V m
x M N"
C “.C.“ 21"‘31 gay . _ . CC . F55”: Wm}. M.
1" t ‘ . ' , . .1. ' ,3. C._C .1‘” u: v
w «W x 7 1;. “4.2mm: f ‘ :‘uw 3%} 3“
my; d"‘~ fiifi‘} l J ’31- 335:” \n ..
‘ $.}'3'C5";C m :. afi" Cc;
“C C . ,. L": ‘6 .C..“ “$2.4.
V QUEE‘W , C". “39:“. aim; ,xdi’guigC
1’17"?"“1rfu «a #3?” 5133112.;
' "M v V “‘1‘.
g z' n. - ' fl 7 fig?
3. " ‘\ 11,-.“ y 33% '
C y fl“ . C . . m
.‘ “7 " 53"“ '1 ‘ ’31- ‘fiV'fif‘ H= (Li-3 ' “It?“ a: "3'; vii:
, _ ,, .C.. _ . , C.- .u x é“? .C
‘ 3.615"?- ,”““‘x*.,. 3:. C” -‘J-' , “795 1-4?“ . u‘é 535: .- . iff-22531§$ ‘
: A...~.‘C’“"‘{”31 "" '7‘ 1‘" ~ ’1‘ ‘ 7' ‘ -~ ' 42.1.2. .. 'P;
C C . C - < ,. .. .. _ ‘ gem.‘-::~. .ww Ho'vv *7'
J.“ C C ' «Jena; M «g... :x‘gvgu
. ”x . . 33:3. ‘ “ 1‘ w ,. ':.“ . 7*..12‘“
A ECL‘E‘Z‘I‘QCQ. :31». y‘“"' ’ “ C ', .3513? '- W . ‘7 "a;
, Camgnfia ' '
‘7‘” \'?~‘,:.’¥3~”g ””3323; ..3.'**
. 'Luxk- I.» . C ’ 1' "V” *3 ,‘ n
' ' " ‘ w "‘ {4a m “1‘? ?:“" "19:55“! v 77::- Ci: «:3: xl‘wlxpu'c‘uv U
“ ' ’ qr" "."M' ,ILE‘ \f :anéi'wt: 513‘?“ -
"ii-k 3W5 "§2¥xgE-E~Z-.‘:‘ ”ZR
‘ ‘ “ 811% Lu-
1 5%; 3‘3}:
} girxfu'f
.
L».
'. “ {T’s-‘1‘ 1..
315%.?"
7. u Neda-4:”
. .- “4r”?
., .
1 3M;- 3*" U.“ .{ti‘fi‘a ...~.
. W" 1
,C“ Jha J C
3 unh
; .mr “7
'w'
’f
'i
‘
‘2; ,, x.- etfig' $5,:
. EU 3 3"-
'l
1
h
I?
3"
.2 2: ~
2. ’.
) 7.‘
':. '... C
4 '9'“
i;
1
1
‘y ‘
§
V
‘f
I
.
'.
. Mg; 1
P i J
‘V‘ .
I ' ’7
’v
1‘
.I
f
I
,
..'
a
‘

 

 

lllllltllllllllllllllllllllllllllllllllllllhllllllllll

3 1293 0087648

This is to certify that the
dissertation entitled

Ecology of Pseudomonas syringag pv. phaseolicola
in northern Tanzania

 

 

presented by

Robert B. Mabagala

has been accepted towards fulfillment .
of the requirements for

Ph.D. degreein Botany & Plant Path.

 

(Mg/,AiJ (/M/ 124% )k

Major professor

Date [fl/Jj/79/

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove thie checkout from your record.
TO AVOID FINES return on or before due due.

  

DATEDUE DATE DUE DATE DUE

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

H— i

ll
W I!

MSU In An Affirmative ActioniEquel Opportunity Institution
chS—p.‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ECOLOGY OF PSEUDOMONAS SYRINGAE PV.
PHASEOLICOLA IN NORTHERN TANZANIA

By

Robert B. Mabagala

. A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Botany and Plant Pathology

1991

, c.
I

53 (1;; .

." ' s
v v

i

Z5 -— 9

ABSTRACT

ECOLOGY OF PSEUDOMONAS SYRINGAE PV.
PHASEOLICOLA IN NORTHERN TANZANIA

By
Robert B. Mabagala

This study was undertaken to investigate the ecology of Pseudomonas m pv.
phaseolicola, the causal agent of halo blight of beans, in the Arusha and Kilimanjaro
regions, northern Tanzania. A total of 118 isolates were collected from November 1988 to
February 1990, and examined for pathogenic variability using four differential bean cultivars.
Three races were found; race 2 occurred at a higher frequency (52.5%) than race 1 (44.9%)
and race 3 (2.5%). Some isolates of race 2, designated race 2P, produced a brown diffusible
pigment on agar media. Some race 2P isolates grew at a higher temperature (34 i 2°C)
than the non-pigment-producing race 2 isolates or race 1 and 3 isolates. Race 2 isolates
obtained from bean debris in Monduli, Arusha region, were more virulent than those
isolated from growing bean plants. They induced systemic chlorosis and stunting on the
usually resistant differential bean cultivar, Edmund.

The ability of fig. pv. phaseolicola to survive in bean debris and in dead standing
bean plants varied depending on race, geographical location, depth of placement in the soil
and the bean genotype used. Volunteer bean plants also provided an alternative survival
site for halo blight bacteria. Of the 17 weed species belonging to 10 families, only
Neonotonia mgi_i_ti_i served as a perennial reservoir of 2s. pv. phaseolicola. The weed
survived the dry periods on fences, hedge rows, by roadsides, on ditch banks near bean
fields and in comers of farmers’ fields. Only race 1 of the bacterium was recovered from

1:1. wightii. However, under artificial inoculation the weed was susceptible to all three races.

Since race 1 of 13.5. pv. phaseolicola was not seed-borne in 1:1, mm this precludes the
weed spreading the halo blight pathogen into new areas through seed.

Intercropping beans with maize favored multiplication of halo blight bacteria on and
in bean leaves, and resulted in more severe disease on pods and in a higher percentage of
seed infection, than occurred in beans grown in pure stands. Maize leaves did not support
high populations of halo blight bacteria, and thus do not seem to provide additional

inoculum for the bean crop when the two are grown in association.

In memory of my advisor Dr. A.W. Saettler and my uncle, Muhindi Musiba.

ACKNOWLEDGEMENTS

I express my sincere gratefulness to Dr. 11.. Lockwood for his encouragement and
guidance throughout the writing phase of this study.

My sincere appreciations are also extended to other members of my committee, Dr.
Jim Kelly, Dr. C. Stephens, and Dr. P. Hart for their suggestions and criticisms during the
course of the study.

I am also grateful to Dr. MJ. Silbemagel, Mr. D.L. Kessy, Ms. B. Gondwe, Dr. DJ.
Allen, Dr. J.D. 'Iaylor, Prof. J.M. Teri, and Joseph Matata for support and encouragement
during the course of this study. Moral support given by my brother, Augustine and his
family during the research phase is highly acknowledged.

Financial support provided by Washington State University/Tanzania Bean/Cowpea
CRSP Program and the Rockefeller Foundation African Dissertation Internship Award is

gratefully acknowledged.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

List of Tables vii
List of Figures ix
CHAPTER 1

GENERAL INTRODUCTION 1

LITERATURE REVIEW 2
Occurrence and severity 3
Synonyms of the pathogen 4
Morphological and physiological characteristics 4
Host range 5
Survival 5
Epiphytic phase 7
Penetration and population dynamics 7
Dissemination 8
Symptoms 8
Pathogenic variation 10
Synergism 13
Cultural control 14
Chemical control 14
Plant resistance— 15

LITERATURE CITED 17

CHAPTER 2

PATHOGENIC VARIATION OF PSEUDOMONAS W PV.

PHASEOLICOLA IN NORTHERN TANZANIA 24

INTRODUCTION 25

MATERIALS AND METHODS 26
Sampling and isolation 26
Bacteriophage tests 28
Serology 29
Reference strains '29
Storage of isolates 79
Soil sterilization 30
Growing of plants 30
Inoculum preparation 30

 

Race identification

iv

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Temperature studies 31
Survival studies 32
Soil pH measurements 33
Volunteer bean plants 34
Weather data 34
RESULTS 34
Prevailing races 34
Tbmperature studies 35
Survival studies 42
pH values 44
Volunteer plants 45
Weather data 45
DISCUSSION 50
LITERATURE CITED 54
CHAPTER 3
THE ROLE OF WEEDS IN SURVIVAL OF PSEUDOMONAS SYRINGAE
PVI PHASEOLICOLA IN NORTHERN TANZANIA 57
INTRODUCTION 58
MATERIALS AND METHODS 60
Sampling and isolation procedures 60
Storage of isolates 61
Sterilization and grinding of samples 62
Identification of bacterial isolates 62
Pathogenicity tests 64
Ice nucleation activity 64
Carbon source utilization 65
Bacteriophage tests 65
Serology 66
Race identification 66
Assay of Neonotonia wightii reproductive parts for Lg. pv. phaseolicola
infection 66
Neonotonia wightii seed infection and transmission assay 67
Reaction of Neonotonia wightii to races 1, 2 and 3 of
24;. pv. phaseolicola 68
Insects as pests of beans and Neonotonia wightii 70
RESULTS 71
Identification of bacterial isolates 76
Pathogenicity tests 76
Race identification 77

 

Assay of N. wightii reproductive parts for infection by Ls. pv. phaseolicola ......... 77

N. wightii seed infection and transmission assay
Reaction of _N. wightii to races 1, 2 and 3 of 2s. pv. phaseolicola
Susceptibility of three bean cultivars to insect pests of N. wightii

DISCUSSION

79
81
81

84

 

LITERATURE CITED

 

CHAPTER4

88

POPULATION DYNAMICS OF PSEUDOMONAS SYRINGAE PVI PHASEOLICOLA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IN TWO BEAN CROPPING SYSTEMS IN NORTHERN TANZANIA 91
INTRODUCTION 92
MATERIALS AND METHODS 95
Location 95
Experimental design 95
Bacterial isolates 96
Inoculation of plants 96
Sampling of leaves 97
Bacterial population trends 97
Evaluation of disease reaction 98
Leaf wetness studies 98
Bean seed infection assay 98
Weather data 99
Data analysis 99
RESULTS 99
Population dynamies at Lambn 99
Population dynamics at Lyarnnngn 100
Populations of Ls. pv. phaseolicola on and in maize 105
Leaf and pod disease ratings 110
Retention of moisture on leaves 112
Climatological data 112
Bean seed infection 118
DISCUSSION .118
LITERATURE CITED 123

 

TABLE

LIST OF TABLES

CHAPTERZ

Distribution of races of Pseudomonas m’ngae pv.
phaseolicola in Arusha and Kilimanjaro regions, northern
Tanzania

Influence of temperature on growth of representative

PAGE

39

isolates of races of Pseudomonas m pv. phaseolicola ....... 41

Survival of Pseudomonas sm'ngae pv. phaseolicola

races 1 and 2 in infected bean debris buried for 6 months
at two depths under field conditions at Ngarash, Monduli,
Arusha region and at Lyamungu, Kilimanjaro region
northern Tanzania

 

CHAPTER3

Weed families and species assayed for presence of
Pseudomonas sm'ngae pv. phaseolicola in northern
Tanzania

Race identification of Pseudomonas m’ngae pv.
phaseolicola strains isolated from Neonotonia wightii
in northern Tanzania ~

Percentage of Neonotonia wigl_1tii seed infected with
Pseudomonas m’ngae pv. phaseolicola and other
bacteria in northern Tanzania

Reaction of Neonotonia wightii to races 1, 2 and 3 of
Pseudomonas gig'ngae pv. phaseolicola from northern
Tanzania

Susceptibility of seed of three bean cultivars grown in

43

72

78

82

northern Tanzania, and _N. wightii seed to bruchid weevils ....... 83

TABLE

CHAPTER 4 PAGE

Halo blight disease severity ratings for foliage and pods of
Canadian Wonder bean in two bean cropping systems in
northern Tanzania 111

Time required for visible moisture on bean and maize leaves
to dry following cessation of rain at Lyamungu, Kilimanjaro
region, northern Tanzania 113

Percentage of Canadian Wonder bean seed infected with
Pseudomonas m’ngae pv. phaseolicola in two bean cropping
systems in northern Tanzania 119

FIGURE

LIST OF FIGURES

CHAPTER 2 PAGE

Map of Tanzania showing location of Arusha and
Kilimanjaro regions 27

 

Origin of Pseudomonas _svg'ngae pv. phaseolicola isolates

in northern Tanzania. 1 = Lambo, 2 = Lyamungu,

3 = Kilema, 4 = Milwaleni, 5 = Sanya J uu, 6 = Thugeru,

7 = Selian, 8 = Monduli 36

 

Relative frequency of races 1, 2 and 3 of Pseudomonas

sm'ngae pv. phaseolicola in collections made in Arusha

and Kilimanjaro regions, northern Tanzania from November,
1988 to February, 1990. Frequencies are expressed as
percentages of the total number of isolates collected.

