INTRODUCTION
Erwinia rhapontici (Millard) Burkholder is a bacterial
pathogen that causes a variety of plant diseases, including
pink seed of cereal and pulse crops, as well as soft rots of
horticultural crops. Examples of diseases of horticultural
crops caused by E. rhapontici include soft rot of wasabi
(Eutrema wasabi Maxim.) (Goto and Matsumoto, 1986),
crown rot of rhubarb (Rheum rhaponticum L.) (Millard,
1924; Metcalfe, 1940; Letal, 1976), soft rot of onion (Al-
lium cepa L.) (Ohuchi et al., 1983) and others (Huang et
al., 2003b). Pink seed disease is found in crops such as
pea (Pisum sativum L.) (Huang et al., 1990; Schroeder et
al., 2002), common bean (Phaseolus vulgaris L.) (Huang
et al., 2002), lentil (Lens culinaris Medik.) (Huang et
al., 2003a), chickpea (Cicer arietinum L.) (Huang et al.,
2003a), common wheat (Triticum aestivum L.) (Howe and
Simmonds, 1937; Campbell, 1958; Roberts, 1974; Forster
and Bradbury, 1990) and durum wheat (Triticum durum
Desf.) (McMullen et al., 1984). The pink discoloration of
seed observed in instances of this disease is attributable
to production of pigments called ferrorosamines by the
pathogen (Feistner et al., 1983). Erwinia rhapontici is an
opportunistic pathogen that depends on plant injury for
initiation of infection (Huang et al., 2003b; Huang and
Botanical Studies (2007) 48: 181-186.
*
Corresponding author: E-mail: ericksons@agr.gc.ca; Tel:
1-403-317-3339; Fax: 0021-403-382-3156.
Erickson, 2004).
Pink seed disease has potential negative impacts on
the production and marketability of crops. For example,
Huang and Erickson (2004) reported that planting pink
seeds of pea infected by E. rhapontici resulted in reduc-
tions in seed yield, seed size, seedling emergence, and
seedling vigor. McMullen et al. (1984) reported that when
durum wheat kernels infected with E. rhapontici were
milled, the resulting semolina had a pink discoloration,
and was therefore unsuitable for pasta production.
A 2-year field study on the overwintering of E. rha-
pontici under Canadian prairie conditions showed that the
pathogen survived winters on infected seeds and stems of
pea regardless of burial depth at 0 or 6 cm (Huang and Er-
ickson, 2003), and therefore such infected seeds or stems
can serve as a source of inoculum for infection of crops
in the subsequent growing season. However, no informa-
tion exists to indicate whether strains of E. rhapontici
from one host crop can infect a different host crop. The
increased use of pulse-wheat rotations in North America
in recent years raises concerns regarding the possibility
of transmission of the bacterial pathogen E. rhapontici
from pulse crops to wheat, a major cereal crop in Canada
and the USA. The purpose of this study was to determine
whether or not strains of E. rhapontici from pea, bean, len-
til, chickpea, wheat, canola or soil are host-specific under
controlled conditions in a growth chamber, and under field
conditions.
mICROBIOlOgy
lack of host specificity of strains of Erwinia rhapontici,
causal agent of pink seed of pulse and cereal crops
Hung-Chang HUANG
1
, R. Scott ERICKSON
1,
*, and Ting-Fang HSIEH
2
1
Agriculture and Agri-Food Canada, Lethbridge Research Centre, P.O. Box 3000, Lethbridge, Alberta, T1J 4B1 Canada
2
Floriculture Research Center, Agricultural Research Institute, Gukeng, Yunlin, Taiwan
(Reveived July 7, 2006; Accepted September 25, 2006)
Abstract. Erwinia rhapontici is the causal agent of pink seed and soft rot diseases of several crops. Laboratory
and field experiments were conducted to study the host specificity of strains of E. rhapontici collected from
diseased seeds of pea, bean, lentil, chickpea, wheat, and canola or from infested field soil in western Canada.
