
29. O’Donnell, K., Ward, T. J., Geiser, D. M., Kistler, H. C., and Takayuki, A.
2004. Genealogical concordance between the mating type locus and seven
other nuclear genes supports formal recognition of nine phylogenetically
distinct species within the Fusarium graminearum clade. Fungal Genet.
Biol. 41:600-623.
30. O’Neill, N. R., and Bauchan, G. R. 2000. Sources of resistance to
anthracnose in the annual Medicago core collection. Plant Dis. 84:261-
267.
31. O’Neill, N. R., Elgin, J. H., and Baker, C. J. 1989. Characterization of
induced resistance to anthracnose in alfalfa by races, isolates, and species
of Colletotrichum. Phytopathology 79:750-756.
32. O’Neill, N. R., van Berkum, P., Lin, J., Kuo, J., Ude, G. N, Kenworthy,
W., and Saunders, J. A. 1997. Application of amplified restriction
fragment polymorphism for genetic characterization of Colletotrichum
pathogens of alfalfa. Phytopathology 87:745-750.
33. Ostazeski, S. A., and Elgin, J. H., Jr. 1980. Physiologic races of
Colletotrichum trifolii. (Abstr.) Phytopathology 70:691.
34. Pain, N. A., O’Connell, R. J., Bailey, J. A., and Green, J. R. 1992. Mono-
clonal antibodies which restricted binding of four Colletotrichum species:
C. lindemuthianum, C. malvarum, C. orbiculare, and C. trifolii. Physiol.
Mol. Plant Pathol. 40:111-126.
35. Perez, N. A., Timmer, L. W., Adaskaveg, J. E., and Correll, J. C. 2005.
Lifestyles of Colletotrichum acutatum. Plant Dis. 89:784-796.
36. Puhalla, J. E. 1985. Classification of strains of Fusarium oxysporum on
the basis of vegetative compatibility. Can. J. Bot. 63:179-183.
37. Rodrigguez-Guerra, R., Ramirez-Rueda, M. T., Cabral-Encisco, M.,
Garcia-Serrano, M., Lira-Maldonado, Z., Gonzalez-Chavira, M., and
Simpson, J. 2005. Heterothallic mating observed between Mexican iso-
lates of Glomerella lindemuthiana. Mycologia 97:793-803.
38. Sherriff, C., Whelan, M. J., Arnold, G. M., Lafay, J., Brygoo, Y., and
Bailey, J. A. 1994. Ribosomal DNA sequence analysis reveals new species
groupings in the genus Colletotrichum. Exp. Mycol. 18:121-138.
39. Stevenson, S. A., Green, J. R., Manners, J. M., and Maclean, D. J. 1997.
Cloning and characterization of glutamine synthetase from Colletot-
richum gloeosporioides and demonstration of elevated expression during
pathogenesis on Stylosanthes guianensis. Curr. Genet. 31:447-454.
40. Sutton, B. C. 1980. The Coelomycetes—Fungi Imperfecti with Pycnidia
Acervuli and Stromata. Commonwealth Mycological Institute, Kew,
Surrey, England.
41. Sutton, B. C. 1992. The genus Glomerella and its anamorph Colletot-
richum. Pages 523-537 in: Colletotrichum: Biology, Pathology and
Control. J. A. Bailey and M. J. Jeger, eds. CAB International, Walling-
ford, UK.
42. Swofford, D. L. 2002. PAUP. Phylogenetic Analysis Using Parsimony
(and Other Methods), version 4. Sinauer Associates, Sunderland, MA.
43. Taylor, J. W., Jacobson, D. J., Kroken, S., Kasuga, T., Geiser, D. M.,
Hibbett, D. S., and Fisher, M. C. 2000. Phylogenetic species recognition
and species concepts in fungi. Fungal Genet. Biol. 31:21-32.
44. Templeton, M. D., Rikkerink, E. H., Solon, S. L., and Crowhurst, R. N.
1992. Cloning and molecular characterization of the glyceraldehyde-3-
phosphate dehydrogenase gene-encoding gene and cDNA from the plant
pathogenic fungus Glomerella cingulata. Gene 122:225-230.
45. Thompson, J. D., Higgins, D. G., and Gibson, T. J. 1994. CLUSTAL W:
improving the sensitivity of progressive multiple sequence alignment
through sequence weighting, positions-specific gap penalties and weight
matrix choice. Nucleic Acids Res. 22:4673-4680.
46. Tuite, J. 1969. Plant Pathological Methods. Burgess Publ. Co., Min-
neapolis, MN.
47. Von Arx, J. A. 1957. Die arten der gattung Colletotrichum cda.
Phytopathol. Z. 29:413-468.
48. Von Arx, J. A. 1970. A revision of the fungi classified as Gloeosporium.
Bibl. Mycol. 24:1-203.
49. Walker, J., Nikandrow, A., and Millar, G. D. 1991. Species of Colletot-
richum on Xanthium (Asteraceae) with comments on some taxonomic and
nomenclatural problems in Colletotrichum. Mycol. Res. 95:1175-1193.
50. Ware, S. B. 2000. Genetic diversity of Colletotrichum sublineolum based
on vegetative compatibility, mtDNA and nuDNA RFLPs, and intron
sequences. M.S. thesis, University of Arkansas, Fayetteville.
51. Wasilwa, L. A. 1997. Characterization of Colletotrichum orbiculare and
allied Colletotrichum species using mtDNA RFLPs, vegetative compati-
bility, and virulence on cucurbits, and screening cucumber and water-
melon for anthracnose resistance. Ph.D. diss., University of Arkansas,
Fayetteville.
52. Wasilwa, L. A., Correll, J. C., Morelock, T. E., and McNew, R. 1993. Re-
examination of races of the cucurbit anthracnose pathogen, Colletot-
richum orbiculare. Phytopathology 83:1190-1198.
53. Welty, R. E., and Mueller, J. P. 1979. Occurrence of a highly virulent
isolate of Colletotrichum trifolii on alfalfa in North Carolina. Plant Dis.
Rep. 63:666-670.
54. Wharton, P. S., and Dieguez-Uribeondo, J. 2004. The biology of
Colletotrichum acutatum. Anals JarBontanico Madrid 61:3-22.
55. Zaumeyer, W. J., and Meiners, J. P. 1975. Disease resistance in beans.
Annu. Rev. Phytopathol. 13:313-334.