2P represents race 2 isolates producing brown diffusible
pigment 38

 

Stunting and systemic chlorosis symptoms produced by

race 2 isolates obtained from bean debris in Monduli,

Arusha region northern Tanzania on a resistant

differential bean cultivar, Edmund 40

Monthly rainfall from June to December at Monduli for
1989 and mean monthly rainfall for 3 years (1986-1988).
Numbers above each bar represent number of rain days
during the indicated month 46

Monthly rainfall and mean monthly maximum and minimum
temperatures from June to December at Lyamungu for

1989 and for 53 years (1935-1988). Numbers aboveeach

bar represent numbers of rain days during the indicated

month 48

FIGURE

CHAPTER 3 PAGE

Typical halo blight symptoms on Neonotonia wightii
leaves 74

 

Halo blight-infected Neonotonia wightii plants surviving
the dry season on a fence at Lyamungu Agricultural
Research Station, Tanzania 75

 

CHAPTER4

Surface and internal population dynamies of Pseudomonas
m pv. phaseolicola race 1 on and in bean foliage

in two bean cropping systems at Lambo, Kilimanjaro region,
northern Tanzania. Bars indicate standard errors of the

means 101

 

Surface and internal population dynamics of Pseudomonas
m’ngae pv. phaseolicola race 1 on and in bean foliage

in two bean cropping systems at Lyamungu, Kilimanjaro

region, northern Tanzania. Bars indicate standard errors

of the means 103

 

Surface and internal population dynamies of Pseudomonas
_syg’ggag pv. phaseolicola race 1 on and in maize foliage

in monocrop and intercrop systems at Lambo, Kilimanjaro
region, northern Tanzania. Bars indicate standard errors

of the means 106

 

FIGURE

CHAPTER 4 PAGE

Surface and internal population dynamics of Pseudomonas
gm’ngae pv. phaseolicola race 1 on and in maize foliage

in monocrop and intercrop systems at Lyamungu, Kilimanjaro
region, northern Tanzania. Bars indicate standard errors

of the means 108

 

Monthly rainfall and mean monthly maximum and minimum
temperatures from March to July at Lambo for 1989 and
for 53 years (1935-1988). Numbers above each bar represent

number of rain days during the indicated month 114

 

Monthly rainfall and mean monthly maximum and minimum
temperatures from March to July at Lyamungu for 1989

and for 53 years (1935-1988). Numbers above each bar

represent number of rain days during the indicated month ....... 116

Chapter 1

GENERAL INTRODUCTION

Beans (Phaseolus m L.) are an important source of dietary protein in many
parts of the world, including Tanzania. Production occurs in a wide range of cropping
systems and environments (Allen gt a_l., 1989; CIAT; 1986). In Tanzania, beans are mainly
grown in the highlands of northern, southern and western parts of the country where
rainfall is adequate for bean production. However, yields have remained relatively low with
an average yield of 600 kg/ha in farmers’ fields (CRSP, 1990). Like many other tropical
bean production regions, diseases, insect pests, and low fertility are the most important
production constraints in the country.

Halo blight incited by Pseudomonas m pv. phaseolicola (Burk) Young, Dye
and Mlkie is one of the most important diseases of beans in Tanzania. The disease is very
prevalent in high, cool, wet areas where beans are mainly grown, and losses due to halo
blight can be severe (Silbemagel and Mills, 1987). Because of the widespread occurrence
and the severity of halo blight in areas where the disease occurs in the country, it is
assumed that the disease causes considerable yield reduction. However, the precise extent
of losses has not been well established. Recently, in field trials conducted at Lyamungu,
northern Tanzania, yield losses of up to 42% in susceptible cultivars were reported (CIAT;

1989). Breeding for halo blight resistance is generally considered the best method of

1

2

controlling the disease in small holder production systems in 'Ihnzania, and efforts are being
made to produce halo blight resistant varieties (Thri e_t 91., 1990). However, the
development of resistant varieties must take into account variability of both the pathogen
and the genetic resistance in the host. In addition, successful halo blight control measures
require a thorough understanding of the ecology of halo blight bacteria in the country.

While much information on Ls. pv. phaseolicola is readily available in other
countries, little is known about the ecology of this bacterial pathogen in Tanzania. The
objective of this dissertation was, therefore, to examine the ecology of fig. pv. phaseolicola
in northern Tanzania. The study was divided into three parts: pathogenic variation and
survival in the soil of halo blight bacteria in northern Tanzania are covered in Chapter 2.
Chapter 3 examines the role of weeds in survival of halo blight bacteria. In northern
Tanzania, as in many other parts of the country, the majority of the bean crop is grown
in association with maize and other crops. Chapter 4 is, therefore, devoted to population
dynamics of fig. pv. phaseolicola in beans when grown in pure stands and in association

with maize.

LITERATURE REVIEW
Halo blight of beans is caused by the bacterium Pseudomonas m pv.
phaseolicola (Burk) Young, Dye & Wilkie (Young 9; a_l., 1978). The disease was first
reported and described from the state of New York by Burkholder in 1926. Since then,
the disease has been reported in many countries of Central and South America, U.S.A.,
Africa, Europe and Australia (Buruchara, 1983; Johnson, 1969; Msuku, 1984; Schwartz,

1980; Schwartz and Pastor-Corrales, 1989). In Thnzania and other tropical countries, halo

3
blight is more prevalent in high, cool, wet areas where beans are produced. In such areas,
the disease can cause severe crop losses (CIA‘I; 1981; Karel e_t _al., 1981; Silbemagel and

Mills, 1987).

Occurrence and severity

Yield losses of 23-43 percent have occurred in research fields in Michigan (Saettler
and Potter, 1970), and the disease has caused serious losses in Colorado (Schwartz and
Legard, 1986). In eastern Africa, the disease can be of economic important in Malawi and
Kenya, but rarely in Uganda, except at the highest altitudes of cultivation (Allen, 1983).
Halo blight disease has generally been considered of minor importance in Tanzania,
especially in some areas of the Kilimanjaro region. However, in recent years severe
outbreaks of the disease have been reported in some parts of this region and in other
regions such as Arusha, Mbeya and West Lake. These outbreaks have been suggested to
be due to the occurrence of a new virulent race, probably race 3 of the pathogen (Gondwe,
1987).

Although it is well known that plant diseases constitute a major constraint to bean
production in Tanzania and elsewhere in Africa, there are very few reliable data quantifying
crop losses. In addition, it is difficult to extrapolate yield losses made under experimental
conditions to those likely to be incurred in agricultural practice, where crop yields are
limited by a complex of stress factors, of which diseases are but a part (Allen, 1983).
Research geared toward estimating yield losses due to halo blight and other bacterial
diseases in Tanzania is generally lacking. Such a situation has resulted in underestimating

yield losses caused by halo blight disease.

Smonms of the pathogen

Other names of Pseudomonas m pv. phaseolicola found in literature include:
Phflomonas medicaginis (Sackett) Bergey gt a_l., var. phaseolicola Burkholder; Bacterium
medicaginis (Sackett) E.F. Smith var. phaseolicola (Burkholder) Link & Hull; Pseudomonas
medicaginis Sackett var. phaseolicola (Burkholder) Stapp and Kotte; Bacterium puerariae
Hedges; and Pseudomonas medicaginis Sackett f.sp. phaseolicola (Burkholder) Dowson

(Bradbury, 1986; CMI, 1965).

Morphological and phfiiological characteristics

Pseudomonas Mag pv. phaseolicola exhibits very distinct morphological,
physiological and nutritional characteristics. Cells appear as single straight rods (Sands e1
a_l., 1970), and are motile with peritrichous flagella. Cells are gram-negative, strictly aerobic
and do not require growth factors. Poly-B-hydroxybutyrate is not accumulated as an
intracellular carbon reserve. In artificial media, especially those deficient in iron, cultures
produce a diffusible fluorescent pigment. Arginine dihydrolase activity is absent (Doudoroff
and Pallerozin, 1974), and it is oxidase negative (Kovacs, 1956).

The bacteria are capable of utilizing D-gluconate, L (+) arabinose, sucrose,
succinate, DL-B-hydroxybutyrate, transconitate, L-serine, L-alanine and p-hydroxy benzoate,
but glutarate, mesotartrate, DL-glycerate, iso-ascorbate, erythritol, sorbitol, meso-inositol and
N-caproate are not utilized. The optimum temperature for growth is 20-23°C. White to
creamy colonies with a bluish hue are produced which may be either smooth or rough
(Adam and Pugsley, 1934; Zaumeyer and Thomas, 1957).

It has been observed that the rough and smooth colony forms of _P_.§_. pv.

phaseolicola are different from one another serologically and in their reaction to

5
bacteriophages. The rough forms were later shown to produce less extracellular
polysaccharides and to be less virulent than the smooth forms (Carey and Starr, 1957).
Extracellular polysaccharides produced by Psp have been associated with pathogenicity on

susceptible bean leaves (Buruchara, 1983; Epton e_t g, 1977).

Host range

Halo blight bacteria can infect various plant species. Moffett (1983), Schwartz,
(1989) and Zaumeyer and Thomas (1957) included the following as hosts: common bean

(Phaseolus vulgaris L), lima bean (P. lunatus L), tepary bean (2. acutifolius A. Gray var.

 

acutifolius), Marcrogtilium bracteatum (Nees ex Mart) Maréchal et Baudet, scarlet runner

bean (E. coccineus L), B. pglyanthus Greenman, P. must—arm (L), pigeonpea (Cajanus

cajan (L) Millsp.), hyacinth bean (Lablab purpureus (L) Sweet), soybean (Glycine 1.7133 (L)

 

Merrill), Vigga angularis (Willd.) Ohwi et Ohasi, mung bean (X. radiata (L) Wilczek var.

 

radiata), Pueraria lobata (Mlld) Ohwi, glycine (Neonotonia wightii (Arn.) Lackey) and

 

siratro (Macroptilium atropugpureum (DC) Urb. (CIAT; 1987).