For the growth chamber experiments, plants of pea, bean, lentil and chickpea were inoculated with each strain
of E. rhapontici by injection of bacterial suspension (1 ¡Ñ 10
9
cfu/mL) into young pods at 0.1 mL/pod, whereas
developing heads of wheat were injured by abrading with a wire brush and inoculated by spraying of bacterial
suspension at 20 mL/plant. Results showed that the E. rhapontici strains were not host specific, since all of the
strains could infect each of the host crops tested, regardless of the origin of strains. The frequency of infected
seeds was high (>50%) for most strain by crop combinations. Field experiments conducted in 2003 and 2004
revealed that the inoculum of E. rhapontici on infected pea seeds was readily transmitted to neighboring crops
of durum wheat, spring wheat, and common bean, if the crops were injured by abrading with a wire brush
at the early pod formation stage. The impact of the lack of host specificity on management of the pink seed
disease caused by E. rhapontici is discussed.
Keywords: Erwinia rhapontici; Host specificity; Pink seed disease; Strain differentiation.
pg_0002
182
Botanical Studies, Vol. 48, 2007
mATERIAlS AND mETHODS
Twelve strains of the pink seed pathogen, E. rhapontici
collected in western Canada were assessed for pathogenic-
ity on common bean, pea, lentil, chickpea, spring wheat,
and durum wheat, under environmentally controlled con-
ditions. The sources and locations of collection of these
strains are listed in Table 1. Seeds of bean cv. US1140, pea
cv. SS2, chickpea cvs. Myles and Sanford, lentil cv. Laird,
spring wheat cv. Fielder, and durum wheat cv. Kyle, were
planted in Cornell peat-lite mix (Boodley and Sheldrake,
1977) in 15 cm-diameter plastic pots, and were kept in a
growth chamber at 20¢XC/18¢XC; 16-h day/8-h night, until
the plants reached the early pod-filling stage for bean, pea,
lentil and chickpea, or the booting stage for wheat. For the
inoculations on bean, pea, lentil, and chickpea, each strain
was inoculated into 30 pods from 6 plants using the meth-
od of Huang et al. (1990). Each pod was inoculated with
0.1 mL of bacterial suspension (10
9
cfu mL
-1
), by injection
through the mid-rib at the basal end. The same number of
uninoculated and water-inoculated pods served as controls.
Plants were kept in the growth chamber until maturity, and
seeds were harvested and plated onto potato dextrose agar
(PDA) (Difco, Detroit, Michigan, USA) at room tempera-
ture (20 ¡Ó 2¢XC) for 3 days to determine the presence or
absence of E. rhapontici, using the method of Huang et al.
(1990) [i.e., observation of culture characteristics such as
pigment production]. Treatments were arranged in a com-
pletely randomized design. For spring wheat and durum
wheat, similar experiments were conducted, except that the
plants were inoculated by spraying 20 mL/plant of bacte-
rial suspension (10
9
cfu mL
-1
), onto developing heads that
had been injured by lightly stroking with a sterilized wire
brush. For each crop and strain, the frequency of seeds
infected by E. rhapontici was calculated. The experiments
were performed twice for each crop.
Field experiments were conducted during 2003 and
2004 at the Agriculture and Agri-Food Canada Research
Centre, Lethbridge, Alberta, Canada, to determine whether
E. rhapontici on infected pea seeds could be transmitted to
adjacent durum wheat, spring wheat, and common bean.
Pea seeds cv. Delta were obtained from a commercial field
near Vulcan, Alberta, Canada that had high incidence of
pink seed following a hailstorm. Seeds were sorted into
the categories of pink and non-pink, and three sub-samples
of 100 seeds from each category were surface sterilized in
70% ethanol for 90 s, air-dried on paper towel, incubated
on PDA in Petri dishes at 20¢XC for 3 days, and examined
for presence of E. rhapontici by the method described by
Huang et al. (1990). The frequency of E. rhapontici in
samples of non-pink seeds from this field was less than
1%, whereas the frequency in pink seeds was 100%. Seed
samples were stored in a cold room at 4¢XC until used for
the field experiments.