M

Knowledge as to how a pathogen survives and is disseminated is very important for
developing successful control strategies. Halo blight bacterium survives in infected seeds
and plant residue on the soil surface, and has been found on volunteer beans in the field
early in the growing season (Legard and Schwartz, 1987; Schuster and Coyne, 1975b,
Schwartz, 1989). The bacterium survives in these habitats until environmental conditions
become favorable for infection. Ls. pv. phaseolicola has also been reported to survive for
nine months after passage through sheep which consumed infested plant debris (Starr and

Kercher, 1969).

6

It has been observed that survival in bean plant debris buried under dry condition
is often longer than when soil is saturated (Allen, 1983). Buddenhagen (1965) suggested
that bacterial pathogens evolving away from the requirements of soil saprophytic existence
have tended to develop sustained plant to plant infection cycles often through insect
transmission. Coupled with survival in seed, they have thus lost the more complex
requirements of survival in the soil. Thus, the halo blight pathogen has essentially been
liberated from the requirements of soil phase through seed transmission or through
association with a perennial host (Allen, 1983).

Schuster and Coyne (1975b) observed that the more virulent strains of fig. pv.
phaseolicola were better adapted for survival than the less virulent strains. This may be due
to the fact that more virulent strains tend to multiply to higher numbers in the host than
the less virulent strains. I_n m P_. pv. phaseolicola cells survived in liquid nitrogen at -
172°C for 30 months or on silica gel at -20°C for 60 months (Leben and Sleesman, 1982;
Moore and Carlson, 1975).

The bacterium has a tremendous potential to cause disease. As few as 0.1 percent
of halo blight infected seed can, under conducive environmental conditions, lead to severe
crop losses in susceptible bean cultivars (Patel 91 _a_l., 1964, Schwartz, 1980, 1989). Under
Wisconsin conditions, as few as 12 infected seeds per hectare may lead to an epidemic of
halo blight disease (Walker and Patel, 1964). Zero tolerance of infected seed appears to
be a necessity where rates of disease spread are likely to be high. Under British conditions
a tolerance level of 1 infected seed per 5 kg has been estimated as the maximum allowable
for the halo blight bacterium (Taylor e_t £11., 1979). In tests of disease transmission from
seed to seedlings, it has been found that heavily infected seeds usually failed to produce
seedlings. Most infections developed from seeds with slight or no symptoms at all (Allen,

1983).

7
Since the halo blight bacteria survive and are transmissible by seed, in Tanzania it
is more severe for farmers who do not purchase high quality seed but rather maintain a
portion of their harvest for the next planting (Gondwe, 1987). Programs to produce

pathogen-free seed by growing the crop under furrow irrigation are lacking.

Epiphflic phase of Rs. pv. phgseolicolg
The epiphytic phase of fig. pv. phaseolicola was first reported by Ercolani e_t £11.,
(1974). Their studies indicated that halo blight bacteria multiplied to some extent on hairy

vetch (Vicia villosa Roth, Leguminoseae) near bean fields, especially in field corners under

 

 

Wisconsin conditions. Using a rifampicin-resistant mutant of 2.5. pv. phaseolicola, Stadt and
Saettler (1981) observed that the pathogen was capable of establishing an epiphytic phase
on bean leaves in Michigan. Under Colorado conditions, Legard and Schwartz (1987)
demonstrated that gs. pv. phaseolicola can occur as an epiphyte on dry beans. They
further observed that the bacterium became a predominant epiphyte late in the season

when bean plants matured.

Penetration and pppulation dypamics

Ls. pv. phaseolicola enters plants through wounds or stomata during periods of high
relative humidity or free moisture (Saettler and Potter, 1970;, Zaumeyer and Thomas, 1957).
Maino (1972) reported that the bacterium produces hemicellulases which degrade host cell
wall material during pathogenesis. The pathogen multiplies epiphytically on blossoms, pods
and stem internodes under experimental conditions. 24;». pv. phaseolicola also multiplies
rapidly on or near the surface of foliage with or without lesions in the presence of dew

(Saettler and Potter, 1970; Stadt and Saettler, 1981).

We

Dissemination of fig. pv. phaseolicola between plants occurs by water splash and
winds during periods of rainfall. Insects, animals and people moving through the crop when
the foliage is wet can also spread the pathogen (Walker and Patel, 1964). Pod infection
with halo blight bacteria, especially of the dorsal vascular tissues can lead to internal seed
contamination which is the primary means of dissemination. Contaminated seed is
important for both local and long distance transport of Ls. pv. phaseolicla (Saettler and

Potter, 1970; Stadt and Saettler, 1981).

Smiptoms

The disease affects all parts of the plant above ground (Burkholder, 1926; Zaumeyer
and Thomas, 1957). Characteristic leaf symptoms initially appear as small, brown,
watersoaked spots on the abaxial surface of the leaves 3-5 days after infection (Omer and
Wood, 1969). Later, a halo-like zone of greenish yellow tissue develops around the water-
soaked areas. Patel and Walker (1965) observed that halos and pronounced systemic
chlorosis develop more commonly at 16-20°C than at 24-28°C. They also noted that age
and nutritional status of the host greatly influenced disease development. Older leaves were
more tolerant than young leaves, while extreme low and high levels of nitrogen, phosphorus
and potassium retarded the development of halo blight.

Systemic chlorosis may occur. Both the halo and systemic chlorosis are due to a
non-host specific toxin called phaseolotoxin produced by the bacterium during infection
(Coyne and Schuster, 1974; Coyne e_t a_l., 1971; Schuster and Coyne, 1975a; Schwartz, 1989;
Zaumeyer and Thomas, 1957). Phaseolotoxin has been purified and characterized as an

ornithine-alanine-arginine tripeptide carrying a phosphosulfamyl group (Ferguson and

9
Johnston, 1980; Mitchell, 1976; Mitchell and Bieleski, 1977; Patil e_t a_l., 1972; Turner and
Mitchell, 1985).

The molecular genetics of phaseolotoxin production and immunity in fig. pv.
phaseolicola have been investigated by Tn; mutagenesis and cosmid cloning procedures
(Peet pt 91., 1985). These procedures indicate that soon after the tripeptide is secreted by
bacteria into the plant, plant enzymes (peptidases) cleave the peptide bonds and release
alanine, arginine and phosphosulfamylornithine. The latter is the biologically functional
moeity of phaseolotoxin (Mitchell and Bieleski, 1977; Patil gt a_l., 1972). The toxin affects
cells by binding to the active site of, and inactivating, the enzyme ornithine
carbamoyltransferase (OCIhse) which normally converts ornithine to citrulline, a precursor
of arginine. By its action on the enzyme, the toxin thus causes accumulation or ornithine
and depleted levels of arginine (Mitchell and Bieleski, 1977; Patil _e_t_ pl, 1972; Turner and
Mitchell, 1985).

Studies by Peet gt a_l. (1985) indicate that a toxin insensitive OClhse activity found
in toxigenic strains of the bacterium may be involved in the organism’s natural immunity to
its own toxin. These researchers also suggest that at least some of the genes involved in
phaseolotoxin production, as well as structural gene(s) for OCIhse may be clustered.
Further research is, however, needed to determine whether the loci affecting phaseolotoxin
production are structural genes involved in phaseolotoxin biosynthesis or in other pathways
indirectly related to phaseolotoxin production.

There is evidence that phaseolotoxin production may be suppressed in infected bean
plants possessing hypersensitive resistance but not in susceptible bean plants (Allen, 1983).
However, it has also been reported that pretreatment of resistant cultivars with the toxin

can suppress the hypersensitive response and phytoalexin accumulation. Thus, it is likely

10
that the factor which determines hypersensitive resistance to halo blight disease is associated
with suppression of phaseolotoxin production (Gnanamanickam and Patil, 1976).

Stern and pod symptoms include typical greasy spots. On pods, lesions appear as
green water-soaked spots which may enlarge and coalasce. Pod lesions in highly susceptible
cultivars are water-soaked and may be colorless, while resistant cultivars may show less
water soaking and reddish brown edges of restricted lesions (Schwartz, 1989; Stadt and
Saettler, 1981; Zaumeyer and Thomas, 1957). Under humid conditions, a cream white
exudate may be present in the center of the lesion and older lesions often have a shiny
surface appearance indicating the presence of dried bacteria (Zaumeyer, 1932; Zaumeyer
and Thomas, 1957). Infected developing pods may rot or become shrivelled and discolored.
Stem girdling or joint rot occurs at nodes above the cotyledons when infection originates
from contaminated seed (Schwartz, 1989; Zaumeyer and Thomas, 1957).

Very small spots may appear on the seed, or the entire seed may be destroyed. In
pigmented seed, it is difficult to distinguish spots caused by halo blight bacteria from those
caused by other bean pathogens. Such pathogens include Colletotrichum lindemuthianum
(Sacc. et Magn.) Briosi et Cav., the causal agent of bean anthracnose, and Xanthomonas
camp_e§tris pv. p_h_as_eofl (EB Smith) Dowson, the causative organism of bean common
bacterial blight (Msuku, 1984). A greater proportion of infected seed occurs when infection

occurs earlier in plant development (Schwartz, 1989).

Pathogenic variation

Pathogenic variation among populations of Ls. pv. phaseolicola is well known.
Using 13 isolates, Jensen and Livingstone (1944) were the first to demonstrate pathogenic

variation in _P_.s. pv. phaseolicola. However, no differences were observed in physiological

11
tests except for slight differences in growth rate. Both qualitative (Patel and Walker, 1965)
and quantitative (Schroth 91 a_l., 1971) variation in the pathogenicity of 11.9. pv. mm
to beans have been reported. Qualitative differences (pathogenicity) have conventionally
depended on the reaction of the bean cultivar Red Mexican UI-3 inoculated artificially.

Patel and Walker (1965) reported the occurrence of race 1 and 2 in the U.S.A.
These races were characterized based on the reaction of the inoculated bean cultivar, Red
Mexican IU-3, which is resistant to race 1 but susceptible to race 2. On the same basis,
the occurrence of these two races has been reported again in the U.S.A. (Schuster 91 91.,
1965); South America (Buruchara and Pastor-Corrales, 1981; Schuster and Coyne, 19753;
Schwartz, 1989); Great Britain (Epton and Deverall, 1965; Wharton, 1967); New Zealand
(Hale and Taylor, 1973); Bulgaria (Poryazov, 1975); Kenya (Kinyua and Mukunya, 1981);
Malawi (Msuku, 1984; Silbernagel and Mills, 1987); Tanzania (Taylor 91 91., 1987) and
South Africa (Edington, 1990).

The existence of another race of 2.9. pv. phaseolicola in the U.S.A. was first
suggested by Schuster _e_t 91. (1979) and Coyne e_t 91. (1979). Recently, Taylor e_t 91. (1987)
reported the existence in Africa of race 3, which was virulent toward bean cultivars with
a single gene for resistance derived from the cultivar Red Mexican UI-3. Race 3 caused
a hypersensitive reaction in the cultivar Rndergreen. Resistance to race 3 has subsequently
been found to be governed by a single dominant gene, which is also present in several
cultivars of U.S.A. origin, such as Seafarer and TEndercrop (Harper e_t 91, 1987).
Experiments examining the effect of bacterial numbers on lesion production have shown
that fewer cells of race 3 than race 1 and 2 were required for symptom development
following inoculation of pods. Antagonism between these isolates was not observed i_n_ Litpq
(Harper 91 91., 1987). D. J. Allen (personal communication) has indicated that race 3 of

2.9. pv. phaseolicola occurs in Colombia.

12

Characterization of halo blight bacteria isolates into race groups has been based on
leaf and/or pod reactions (Buruchara, 1983; Hale and Taylor, 1973; Patel and Walker, 1965)
following inoculation with bacteria, and on bacteriophage tests (Hale and Taylor, 1973;
Thylor, 1970). Schroth 91 91. (1971) indicated that possibly, there are many strains of 2.9.
pv. phaseolicola which vary in virulence. Neither race 1 nor race 2 were homogeneous with
respect to virulence when tested on leaves of certain bean cultivars. They considered that
the practice of separating isolates of 2.9. pv. phaseolicola into race 1 and 2 on the basis of
their reaction to Red Mexican UT-3 only distinguished strains of bacteria with ditIerent
degrees of virulence. Similar observations were reported by Szarka and Velich (1979).

Some workers argue that the establishment of pathotypes on a host differential basis
is subject to variation in inoculation techniques, environmental growth conditions, and the
subjectivity of scoring for the disease (Coddington 91 91., 1987). Others feel that race
designation is not valid because serological tests show that 2.9. pv. phaseolicola antiserum
is not race specific (Guthrie, 1968). A 32P-labelled DNA probe carrying a gene(s) involved
in phaseolotoxin production by 2.9. pv. phaseolicola has been used to detect and identify
the bacterium in pure and mixed cultures. Hybridization tests were highly reliable and no
race specificity was reported (Schaad e_t 91., 1989).

Quantitative differences (virulence) have been related to variation in toxin
production and in motility (Patel e_t a_l., 1964; Russell, 1975; Saettler e_t 91., 1981). Jensen
and Livingstone (1944) and Johnson (1969) reported that halo-less 2.9. pv. phaseolicola
isolates were less virulent than those that produced halos. Studies by Mulrean and Schroth
(1979) indicated that motility in this pathogen may be regulated by chemicals produced by
plant tissues in potential infection sites on leaves. Some researchers have attributed the

variation in virulence of strains of halo blight bacteria to differences in motility.

13

Panopoulos and Schroth (1974) observed that motile strains caused up to twelve times as
many lesions as non-motile stains. However, Msuku (1984) found no relationship between
motility and virulence in a study of four pathotype groups in Malawian isolates of halo
blight bacteria. This study also indicated that most of the group 1 isolates were obtained
from warmer areas as opposed to the more virulent isolates in groups 2, 3, and 4, which
were obtained from cooler areas of Malawian bean growing regions. The author noted
that with an exception of one isolate in pathotype 4, all Malawian isolates serologically
shared the same antigenic characteristies.

Appearance of new virulent strains has been attributed to a number of factors.
These include genetic changes in the pathogen that has long occupied a given area,
introduction of an organism into a new area, and changes in the cropping systems which
affect the ecological niche of the pathogen (Buruchara, 1983; Schuster and Coyne, 19753).
Studies have also indicated that the virulence of a bacterial population was increased
through mutation and selection during passage through a resistant host (Buruchara, 1983).
In other host-pathogen systems variation in pathogenicity may also evolve with the host
genotype. Correa (1987), using isozyme studies, suggested that each pathogen population
is subject to a selection pressure in favor of one allele or another, probably depending on

the genotypes serving as hosts.

Sypergism

Synergism between 2.9. pv. phaseolicola and Uromyces M (Reben) Wint., the
bean rust pathogen, has been reported (Schwartz, 1989). Yarwood (1969) observed that
lesion size became larger when plants were infected with a bean rust pathogen followed by
infection with halo blight bacteria. Lesion numbers may also be increased by inoculating

2.9. pv. phaseolicola mixed with Achromobacter sp. (Maino, 1972).

14
Cultural control

Halo blight bacteria are known to survive between growing seasons in bean debris
on the soil surface and on volunteer beans (Schuster and Coyne, 1975b). Deep plowing
and crop rotation are thus recommended to reduce initial inoculum. The pathogen is also
seed-borne. In order to reduce initial inoculum density, the use of pathogen-free seed
produced under conditions unfavorable to the bacterium is highly recommended (Schwartz,
1989; Zaumeyer and Thomas, 1957). Clean seed production is a major method for
controlling halo blight and other bean bacterial diseases in the U.S.A., where clean seeds
are produced in the arid west under irrigation. The production of clean seed in the arid
west (Idaho) depends on field inspection for visible evidence of plant infection, laboratory
inoculation of susceptible pods with suspensions from seed lots, serological tests and
quarantine to prevent importation of bean seed from areas where the pathogen exists
(Schwartz, 1989).

Seeds should also be thoroughly cleaned of dust after threshing because they can
be contaminated by halo blight bacteria present in powdered plant tissue. Contaminated
seed can also be treated with chemicals, including antibiotics to kill bacteria present on the
surface (Grogan and Kimble, 1967; Russell, 1975; Saettler _e_t 9]., 1981). While field
inspection and roguing of diseased plants may help to ensure that seeds are pathogen free,

contamination of seed may still occur, even when halo blight incidence is negligible,even

in certain resistant cultivars (Allen, 1983; Katherman g1 91., 1980; Stadt and Saettler, 1981).

Chemical control
Various researchers have indicated that halo blight can be controlled chemically by

foliar applications of Bordeaux mixture, copper oxychloride, copper sulfate, copper oxide,

15
streptomycin sulfate, or dihydro-streptomycin sulfate. These chemicals may be applied at
7 to 10 day intervals at rates of 200-400 g per 1000 square meters. Application may also
be done at first flower and pod-set at the rate of 0.1 percent a.i. per 675 liters per hectare,
to prevent spread and development of halo blight on leaves and pods (Hagedorn e_t 91.,
1969; Ralph, ‘1976; Saettler and Potter, 1970; Taylor and Dudley, 1977, Zaumeyer and
Thomas, 1957).

Russell (1975) and Schwartz (1989) suggest that application of antibiotics to the
foliage may induce the development of resistant mutants, and their use should therefore be
avoided. Moreover, control of halo blight disease by use of chemicals may not always be
practical in tropical areas (Allen, 1983). In addition, chemicals are relatively expensive for
small holder growers in Tanzania and other developing countries, and may also be difficult

to obtain (Saettler e_t 91., 1981), or to apply appropriately.

Plant resistance

The use of resistant varieties as a means to control halo blight has been successful
in some areas of the world (Baggett and Frazier, 1967; Coyne _e_t_ a_l., 1967; Schwartz, 1989;
Zaumeyer and Meiners, 1975). However, development of resistant varieties must take into
account variability of both the pathogen and genetic resistance in the host. Independent
genes separately govern leaf resistance, pod resistance and plant systemic chlorotic reactions
(Baggett and Frazier, 1967; Coyne and Schuster, 1974; Coyne e_t 91., 1971). For example,
pod susceptibility frequently occurs in plants which possess leaf resistance. However,
linkage has been observed between the different genes that control leaf and plant systemic
chlorotic reactions (Hill _e_t_ a_l., 1972). Breeding programs should, therefore, select
germplasm which provides resistant reactions against leaf and pod infection, and in which

the pathogen is non-systemic (Coyne and Schuster, 1974).

16

Plant resistance to 2.9. pv. p11£00_lic_ol_a has been reported to involve suppression
of toxin production and inhibition of bacterial growth (Russell, 1977). Bean germplasm
resistant to race 1 and 2 has been identified in field and greenhouse tests (Schwartz, 1989)
and is under the control of both dominant and recessive genes. Dominant resistance to
race 1 was reported in Red Mexican UI-3 by Patel and Walker (1966). Varieties Wisconsin
HBR40 and 72 developed by Hagedorn 91 91,. (1974) are reported to be resistant to both
races 1 and 2 of Psp, Xanthomonas plla9epli (common bacterial blight); 2.9. pv. 9vg'pga_e
(bacterial brown spot), and various fungal pathogens. However, for successful long term
control of halo blight disease, integrated control programs should be adopted (Schwartz,
1989).

In an integrated control program, several disease control methods are employed,
including regulatory inspections for healthy seed production, quarantine measures, cultural
practices (rotation, sanitation), biological control (resistant varieties) and chemical control
(seed treatment and foliage sprays). The optimum use of each of these control measures
makes the others relatively more effective. However, such integrated disease control
measures are most successful and economical when all relevant information regarding the
crop, its pathogen, the environmental conditions expected to prevail, locality, availability of

materials, and costs are taken into account.

LITERATURE CITED

Adam, DB. and AT Pugsley. 1934. Smooth-rough variation in Phytomonas medicaginis
phaseolicola Burk. Austr. J. Expt. Biol. Med. Sci. 12:193-202.

Allen, DJ. 1983. The Pathology of Tropical Food Legumes: Disease Resistance in Crop
Improvement. John Wiley and Sons. Chichester. 413pp.

Allen, DJ., M. Dessert, P. Ti'utmann, and J. Voss. 1989. Common beans in Africa and their
constraints. pp. 9-31. I_n: Schwartz, HE and MA. Pastor-Corrales (Eds). Bean
Production Problems in the Ti'opies. CIA'I; Cali, Colombia. 726pp.

Baggett, J .R. and WA. Frazier. 1967. Sources of resistance to halo blight in Phaseolus
vulgaris. Plant Dis. Reptr. 51:661-665.

Bradbury, J.F. 1986. Guide to Plant Pathogenic Bacteria. CAB. International Mycological
Institute. Kew. 332pp.

Buddenhagen, I.W. 1965. The relation of plant pathogenic bacteria to the soil. pp. 269-
282. I_n: Baker, KB and WC. Snyder (Eds). Ecology of Soil-Borne Plant
Pathogens, Prelude to Biological Control. John Murray, London. 571 pp.

Burkholder, WH. 1926. A new disease of bean. Phytopathology 16:915-927.

Buruchara, RA. 1983. Determination of Pathogenic Variation in Isariomis gr_1_s' cola Sacc.
and Pseudomonas sm'ngae pv. phaseolicola (Burk.) Young, Dye & Wilkie. Ph.D.
Dissertation. University of Nairobi, Kenya. 188pp.

Buruchara, RA. and MA. Pastor-Corrales. 1981. Variation, virulence of Psepcflnonas
sm'ngae pv. phaseolicola on bean in Colombia. 19: Lozano, J.C. (Ed). Proceedings
of the Fifth International Conference on Plant Pathogenic Bacteria. August 16-23.
CIAT Cali, Colombia. pp. 341-351.

CIAT 1981. Potential for field beans in eastern Africa. Proceedings of a Regional
Workshop held in Lilongwe, Malawi. March 9-14. 1980. 226pp.

CIA'II 1987. Annual Report. Bean Program. Working document No. 14, 1986, Cali,
Colombia. 352pp.

CIAT 1986. Bean Production Systems in Africa. CIAT Cali, Colombia. 16pp.

CMI. 1965. Pseudomonas phaseolicola. Description of plant pathogenic fungi and bacteria.
No. 45. 2pp.

Coddington, A., P.M. Mathews, C. Cullis, and KH. Smith. 1987. Restriction digest pattern

of total DNA from different races of Fusarium ommrum f.sp. p131, an improved
method for race classification. J. Phytopathol. 118:9-20.

17

18

Corey, RR. and M.P. Starr. 1957. Colony types of Xanthomonas phaseoli. J. Bacteriol.
74:137-140.