Field plots were established in late May of each year,
in an area of an irrigated field that was fallowed during
the previous growing season. For experiment 1, each plot
consisted of 4 rows of healthy beans (cv. US1140) on the
south side of the plot, 4 rows of peas (cv. Delta, healthy
or infected with E. rhapontici) in the middle of the plot,
and 4 rows of healthy wheat (spring cv. Fielder) on the
north side of the plot. For experiment 2, the plots were
the same, except that the beans were cv. AC Skipper, and
the wheat was durum cv. Kyle. Both experiments were
conducted in both years. For all crops in each experiment,
row length was 5 m and row spacing was 22 cm. Treat-
ments were arranged in a randomized block design with
6 replicates. Plots were maintained until the pea plants
reached the young pod stage (mid-July), and the wheat and
bean plants in each plot were injured by gently abrading
with a sterilized wire brush. Plots with non-injured wheat
and bean plants were used as controls. At maturity (early
Table 1. Source and location of Erwinia rhapontici strains used for the study.
Strain of E. rhapontici
Source (host)
Host cultivar
Location
LRC 8251
Bean
Othello
Bow Island, Alberta
LRC 8252
Bean
US1140
Bow Island, Alberta
LRC 8253
Bean
Viva
Bow Island, Alberta
LRC 8289
Bean
Othello
Carman, Manitoba
LRC 8345
Canola
Hyola 401
Bow Island, Alberta
LRC 8266
Chickpea
Myles
Beechy, Saskatchewan
LRC 8265
Lentil
Laird
Bladworth, Saskatchewan
LRC 733
Pea
Marrowfat
Grassy Lake, Alberta
LRC 965
Pea
Radley
Saskatoon, Saskatchewan
LRC 7954
Pea
Trapper
Lethbridge, Alberta
LRC 1076
Soil
-
Lethbridge, Alberta
LRC 8314
Wheat
Unknown
Ponteix, Saskatchewan
pg_0003
HUANG et al. ¡X Host specificity of
Erwinia rhapontici
183
September) the plots were harvested with a Nurserymaster
Elite 2000 plot combine (Wintersteiger, Ried im Innkreis,
Austria). The frequency of infection of bean, pea, and
wheat seeds by E. rhapontici was determined for each plot
by sorting seed samples into pink and non-pink seeds, and
confirming the accuracy of visual sorting by plating a sub-
sample of 100 seeds from each plot as described previ-
ously.
In each experiment each year, differences between
treatments in frequency of infection of seed by E. rhapon-
tici in each crop were statistically analysed using analysis
of variance (ANOVA), and means were separated using
Duncan¡¦s multiple range tests. SAS/STAT
TM
computer
software, version 8.2, was used for the statistical analyses
(SAS Insitute, 1999).
RESUlTS
In the growth chamber experiments, testing of the
strains of E. rhapontici from bean, canola, chickpea, lentil,
pea, wheat, and soil showed that none of the strains were
host specific, since they could infect all of the tested crops
of bean, chickpea, lentil, and wheat, regardless of the
origin of the strains (Table 2). The frequency of infection
of seeds by E. rhapontici was high (>50%) for most of
the strains on most of the crops. Although none of the
strains was consistently more virulent than the others,
some variation in the susceptibility of crops was observed.
For example, the frequency of infected seeds for chickpea
ranged from 94-100% for cv. Myles, and from 92-100%
for cv. Sanford, compared to a frequency of 28-67% for
wheat cv. Kyle (Table 2). Lentil was also very susceptible
to E. rhapontici, with the frequency of infected seeds
ranging from 88-100%.