Correa, FJ. 1987. Pathogenic Variation, Production of Tbxin Metabolites, and Isozyme
Analysis in Phaeoisariopsis gr_r_s' eola (Sacc.) Ferr. Ph.D. Dissertation, Michigan State
University, East Lansing. 154pp.

Coyne, DP. and M.L Schuster. 1974. Breeding and genetic studies of tolerance to several
bean (Phaseolus vulgaris L) bacterial pathogens. Euphytica 23:651-656.

Coyne, D.P., M.L Schuster, and CC. Gallegos. 1971. Inheritance and linkage of the halo
blight systemic chlorosis and leaf watersoaked lesions in Phaseolus vulgaris L variety
crosses. Plant Dis. Reptr. 55:203-207.

Coyne, D.P., M.L Schuster, and C. Erwin. 1979. Reaction of Phaseolus vulgaris germplasm
to a new virulent strain of Pseudomonas phaseolicola Ann. Rept. Bean Improv.
Coop. 22:20-21.

Coyne, D.P., M.L Schuster, and R. Fast. 1967. Sources of tolerance and reaction of bean
to races of halo blight bacteria. Plant Dis. Reptr. 51:20-24.

CRSP. 1990. Summary Report for Sokoine University of Agriculture. Presented at the
Bean/Cowpea Collaborative Research Support Program Research Meeting.
University Place Holiday Inn, East Lansing. April 30—May 3 (Abstr.).

Doudoroff, M. and NJ. Pallerozin. 1974. Genus I: Pseudomonas Migula 1894. pp. 217-
243. 19: Buchanan, RE. and NE. Gibbons (Eds). Bergey’s Manual of
Determinative Bacteriology, 8th Edition, Bailliéré, London.

Edington, BR. 1990. The identification of race 1 of Pseudomonas m’ngae pv. phaseolicola
in South Africa. Ann. Rept. Bean Improv. Coop. 33:171.

Epton, H.A.S. and BJ. Deverall. 1965. Physiological races of Pseudomonas m’ngae pv.
phaseolicola causing halo blight of beans. Plant Path. 14:53-54.

Epton, H.S.A., D.C. Sigee, and M. Passmoore. 1977. The influence on pathogenicity of
ultrastructural changes in Pseudomonas phaseolicola during lesion development. pp.

301-303. I_n: Kiraly, Z. (Ed). Current prics in Plant Pathology. Academial Kiado,
Budapest.

Ercolani, G.L, D.J. Hagedorn, and RE. Rand. 1974. Epiphytic survival of Pseudomonas
m‘ngae on hairy vetch in relation to epidemiology of bacterial brown spot of bean
in Wisconsin. Phytopathology 64:1330-1339.

Ferguson, AR. and J.S. Johnston. 1980. Phaseolotoxin, chlorosis, ornithine accumulation
and inhibition of ornithine carbamoyltransferase in different plants. Physiol. Plant
Path. 16:269-275.

19

Gnanamanickam, SS. and SS. Patil. 1976. Bacterial growth, toxin production, and levels of
ornithine carbamoyltransferase in resistant and susceptible cultivars of bean
inoculated with Pseudomonas phaseolicola. Phytopathology 66:290-294.

Gondwe, B. 1987. Halo blight of Phaseolus beans. pp. 70-74. 19: Salema, MP. and Minjas,
A.N. (Eds). Bean Research 1. Proceedings of the Fifth Bean Research Workshop
held at Sokaine University of Agriculture, Morogoro, Tanzania, September 9-12,
1986. Benedictine Publications, Ndanda, Peramiho. 167pp.

Grogan, RG. and KA. Kimble. 1967. The role of seed contamination in the transmission
of Pseudomonas phaseolicola in Phaseolus vulgaris. Phytopathology 57:28-31.

Guthrie, J.W. 1968. The selorogical relationship of races of Pseudomonas phaseolicola.
Phytopathology 58:716-717.

Hagedorn, D.J., E.K. Wade, and G. Weis. 1969. Chemical control of bean bacterial diseases
in Wisconsin. Plant Dis. Reptr. 53:178-181.

Hagedorn, D.J., J.C. Walker, and RE. Rand. 1974. Wis. HBR40 and Wis. HBR 72 bean
germplasm. Hort. Sci. 9:402.

Hale, C.N. and JD. Thylor. 1973. Races of Pseudomonas phaseolicola causing halo blight
of beans in New Zealand. New Zealand J. Agric. Res. 167:147-149.

Harper, 8., N. Zewdie, I.R. Brown, and J.W. Mansfield. 1987. Histological, physiological,
and genetical studies of the responses of Pseudomonas gri_r_1g99 pv. phaseolicola and
Pseudomonas m’ngae pv. coronafaciens. Physiol. Mol. Plant Path. 31:153-172.

Hill K., D.P. Coyne, and M.L Schuster. 1972. Leaf, pod, and systemic chlorosis reactions
in Phaseolus vulgaris to halo blight controlled by different genes. J. Am. Soc. Hort.
Sci. 97:494-498.

Jensen, J.H. and J .E. Livingstone. 1944. Variation in symptoms produced by isolates of
Phy10monas medicaginis var. phaseolicola. Phytopathology 34:471-480.

Johnson, J.C. 1969. Halo-less halo blight of French bean in Queensland. Queensland J.
Agric. and Anim. Sci. 26:293-302.

Karel, A.K., B.J. Ndunguru, M. Price, S.H. Semuguruka, and BB. Singh. 1981. Bean
production in Thnzania. _I_r_1_: CIAT (Ed.). Potential of FiCId Beans in Eastern Africa.
Proceedings of a Regional Workshop held in Lilongwe, Malawi, March 9-14, 1980.
Cali, Colombia. 226p.

Katherman, M.J., R.E. Wilkinson, and S.V. Beer. 1980. Resistance and seed infection in
three dry bean cultivars exposed to a halo blight epidemic. Plant Dis. 64:857-859.

20

Kinyua, GK. and D. Mukunya. 1981. Variability in isolates of Pseudomonas phaseolicola
(Burk.) Young, Dye and Wilkie in Kenya and genetic studies on resistance in dry
food beans. pp. 352-357. I_p: Lozano, J.C. (Ed) Proc. Fifth Int. Conf. on Plant
Pathogenic Bacteria, August 16-23, CIAT Cali, Colombia.

Kovacs, N. 1956. Identification of Pseudomonas pyocyanin by the oxidase reaction. Nature
178:703.

Leben, C. and J .P. Sleesman. 1982 Preservation of plant pathogenic bacteria on silica gel.
Plant Dis. 66:327.

Legard, DE. and HF Schwartz. 1987. Sources and management of Pseudomonas sm'ngae

pv. phaseolicola and Pseudomonas m’ngae pv. m’ngae epiphytes on dry beans in
Colorado. Phytopathology 77:15-3-1509.

Maino, AL 1972. Degradation of bean cell walls during early stages of halo blight

infections caused by Pseudomonas phaseolicola and interactions with Achromobzlcter
sp. Phytopathology 62:775 (Abstr.)

Mitchell, R.E. 1976. Isolation and structure of a chlorosis inducing toxin of Pseudomonas
phaseolicola. Phytochem. 15:1941-1947.

Mitchell, R.E. and R.L Bieleski. 1977. Involvement of phaseolotoxin in halo blight of
beans. Transport and conversion to functional toxin. Plant Physiol. 60:723-729.

Moffett, M.L 1983. Bacterial plant pathogens recorded in Australia. pp. 317-336. I_n: Fahy,
PC. and GJ. Persley (Eds). Plant Bacterial Diseases, A Diagnostic Guide. Academic
Press. Sydney New York London, 393pp.

Moore, LW. and RV. Carlson. 1975. Liquid nitrogen storage of phytopathogenic bacteria.
Phytopathology 65:246-250.

Msuku, W.A.B. 1984. Pathogenic Variation in Malawian Isolates of Pseudomonas syg'ngae
pv. phaseolicola (Burk.) Young, Dye and Wilkie and Implication for Breeding
Disease Resistant Beans. Ph.D. Dissertation, Michigan State University, East

Lansing, 90pp.

Mulrean, E.N. and M.N. Schroth. 1979. 19 vitro and 19 M chemotaxis by Pseudomonas
phaseolicola. Phytopathology 69:1039 (Abstr.)

Omer, M.E.H. and R.K.S. Wood. 1969. Growth of Pseudomonas phaseolicola in susceptible
and in resistant bean plants. Ann. Appl. Biol. 63:103-116.

Panopoulos, NJ. and M.N. Schroth. 1974. Role of flagela motility in the inversion of bean
leaves by Pseudomonas syg'ngac pv. phaseolicola. Phytopathology 64:1389-1397.

21

Patel, RN. and J.C. Walker. 1965. Resistance in Phaseolus to halo blight. Phytopathology
55:889-894.

Patel, RN. and J .C. Walker. 1966. Inheritance of tolerance to halo blight in beans.
Phytopathology 56:681-682.

Patel, P.N., J.C. Walker, D.J. Hagedorn, C. Deleon-Garcia, and M. Tbliz-Ortiz. 1964.
Bacterial brown spot of beans in central Wisconsin. Plant Dis. Reptr. 48:335-337.

Patil, S.S., LO. Pam, and WS. Sakai. 1972. Mode of action of the toxin from Pseudomonas
phaseolicola part 1. Plant Physiol. 49:803-812.

Peet, R.C., R.B. Lindgren, and NJ. Panopoulos. 1985. Molecular genetics of phaseolotoxin
production and immunity of Pseudomonas m‘ngae pv. phaseolicola. pp. 487-497.
I_n: Civerolo, E.L, A. Collmer, R.E. Davis, and AG. Gillaspie (Eds), Plant
Pathogenic Bacteria. Proceedings of the Sixth International Conference on Plant
Pathogenic Bacteria, Maryland, June 2-7, Martinus Nijhoff Publishers, Dordrecht.
1050pp.

Poryazov, I. 1975. Physiological races of Pseudomonas 9m'ngae pv. phaseolicola (Burk.)
Young, Dye & Wilkie in Bulgaria. Ann. Rept. Bean Improv. Coop. 18:57-58.

Ralph, W. 1976. Pelleting seed with bacteriocides: The effect of streptomycin on seed-
borne halo blight of French bean. Seed Sci. Technol. 4:325-332.

Russell, RE. 1975. Variation in virulence of some streptomycin resistant mutants of
Pseudomonas phaseolicola. J. Appl. Bacteriol. 39:175-180.

Russell, RE. 1977. Observation on the i9 vitro growth and symptom production of
Pseudomonas phaseolicola on Phaseolus vulgaris. J. Appl. Bacterial. 43:167-170.

Saettler, AW. and HS. Potter. 1970. Chemical control of halo blight in field beans.
Research Report No. 98. MSU Agric. Expt. Sta. Extension Bull. 8pp.

Saettler, A.W., SJ. Stadt, and LT Pontius. 1981. Effect of inoculation time and cultivar

on internal infection of bean seed by Pseudomonas phaseolicola. J. Seed Tbchnol.
6:23-30.

Sands, D.C., M.N. Schroth, and DC. Hildebrand. 1970. Taxonomy of phytopathogenic
pseudomonads. J. Bacteriol. 101:9-23.

Schaad, N .W., H. Azaad, R.C. Peet and NJ. Panopoulos. 1989. Identification of
Pseudomonas m’ngae pv. phaseolicola by a DNA hybridization probe.
Phytopathology 79:903-907.

22

Schroth, M.N., V.B. Vitanza, and DC Hildebrand. 1971. Pathogenic and nutritional
variation in the halo blight group of fluorescent pseudomonads of bean.
Phytopathology 61:852-857.