Results of the field experiments revealed that E.
rhapontici in naturally infected pea seeds can spread
onto adjacent spring wheat, durum wheat, and common
bean (Table 3). For all crops and years except bean
in 2003, the rate of transmission of E. rhapontici was
significantly (p<0.05) higher for the treatments of plant
injury, compared to the uninjured controls. For example,
the frequency of durum wheat cv. Kyle seeds infected by
E. rhapontici in 2003 was 16% for injured plants grown
adjacent to pink peas (diseased seeds used as a source
of inoculum), 12% for injured plants grown adjacent to
healthy peas, 4% for non-injured plants grown adjacent to
pink peas, and 1% for non-injured plants grown adjacent
to healthy peas (Table 2). Although transmission of E.
rhapontici from infected pea plants to common bean seeds
was not observed in the 2003 field experiment, it was
observed in the 2004 field experiment.
DISCUSSION
This study demonstrates for the first time that the strains
of E. rhapontici from pea, bean, chickpea, lentil, canola,
Table 2. Infection of cereal and pulse crops by strains of Erwinia rhapontici (growth chamber experiments).
Strain of E. rhapontici (source)
Frequency of seed infection (%)
Bean cv.
US1140
1
Chickpea
cv. Myles
1
Chickpea
cv. Sanford
1
Lentil cv.
Laird
1
Pea cv.
SS2
1
Wheat cv.
Fielder
2
Wheat cv.
Kyle
2
LRC 8251 (bean)
60
100
100
93
62
59
51
LRC 8252 (bean)
69
96
100
96
54
51
46
LRC 8253 (bean)
63
100
92
88
43
53
42
LRC 8289 (bean)
62
100
100
100
65
58
28
LRC 8345 (canola)
73
97
97
89
54
68
47
LRC 8266 (chickpea)
61
100
100
94
70
69
67
LRC 8265 (lentil)
66
97
100
94
45
76
50
LRC 733 (pea)
58
100
97
100
61
52
44
LRC 965 (pea)
70
94
100
93
64
60
47
LRC 7954 (pea)
61
100
100
100
54
69
43
LRC 1076 (soil)
58
100
94
91
43
68
38
LRC 8314 (wheat)
56
100
96
95
69
66
63
1
Isolates were inoculated into 30 pods by injection of 0.1 mL pod
-1
of bacterial suspension, 10
9
cfu mL
-1
.
2
Isolates were inoculated onto 30 heads by spray of bacterial suspension, 10
9
cfu mL
-1
, to the runoff point.
pg_0004
184
Botanical Studies, Vol. 48, 2007
wheat and soil, are not host-specific, and are capable of
infecting a wide range of crop plants, including great
northern bean cv. US1140, desi chickpea cv. Myles, kabuli
chickpea cv. Sanford, lentil cv. Laird, pea cv. SS2, spring
wheat cv. Fielder, and durum wheat cv. Kyle, regardless of
the host origin of the pathogen. The lack of host specifi-
city among strains of E. rhapontici is further proven in the
field experiments, as the pathogen on naturally infected
pea seeds can readily spread onto adjacent plants of great
northern bean, spring wheat and durum wheat in the field.
The indoor and field studies further confirm previous
reports that plant injury at the seed formation stage is
critical for E. rhapontici to gain entrance into host plants
and cause formation of pink seeds (Huang et al., 2003b;
Huang and Erickson, 2004). This strongly suggests that
E. rhapontici is an opportunistic pathogen, and that
its dissemination may be limited by the occurrence of
circumstances that cause plant injury. Wounding of plant
tissue due to wind or hail damage, or insect damage, likely
provides opportunities for E. rhapontici to infect and
spread in the field. Other factors such as temperature or
humidity within the crop canopy could also be important
in dissemination of E. rhapontici in the field. The
observation that infection can occur on injured plants that
are adjacent to infected plants, suggests that movement
of the pathogen occurs within a certain area. The actual
mechanism of dispersal remains to be determined, but may
be related to movement of insects or splashing of droplets
during irrigation.