Schuster, M.L and DR Coyne. 1975a. Genetic variation in bean bacterial pathogens.
Euphytica 24:143-147.

Schuster, M.L and DP Coyne. 1975b. Survival factors of plant pathogenic bacteria.
Research Bull. No. 268, Agric. Expt. Sta., University of Nebraska, Lincoln, NE,
USA. 53pp.

Schuster, M.L, D.P. Coyne, and ED. Kerr. 1965. New virulent strain of halo blight
bacterium overwinters in the field. Phytopathology 55:1075 (Abstr.).

Schuster, M.L, D.P. Coyne, and C. Smith. 1979. New strain of halo blight bacterium in
Nebraska. Ann. Rept. Bean. Improv. Coop. 22:19-20.

Schwartz, HF. 1980. Halo blight. pp. 175-182. 19: Schwartz, HP. and EG. Guillerno
(Eds). Bean Production Problems. Disease, Insect, Soil and Climatic Constraints of
Phaseolus vulgaris. CIAT Cali, Colombia, 424pp.

Schwartz, HP. 1989. Halo blight. pp. 285-301. I_n_: Schwartz, H.E and MA. Pastor-
Corrales (Eds). Bean Production Problems in the 'B'opics. 2nd Edition. CIAT Cali,
Colombia, 726pp.

Schwartz, HF. and DE. Legard. 1986. Bacterial Diseases of Beans. Colorado State
University Serv. Act. Bull. 2941, 4pp.

Schwartz, HP. and MA. Pastor-Corrales. 1989. Bean Production Problems in the Ti'0pics.
2nd Edition. CIAT; Cali, Colombia, 726pp.

Silbernagel, MJ. and MJ. Mills. 1987. Halo blight of Phaseolus vulgaris. pp. 81-83. 19:
Salema, MP. and Minjas, A.N. (Eds). Bean Research 3. Papers presented at a
workshop held at Sokoine University of Agriculture, Morogoro, Tanzania, September
9-12, 1986. Benedictine Publications, Ndanda, Peramiho, 167pp.

Stadt, SJ. and AW. Saettler. 1981. Effect of host genotype on multiplication of
Pseudomonas phaseolicola. Phytopathology 71:1307-1310.

Starr, G.H. and CJ. Kercher. 1969. Passage of Pseudomonas phaseolicola in bean plants
through sheep. Phytopathology 59:1976.

Szarka, J. and I. Velich. 1979. A study of the aggressivity of isolates belonging to
Pseudomonas sm'ngae pv. phaseolicola (Burkh.) Dowson. Ann. Rept. Bean Improv.
Coop. 22:64-65.

23

Taylor, JD. 1970. Bacteriophage and serological methods for the identification of
Pseudomonas phaseolicola (Burkh.) Dowson. Ann. Appl. Biol. 66:387-395.

Taylor, J. D. and C. L Dudley. 1977. Effectiveness of late copper and streptomycin sprays
for the control of halo blight of beans (Pseudomonas phaseolicola ). Ann. Appl.
Biol. 85: 217-221.

Taylor, J.D., C.L Dudley, and L Presly. 1979. Studies of halo blight seed infection and
disease transmission in dwarf beans. Ann. Appl. Biol. 93:267-277.

Taylor, J.D., D.M. Tbverson and J.H.C. Davis. 1987. Diseases of Phaseolus beans - biology,
resistance and control (RO 007035; RO 032055). pp 63-64. 19: Inst. Hort. Res.
Ann. Rept. Wellsborne, Warwick, England, 1986/87.

Teri, J.M., M.J. Silbernagel and S. Nchirnbi. 1990. The SUA/WSU CRSP: A review of the
progress in disease resistance. Paper Presented at the Bean Workshop Held at
Sokoine University of Agriculture, Morogoro, Tanzania. September 17-19. (Abstr.)

Turner, J.G. and R.E. Mitchell. 1985. Association between symptom development and
inhibition of ornithine carbamoyltransferase in bean leaves treated with
phaseolotoxin. Plant Physiol. 79:468-473.

Walker, J .C. and RN. Patel. 1964. Splash dispersal and wind as factors in epidemiology of
halo blight of beans. Phytopathology 54:140-141.

Wharton, AL 1967. Sources of infection by physiological races of Pseudomonas gm'ngae
pv. phaseolicola in dwarf beans. Plant Path. 16:27-31.

Yarwood, CE. 1969. Association of rust and halo blight on beans. Phytopathology 59:1302-
1305.

Young, J.M., D.W. Dye, J.E Bradbury, C.G. Panagopoulos, and CE Robbs. 1978. A
proposed nomenclature and classification for plant pathogenic bacteria. N.Z.J.
Agric. Res. 21:153-177.

Zaumeyer, WJ. 1932. Comparative pathological, histology of three bacterial diseases of
bean. J. Agric. Res. 44:605-632.

Zaumeyer, WJ. and J.P. Meiners. 1975. Disease resistance in beans. Ann. Rev. Phytopath.
13:313-334.

Zaumeyer, WJ. and HR. Thomas. 1957. A Monographic Study of Bean Diseases and
Methods of their Control. U.S.D.A. Agric. Techn. Bull. No. 868. 255pp.

Chapter 2

PATHOGENTC VARIATION OF PSEUDOMONAS SYRINGAE

PVI PHASEOLICOLA IN NORTHERN TANZANIA

INTRODUCTION

Pathogenic variation among populations of Pseudomonas m pv. phaseolicola
is well known. Using 13 isolates Jensen and Livingstone ( 1944) were the first to
demonstrate pathogenic variation in 2.9. pv. phaseolicola. Patel and Walker (1965) reported
the occurrence of races 1 and 2 in the U.S.A. These races were characterized based on the
reaction of inoculated bean cultivar Red Mexican UI-3, which is resistant to race 1 but
susceptible to race 2. On the same basis, the occurrence of these two races has been
reported in Tanzania and many other countries (Buruchara and Pastor-Corrales, 1981; Hale
and Taylor, 1973; Kinyua and Mukunya, 1981; Silbernagel and Mills, 1987; Taylor _t _1.,
1987). The existence of a new race in Nebraska, U.S.A. was first suggested by Schuster 91
91., ( 1979) and Coyne 91 91., (1979), who recovered new strains of halo blight bacteria which
were virulent to the usually resistant Great Northern bean cultivars such as Great Northern
111-59 and California Pink. Recently, Taylor 91 a_l., (1987) confirmed the erdstence in Africa,
of race 3 of 2.9. pv. phaseolicola, which was virulent on bean cultivars with a single gene
for resistance derived from cultivar Red Mexican UI-3. Race 3 caused a hypersensitive
reaction in the cultivar Tendergreen, which is susceptible to races 1 and 2. Resistance to
race 3 has been found to be governed by a single dominant gene, which is also present in

several cultivars of U.S.A. origin such. as Seafarer and Thadercrop (Harper 91 91., 1987).

DJ. Allen (personal communication) indicated that race 3 occurs in Colombia.

26

Breeding for halo blight resistance is generally considered the best method of
control, and the use of resistant varieties as a means to control halo blight has been
successful in some areas of the world (Baggett and Frazier, 1967; Coyne 91 a_l., 1967;
Hagedorn 91 91., 1974; Schwartz, 1989). However, the development of resistant varieties
must take into account variability of both the pathogen and the genetic resistance in the
host. Therefore, a knowledge of the distribution of races of halo blight bacteria is important
in relation to programs of breeding resistant bean varieties, including those in Tanzania.
The purpose of this study, therefore, was to investigate races of 2.9. pv. phaseolicola
prevailing in northern Tanzania and their survival ability under different field conditions.
Information generated from this study will be very useful for bean breeding programs for
disease resistance, and their development will also provide an alternative management tool

for halo blight disease in the country.

MATERIALS AND METHODS

Sampling and isolation

To determine the relative prevalence of races of 2.9. pv. phaseolicola in Arusha and
Kilimanjaro regions, northern Tanzania (Fig. 1) where halo blight is one of the important
diseases of beans, a survey was conducted during the period of November 1988 to February,
1990. Samples were collected periodically at random from several farmers’ bean fields. A
total of 118 isolates were collected. Isolations were made from diseased plant material with
halo blight-like symptoms. Pieces of diseased tissue including small areas of surrounding
healthy tissue were excised and surface-sterilized for 1 minute in 2.6% NaOCl and rinsed
in sterile distilled water. The pieces were crushed on flame-sterilized glass slides containing

one or 2 drops of sterile distilled water. Using a sterile wire loop, the resulting suspensions

27

 

 

 

 

 

 

g
g
J g‘

28
were streaked onto medium B of King 91 91. (1954) (KMB). F1uorescing colonies were
purified by a series of single colony transfers and their identity confirmed by biochemical
and physiological tests using procedures of Lelliot 91 91. (1966); pathogenicity tests on
Canadian Wonder bean seedlings; carbon source utilization tests (Schaad, 1988) using
mannitol, sorbitol and inositol; ice nucleation activity using the single temperature droplet
method of Lindow (1988) and Lindow 91 91. (1978, 1982) at -5°C; sensitivity to

bacteriophage and serology.

Bacteriophage tests

Sensitivity to bacteriophage was tested using the method described by J .D. Taylor
and D.M. Thverson (personal communication), as follows. Three bacteriophages (11P, 12P,
48P), specific to 2.9. pv. phaseolicola, were used. Log phase cultures of each isolate grown
on KMB and nutrient agar were used. Thick bacterial suspensions were prepared by adding
4 ml nutrient broth to each slant culture. Two to 3 drops of the resulting bacterial
suspension of each isolate were added to glass vials containing 2.5 ml of sterile soft glycerol
agar (SGA) (g/l:proteose peptone (Sigma), 5.0; yeast extract (Difco), 3.0; glycerol, 20 ml;
Bacto-agar, 7.0) maintained molten in a water bath at 45-50°C. Suspensions were thoroughly
mixed and poured on cool, solidified nutrient agar plates, followed by swirling gently to
allow a thin layer of SGA with bacterial suspension to form over the plate. Plates were
dried in a laminar flow chamber. One 5-ul drop of each bacteriophage was applied to the
surface of the medium at designated spots and the drops were allowed to dry on a
laboratory bench. Plates were then incubated upside down for 24-48 hours after which
results were recorded. Each isolate was replicated three times. Clear zones (plaques) due

to lysis of bacterial cells by phages indicated a positive 2.9. pv. phaseolicola identity. Three

29
tests were conducted for each bacterial isolate, and reference strains 1299A and 1375A

were included as positive controls.

Serology

2.9. pv. phaseolicola antisera was provided by Dr. J.D. Taylor. The agglutination
method of J .D. Taylor and D.M. Thverson (personal communication) was used. One 0.07-
ml drop of the antiserum was placed on a clean glass microscope slide, and a small amount
of 24-48 hour bacterial growth picked up with a platinum wire loop was mixed into the
drop of the antiserum, which was stained pink. Agglutination was observed against a light
colored background. Occurrence of agglutination was recorded as positive. Three tests were

conducted for each isolate.

Reference strains

Reference strains of 2.9. pv. phaseolicola (882, 1281A, 1299A, 1301A, 1302A and
1375A) were included in this investigation. Bacteriophage, antisera, and reference strains
were provided by Dr. J .D. Taylor (Institute of Horticultural Research, Wellsborne, Warwick,

England).

Storage of Isolates

Bacterial isolates were maintained as very thick suspensions in glycerol:0.1M
phosphate buffer (50:50, v/V) and on nutrient agar slants at 0°C in an incubator (A.
Gallenkamp and Co., London). Isolates were also stored in bean leaf powder in sterile vials
at room temperature (22 9; 2°C) in darkness. The three storage methods were used to

reduce the chances of loss of viability.

30
Soil sterilization
Soil used for screen house experiments was heat-sterilized by exposing soil to a high
temperature kerosene flame (198°C) using a Tbrra Force Tarralizer (Kent Horticultural
Engineers, Kent, England). A forest soil/sand mixture (2/1, v/v) was used. The two soil
components were mixed and sieved to remove gravel and debris before the sterilization

process.

Growing of plants

All plants used in this study were grown in the screen house in 16 cm-diameter
sterile plastic pots or in 7 x 8 x 22 cm3 sterile plastic flats containing a sterile mixture of
forest soil/sand. Plants were maintained at temperatures ranging from 19 to 27°C and a
natural 12 i 0.5 hour day-length, and watered as required with tap water. The halo blight

susceptible cultivar, Canadian Wonder, was used as a control throughout the study.

Inoculum preparation

Each 2.9. pv. phaseolicola isolate was grown on KMB for 24-48 hours at 24 i 2°C.
Suspensions were made by adding 0.01 M phosphate buffer, pH 7.2, to each culture and
dislodging the bacteria. Concentrations of bacterial suspensions were adjusted

turbidimetrically to contain about 107 -10° colony forming units (CFU) per ml.

Race identification
Races were determined from pod and foliage reactions of four differential bean
cultivars. These were Canadian Wonder (universal susceptible), Edmund (universal

resistant), Red Mexican UI-3 (resistant to race 1) and Tandergreen (resistant to race 3).

31

Pathogen-free seeds of these differential bean cultivars were provided by Dr. M.J.
Silbernagel (USDA/ARS, Prosser, WA) and were then multiplied at Lyamungu Agricultural
Research Station, Moshi, Tanzania. For leaf reaction, 7-10 day old seedlings containing
fully expanded primary leaves were injected with bacterial suspensions at the first node
using a sterile 25 gauge needle attached to a sterile 100C hypodermic syringe. Following
needle inoculation, bean seedlings were also spray-inoculated abaxially with bacterial
suspensions without water-soaking, using a half-liter plastic hand-operated atomizer. After
inoculation, plants were covered with a plastic bag for 24 hours in the screen house and
observed for halo blight symptom development during the following 10 days. Each single
plant was considered a replicate and six plants were used for each isolate; two experiments
were conducted.

Pod reactions were determined using procedures of Ekpo (1975). Pods at the flat
stage were harvested, surface-sterilized, rinsed and dried aseptically on sterile filter papers.
Three pods from each differential bean cultivar were inoculated by placing a S-ul drop of
inoculum at three sites per pod, followed by pricking the pod five times through the
inoculum drop to a depth of about 1.0 m using a 25-gauge sterile disposable needle.
Inoculated pods were placed side by side in 7 x 8 x 22 cm3 plastic flats containing moist
filter papers with the inoculated side up and incubated for 7-10 days. Data were compiled
from two repeated experiments. Isolates of known 2.9. pv. phaseolicola races and sterile

phosphate buffer were included as positive and negative controls, respectively.

Temxrature studies

Tb determine the efl‘ect of temperature on growth of 2.9. pv. phaseolicola races,

growth studies were conducted on Km agar plates. Plates were inoculated in quadruplicate

32
with different serial dilutions of each race. After inoculation, plates were placed upside
down in an incubator (Fi-totron 600H, Fisons Environmental Equipment, Loughborough,
England) without light, at 24, 28 and 34°C with a temperature variation of _-_l-_ 2°C. At 28
and 34°C, humidity in the incubator was adjusted at 70% to avoid excessive drying of agar
plates. Colony diameter was measured after 4-5 days. The ability to produce a brown
diffusible pigment at different temperatures on KMB and NA was also examined by growing
the isolates producing the pigment at the above mentioned temperatures. Subjective

pigment production determinations were recorded during the following 5 days.

Surviv91 studies

Survival of races 1 and 2 of halo blight bacteria was studied following the
procedures of Saettler 91 91. (1986) and Groth and Braun (1989). Three sites were chosen
representing different ecological environments in which beans are grown in Arusha and
Kilimanjaro regions. The coffee/banana intercrop environment was also included. Three
cultivars, Canadian Wonder (susceptible), Masai Red (slightly susceptible) and GO 7928
(resistant) were grown in 10 m rows in the field at Lyamungu and Monduli in March, 1989.
Plants were inoculated twice with race 1 (isolate #2) or race 2 (isolate #4) 18 and 30 days
after planting. Control plants grown 25 m away were left uninoculated. Severely halo blight
diseased plants were collected in early June, air-dried at room temperature and separated
into leaves and stems. Some plants were left standing in the field. Stems with typical halo
blight symptoms were cut into pieces 2-3 cm long. Samples of leaves (0.5g) and stems (10
pieces) were placed in fine-mesh nylon bags and tied with nylon threads. A different color
of nylon-mesh was used for each race to avoid mixing the two races, and within a race

threads of different colors were used to distinguish cultivars. Samples were then taken to

33
the field at the end of June, the end of the main bean growing season. Half of the samples
were placed 2-5 cm beneath the soil surface and the remaining half placed at 25 cm depth.
Samples of each race were separated by a distance of 10 m. Healthy control samples were
placed 5 m away from diseased samples. Diseased samples were also stored in the
laboratory at 22 i 2°C for a similar period.

At monthly intervals starting in July, when beans are usually harvested, three
samples for each race and for each plant part, including samples from standing plants, were
retrieved. Samples for standing plants included stem tissue and fallen leaves. After removing
the nylon mesh, samples were ground in sterile mortars and pestles and 10 ml of sterile
phosphate buffer added. The resulting suspensions were left to stand for 15 minutes
followed by a 2-minute centrifugation at 20,000xg using a Marusan centrifuge (Sekuma
Seisakusho Ltd., Tbkyo, Japan) to sediment the plant debris. The supernatants were diluted
serially and plated in triplicate on KMB agar containing cycloheximide (100 ug/ml). Plates
were observed under UV light after 4-5 days of incubation. For each sampling period, 5
presumed 2.9. pv. phaseolicola colonies were purified by single colony transfers and tested
for sensitivity to bacteriophage and for pathogenicity on 7-10 day-old Canadian Wonder
been seedlings, by injecting stems and by infiltrating leaves with suspensions of
approximately 107 cfu/ml. Plants were incubated in the screen house and observed for halo

blight symptom development for up to 14 days.

Soil pH measurements

Soil pH for each placement site at each geographical location was measured
electrometrically using a pH meter (BIL-7015, Kent Industrial Measurements Ltd., England)
at the time of sample placement in the field and at monthly intervals when samples were

retrieved. Three soil samples were taken for each placement depth.

34

Volunteer bean plants

During the bean growing season, farmers’ fields with severe halo blight symptoms
were identified. Volunteer bean plants from such fields and around farmers’ backyards in
Monduli (November-December) and at Lyamungu (September-October) were observed for
the presence of halo blight symptoms. Isolation attempts were made from suspect plants as
described earlier. Presumed 2.9. pv. phaseolicola colonies which were fluorescent were
subject to biochemical, pathogenicity and bacteriophage tests for species identification.
Confirmed 2.9. pv. phaseolicola colonies were tested on the four bean differential cultivars

for race identification.

Weather data
Rainfall and temperatures for Lyamungu and Monduli were obtained from the
agrometeorology section at the Agricultural Research Institute, Lyamungu, Moshi, and from

the Monduli District Agricultural Development office, respectively.

RESULTS
Prevailing races
Within the two regions surveyed in northern Tanzania (Fig. 2), three pathogenic
races of halo blight bacteria were distinguished based on the reaction of the four
differential bean cultivars used. The distribution of these races in the surveyed area is
shown in Table 1. Race 2 was the most prevalent of the three races in the area, followed
by race 1; race 3 occurred at a very low frequency (Fig. 3). Besides producing the usual

blue-green diffusible pigment, some race 2 isolates also produced a brown diffusible

35
pigment, which was more pronounced on nutrient agar than on KMB. Such isolates were
designated as race 2P (Table 1 and Fig. 3). Race 2P isolates occurred infrequently, and
were restricted to Lambo, Kilimanjaro region and Monduli and Selian in Arusha regions.
The bean cultivars from which race 2P isolates were obtained included Canadian Wonder,
the breeding line FB-GP307-2 and other local varieties.

Brown diffusible pigment production could be detected as early as 48 hours on
nutrient agar and after 4-5 days on KMB. The delayed appearance of the pigment on KMB
might be due to the masking effect by the blue-green diffusible pigment which is produced
in large quantities on this medium. The brown diffusible pigment-producing isolates were
not distinguishable from other race 2 isolates that did not produce the pigment, based on
biochemical and physiological tests. Race 2P isolates induced a reaction similar to that of
race 2 on foliage and on pods of the four differential bean cultivars used.

All race 2 isolates obtained from bean debris collected from farmers’ fields in
Monduli in November, 1988 and September, 1989 were highly virulent and produced very
large water-soaked lesions on the cultivar Tandergreen. They also induced stunting and
extensive systemic chlorosis in the resistant differential cultivar Edmund. However, no water-
soaked lesions were produced in Edmund and inoculated leaves produced a hypersensitive
reaction (Fig. 4). By contrast, race 2 isolates obtained from halo blight-infected bean plants

produced only a hypersensitive reaction on Edmund.

Temperature studies

The growth responses of different isolates of the three races of halo blight bacteria
are shown in Table 2. Growth of all races did not differ much at 24 i 2 and 28 i 2°C.

However, 2 out of 3 race 2P isolates were able to grow at 34 :1; 2°C. These isolates also

36

Fig. 2. Origin of Pseudomonas yfingae pv. phaseolicola isolates in northern Tanzania. 1 =
Lambo, 2 = Lyamungu, 3 = Kilema, 4 = Miwaleni, 5 = Sanya Juu, 6 = Tangeru, 7 =
Selian, 8 = Monduli.

37

 

 

 

 

 

 

 

 

 

 

38

 

 

 

 

 

 

 

1; 50- w
;
32) 4o~— g
m /
a /
8 %
E 30‘" Z fllIIIHIZP
g é
E 20—~ g
- /
“J /
‘I —— ¢
10 ¢
/
. , // , r1 ,
‘ 1 2 3
RACE

 

Fig. 3. Relative frequency of races 1, 2 and 3 of Pseudomonas mingae pv. phaseolicola
in collections made in Arusha and Kilimanjaro regions, northern Tanzania from November,
1988 to February, 1990. Frequencies are expressed as percentages of the total number of
isolates collected. 2P represents race 2 isolates producing brown-diffusible pigment.

39

Table 1. Distribution of races of Pseudomonas m‘ngae pv. phaseolicola in Arusha and
Kilimanjaro regions, northern Tanzania.

 

 

Altitude Tbtal No. Number of isolates of race

Location (m) of isolates 1 2 2P 3
Arusha

Monduli 1630 39 8 26 3 2

Selian 1387 45 27 16 1 1

Tengeru ? 2 2 -° - -
Kilimanjaro

Lambo 1020 13 3 8 2 —

Lyamungu 1268 13 12 1 - -

Kilema 1422 4 — 4 — —

Miwaleni 765 - - - - —

N arumu 1268 1 - 1 - -

Sanya J uu 1400 1 1 - - -
Total 118 53 56 6 3

 

? Information on altitude lacking.