A previous study indicates that E. rhapontici can
survive Canadian winters in infected seeds and crop debris
(Huang and Erickson, 2003). The lack of host specificity
observed in this study suggests that the inoculum of E.
rhapontici from infected pea seeds or stems can serve as
a potential source of inoculum for other crops including
wheat. Since pulse crops such as peas are important crops
for rotation with wheat on the North American prairies
(Biederbeck et al., 1999), the transmission of E. rhapontici
from infected pea seeds to wheat crop observed in this
study raises new concerns about the appropriateness of
using peas in a rotation sequence with wheat in areas
where the pink seed pathogen is prevalent.
Huang and Erickson (2004) reported that planting pea
seeds infected by E. rhapontici can have serious impact,
including losses in stand establishment, seedling vigor,
seed yield, and seed quality. In addition, a study on
durum wheat infected with E. rhapontici showed that the
milled wheat kernels resulted in pink semolina that was
unsuitable for producing pasta (McMullen et al., 1984).
Since the cereal and pulse crops used in this study are
Table 3. Spread of Erwinia rhapontici from infected pea to bean and wheat (field experiments, 2003-2004).
Experiment 1
Frequency of seed infection (%)
Treatment
Beans cv. US1140
Peas cv. Delta
Wheat cv. Fielder
2003 2004
2003 2004
2003 2004
Injured
1
, pink
2
0 a
3
8 a
17 a
15 a
14 a
9 a
Non-injured, pink
0 a
0 b
4 b
3 b
3 b
1 b
Injured, non-pink
0 a
5 a
15 a
18 a
9 a
10 a
Non-injured, non-pink
0 a
0 b
3 b
2 b
2 b
2 b
Standard error
0.0
1.1
1.3
1.8
1.4
1.5
Experiment 2
Frequency of seed infection (%)
Treatment
Beans cv. AC Skipper
Peas cv. Delta
Wheat cv. Kyle
2003 2004
2003 2004
2003 2004
Injured
1
, pink
2
0 a
3
4 a
14 a
18 a
16 a
14 a
Non-injured, pink
0 a
1 b
1 b
2 b
4 b
2 b
Injured, non-pink
0 a
6 a
14 a
12 a
12 a
11 a
Non-injured, non-pink
0 a
0 b
2 b
1 b
1 b
3 b
Standard error
0.0
0.8
1.6
2.1
1.0
1.3
1
Plants were injured by gently abrading with a wire brush at the young pod stage.
2
Naturally infected (pink) pea seeds were obtained from a commercial field and used for the study.
3
Means within each column followed by the same letter are not significantly different (Duncan¡¦s multiple range test; p>0.05).
pg_0005
HUANG et al. ¡X Host specificity of
Erwinia rhapontici
185
used worldwide as sources of food and food products for
people and animals, and since the effect of consumption
of seeds infected by E. rhapontici on the health of humans
and livestock is still largely unknown, further research is
needed on the quality and safety of crops infected by pink
seed disease.
Acknowledgements. This study was supported by funding
from the Canadian Seed Growers¡¦ Association and the
Agriculture and Agri-Food Canada Matching Investment
Initiative Fund, project number A03898. Dr. T. F. Hsieh
was a visiting scientist at Lethbridge Research Centre
under NSERC Postdoctoral Fellowship during 2003.
The technical assistance of L. Hermann is gratefully
acknowledged. This is LRC contribution number
38706035.
lITERATURE CITED
Biederbeck, V.O., N.Z. Lupwayi, W.A. Rice, K.G. Hanson, and
R.P. Zentner. 1999. Crop rotation effects on soil microbial
populations, biomass and diversity under wheat in a brown
loam. In Proceedings of Soils and Crops, February 25-26,
1999. Saskatoon, Saskatchewan, University of Saskatch-
ewan.
Boodley, J.W. and R. Sheldrake. 1977. Cornell Peat-Lite Mixes
for commercial plant growing. New York State College of
Agricultural and Life Science, Information Bulletin 43, 8
pp.
Campbell, W.P. 1958. A cause of pink seeds in wheat. Plant Dis.
Reptr. 42: 1272.
Feistner, G., H. Korth, H. Ko, G. Pulverer, and H. Budzikiewicz.
1983. Ferrorosamine A from Erwinia rhapontici. Current
Microbiol. 8: 239-243.
Forster, R.L. and J.F. Bradbury. 1990. Pink seed of wheat caused
by Erwinia rhapontici in Idaho. Plant Dis. 74: 81.
Goto, M. and K. Matsumoto. 1986. Taxonomic study on soft
rot bacteria isolated from diseased rhizomes and roots of
wasabi (Eutrema wasabi Maxim.). Ann. Phytopathol. Soc.
Japan 52: 69-77.
Howe, E.T. and P.M. Simmonds. 1937. Bacterial pink blotch of
wheat. In I.L. Conners, J.H. Graigie, and T.C. Vanterpool
(eds.), Proceedings of the Canadian Phytopathological Soci-
ety. Seventh Session. June 28-30, 1937, Ottawa, ON.
Huang, H.C. and R.S. Erickson. 2003. Overwintering of Erwinia
rhapontici, causal agent of pink seed of pea, on the Cana-
dian prairies. Plant Pathol. Bull. 12: 133-136.
Huang, H.C. and R.S. Erickson. 2004. Impact of pink seed of
pea caused by Erwinia rhapontici in Canada. Plant Pathol.
Bull. 13: 261-266.
Huang, H.C., R.S. Erickson, L.J. Yanke, T.F. Hsieh, and R.A.A.
Morrall. 2003a. First report of pink seed of lentil and chick-
pea caused by Erwinia rhapontici in Canada. Plant Dis. 87:
1398.
Huang, H.C., R.S. Ericks on, L.J. Yanke, and H.-H. Mundel.
2002. First report of pink seed of common bean caused by
Erwinia rhapontici. Plant Dis. 86: 921.
Huang, H.C., T.F. Hsieh, and R.S. Erickson. 2003b. Biology and
epidemiology of Erwinia rhapontici, causal agent of pink
seed and crown rot of plants. Plant Pathol. Bull. 12: 69-76.
Huang, H.C., R.C. Phillippe, and L.M. Phillippe. 1990. Pink
seed of pea: A new disease caused by Erwinia rhapontici.
Can. J. Plant Pathol. 12: 445-448.
Letal, J.R. 1976. Crown rot of rhubarb in Alberta. Can. Plant
Dis. Surv. 56: 67-68.
McMullen, M.P., R.W. Stack, J.D. Miller, M.C. Bromel, and V.L.
Youngs. 1984. Erwinia rhapontici, a bacterium causing pink
wheat kernels. Proc. North Dakota Academy S ci. Grand
Forks, ND, 38: 78.
Metcalfe, G. 1940. Bacterium rhaponticum (Millard) Dowson, A
cause of crown-rot disease of rhubarb. Ann. Appl. Biol. 27:
502-508.
Millard, W.A. 1924. Crown rot of rhubarb. Bull. Univ. Leeds.
138: 28.
Ohuchi, A., T. Ohsawa, and J. Nishimura. 1983. Two pathogenic
bacteria, Erwinia rhapontici (Millard 1924) Burkholder
1948 and Pseudomonas marginalis pv. marginalis (Brown
1918) Stevens 1925, causing a soft rot of onion. Ann. Phy-
topathol. Soc. Japan 49: 619-626.
Roberts, P. 1974. Erwinia rhapontici (Millard) Burkholder as-
sociated with pink grain of wheat. J. Appl. Bacteriol. 37:
353-358.
SAS Insitute. 1999. Statistical Analysis Software, version 8.2.
Cary, NC.
Schroeder, B.K., S.L. Lupien, and F.M. Dugan. 2002. First re-
port of pink seed of pea caused by Erwinia rhapontici in the
United States. Plant Dis. 86: 188.
pg_0006
186
Botanical Studies, Vol. 48, 2007