‘- Represents absence of indicated race among 2.9. pv. phaseolicola isolates.

 

Fig. 4. Stunting and systemic chlorosis symptoms produced by race 2 isolates obtained from
bean debris in Monduli, Arusha region northern Tanzania on a resistant differential bean
cultivar, Edmund.

41

Table 2. Influence of temperature on growth of representative isolates of races of
Pseudomonas m’ngae pv. phaseolicola on KMB.

Mean colony diameter (mm) at temgrature (°C)“'

 

Race Isolate No. 24 _+_- 2 28 i 2 34 i 2
1 50 3.6 _+_- 0.2 2.7 i 0.2 -"

1 70 2.5 :1; 0.2 2.4 i 0.1 -

2 12 2.7 i 0.2 2.6 :l_- 0.1 -

2 882’ 22:20.4 2.1 i 0.2 —

2 1299A" 2.6 i 0.0 2.4 i 0.1 -
2P2 5 2.3 _+_ 0.1 2.5 i 0.1 3.0 _-l_-_ 0.2
2Pz 19 2.5 i 0.1 2.4 5; 0.0 2.3 i 0.1
2Pz 44 2.5 i 0.2 2.4 i 0.0 -

3 46 2.6 i 0.1 2.4 i 0.1 -

3 1302Ay 2.6 i 0.1 2.3 i 0.2 -

 

”Values are means of four replications :1; standard errors.
"- = No growth was observed.

YIsolates were kindly supplied by Dr. J .D. Taylor, Institute of Horticultural Research,
Wellsborne, Warwick, England.

z2P = Race 2 brown-diffusible pigment producing isolates.

42
produced the brown diffusible pigment at all three temperatures. The pigment was easily
seen on nutrient agar plates due to the absence of the blue-green diffusible pigment which

occurs on KMB.

Survival studies

The ability of 2.9. pv. phaseolicola to survive in bean debris buried in the soil and
in standing bean plants left in the field varied depending upon the race, geographical
location, depth of debris placement and the bean genotype used. Mthin the same bean
genotype 2.9. pv. phaseolicola survived for a longer period in stems than in leaf debris. For
example, at Ngarash, Monduli, Arusha region, race 1 survived for 5 months in Canadian
Wonder stem debris and only 4 months in foliage debris (Table 3). This race also survived
well for 5 months at Monduli in standing bean plants and in bean debris placed in soil at
a depth of 2-5 cm, while it survived for only 3 months when debris was placed 25 cm deep.
On the other hand, at the same location race 2 survived in stems of cultivar Canadian
Wonder one month longer than race 1 (Table 3). Race 2 was readily recovered from bean
stem tissue samples of plants left standing in the field and those buried at 2-5 cm
throughout the 6 month period (July to December). In all cases, the susceptible cultivar,
Canadian Wonder, supported longer survival of halo blight bacteria, probably due to more
inoculum present, than the other bean genotypes used. Pathogenicity tests and sensitivity
to bacteriophages 11P, 12F and 48P confirmed the identity of recovered isolates as 2.9. pv.
phaseolicola. .

In the coffee/banana intercrop environment at Lyamungu survival of halo blight
bacteria declined rapidly in bean plants left standing after the harvest season and in bean

debris buried in the soil. Race 1 was recovered after only 1 month in leaves and 2 months

43

Table 3. Survival of Pseudomonas mingae pv. phaseolicola races 1 and 2 in infected bean
debris represented by standing plants and debris buried for 6 months at two depths under
field conditions at Ngarash, Monduli, Arusha region and at Lyamungu, Kilimanjaro region,
northern Tanzania.

S9rvival duration (mom)
Standing plants 2-5 cm deep 25 cm deep

 

Location Race Genotype" Leaves Stems Leaves Stems Leaves Stems

 

Monduli" 1 CW 4 5 4 5 2 3
MR 2 3 2 2 1 1

607928 0 0 0 0 0 0

2 CW 5 6 5 6 2 4

MR 2 3 2 2 1 1

G07928 0 0 0 0 0 0

Lyamunguy 1 CW 1 2 0 0 0 0
MR 0 2 0 0 0 0

G07928 0 0 0 0 0 0

2 CW 2 3 0 0 0 0

MR 1 2 0 0 0 0

607928 0 0 0 0 0 0

Lyamunguz 1 CW 2 3 1 1 0 0
MR 1 2 0 0 0 0

G07928 0 0 0 0 0 0

2 CW 2 3 1 1 0 0

MR 2 2 0 1 0 0

G07928 0 0 0 0 0 0

 

"CW = Canadian Wonder (susceptible), MR = Masai Red (slightly susceptible),
G07928 = breeding line (resistant).

xField used for the study was cultivated to maize the previous year.

yThe field was under coffee/banana association, and no bean crop had been grown in the
field.

2The field used was under fallow the previous 3 years.

44

in stems of standing plants of cultivars Canadian Wonder and Masai Red, whereas race 2
remained viable for 2 and 3 months in foliage and stems of standing plants, respectively.
The two races (1 and 2) were not recovered from debris of either bean genotype after 1
month burial at 2-5 cm and at 25 cm deep (Table 3).

In the field left fallow the previous 3 years at Lyamungu, races 1 and 2 of 2.9. pv.
phaseolicola were recovered after 2 and 3 months in leaves and stems, respectively, of
cultivar Canadian Wonder left standing in the field. In samples placed at 2-5 cm depth, race
1 was viable for 1 month (July) only in Canadian Wonder, while race 2 survived for the
same duration in Canadian Wonder and in stem pieces of Masai Red (Table 3). As in the
coffee/banana intercrop environment, halo blight bacteria were not detected after 1 month
in samples buried at 25 cm beneath the soil surface, the normal plowing depth.

At both geographical locations (Monduli and Lyamungu), the pathogen remained
viable for a longer period in plants left standing in the field than in buried debris (Table
3). Comparing the two locations, however, 2.9. pv. phaseolicola survived for a longer period
at Ngarash, Monduli than at Lyamungu. In dry infected debris stored in the laboratory at
22 i 2°C, the bacterium remained viable for more than 8 months in cultivars Canadian
Wonder and Masai Red. However, in the resistant line, 607928, both races were detected
at very low numbers only at the beginning of the experiment in June; thereafter the

pathogen was not detected.

pH values

During the period of the survival study, the soil pH values at N garash, Monduli
ranged from 6.8-6.9 at both placement depths. In the field under the coffee/banana

intercrop at Lyamungu, pH values ranged from 4.6 in July to 6.2 in September for the

45
topsoil, and from 4.7 in July to 6.3 in September for the 25 cm depth. The field left fallow
the previous 3 years had pH values ranging from 4.9 to 5.0 for the two placement depth
for July and September, respectively. At Lyamungu the low pH in July may be attributed

to ammonium sulfate which is applied as a nitrogen fertilizer for coffee.

Volunteer plants

2.9. pv. phaseolicola was recovered from infected foliage of volunteer bean plants
in two farmers’ fields in Monduli and around several backyards at Lyamungu. Due to high
moisture content in the soil, shatter loss bean seeds germinated almost immediately after
harvest (July/August) at Lyamungu. However at Monduli, where the soil moisture was very
low due to lack of rain, most of the shatter loss bean seeds remained dormant in the field
until October/November when soil moisture conditions allowed germination. Volunteer
bean plants around farmers’ backyards arose from seeds dropped or discarded during

threshing.

Weather data

Rainfall and temperature changes at Monduli and at Lyamungu from June to
December are shown in Fig. 5 and Fig. 6, respectively. Lyamungu received rain throughout
the survival study period (J une-December, 1989). On the Other hand, Monduli remained
dry for most of this period and received little or no rain until November and December
(Fig. 5). The monthly mean maximum temperature at Lyamungu ranged from 20.5 to 26.5°C

and the monthly mean minimum temperature ranged from 12.5 to 15°C.

Figure 5. Monthly rainfall from June to December at Monduli for 1989 and mean monthly
rainfall for 3 years (1986-1988). Numbers above each bar represent number of rain days
during the indicated month.

 

RAINFALL (MM)

R50 1

woo 1

N00 ..

Loo-

 

 

D Ammo

E ><mm>0m Om w <mw

m w

m

w

A

 

p

 

 

S\\\\\V¢O

 

 

 

 

 

_<_OZ._.I

eczema... {coxmmvwooizo

<

_om

m

47

Figure 6. Monthly rainfall and mean monthly maximum and minimum temperatures from
June to December at Lyamungu for 1989 and for 53 years (1935-1988). Numbers above
each bar represent number of rain days during the indicated month.

 

49

RAINFALL (MM)

 

 

 

 

 

moo mo
80... \l :8
I\\\

.50-- l :8
@001: \0 114m
s.iI inovthls

fl|79lll5||l+lllb\l.
moo-. --ao
8
l 3 i
so : a s :3 m
§ Q S we mm Q
.62 . 2? .25 . mmu . 03. 22 , omo
:02?
m>_z_u>_._. amznmmficmm
D 58 ll. :92 z>x§c§ 68

g §m>z Om mm <mm.

oils §m>z _<=Z=<_C_<_ homo

I §m>z _<_>x=<_CZ_ mm <mm.

bulb §m>z Z:Z=<_C_<_ mm <Em.

TEMPERATURE co)

DISCUSSION

The current survey has revealed that in Arusha and Kilimanjaro regions, three races
of 2.9. pv. phaseolicola exist, and race 2 including race 2P occurred at a relatively higher
frequency (52.5%) than race 1 (44.9%) and race 3 (2.5%). These findings agree with
previous work by Taylor 91 91. (1987) who reported the existence of three races of this
pathogen in Tanzania and in other countries, especially around the African Great Lakes
region. Although the frequency of race 3 was very low in northern Tanzania, Taylor _t _l.
(1987) observed that it is the dominant race in the neighboring country of Rwanda. Due
to its high prevalence in northern Tanzania, race 2 appears to be the most threatening race
to bean production in that area. However, the increased movement of bean germplasm
between countries and the need for continued exchange of bean breeding material between
research stations within countries may assist in spreading other races of 2.9. pv. phaseolicola
in areas where they presently occur at low frequencies.

Mthin race 2, there appeared to be a great deal of variation, with some isolates
producing a brown diffusible pigment on NA and KMB. Brown diffusible pigment-
producing isolates were designated as race 2P, and were found at altitudes ranging from
1020 to 1630 m above sea level (Table 1). In addition, this study has revealed that at least
2 out of 3 race 2P isolates were able to grow at a higher temperature (34 i 2°C) than the
non-pigment-producing race 2 isolates or the race 1 and 3 isolates. However, it is not
known whether such isolates may also tolerate high temperatures under natural conditions.
Two important questions also remain unanswered. First, whether this brown diffusible
pigment is also produced i_n 9199 under natural conditions or whether it is a nutritional
artifact of growth i_n_ M. Second, whether the ability to grow at high temperature i v.99;

is associated with pigment production. More studies on the physiology of brown diffusible

50

51
pigment production in race 2 isolates are needed in order to answer these questions.
Phytopathogenic bacteria producing brown diffusible pigments are very rare in nature, and

so far have been known to occur only in Xanthomonas camgtris pv. phaseoli var. fuscans

 

(CMI, 1965).

King e_t 91. (1954) observed that bacteria of the genus fiudomonas produce a
number of pigments on medium A and B, including fluorescein, pyocynin and pyorubin. In
addition, a dark-brown pigment was also produced by some strains, which they speculated
was not an oxidation-reduction product of pyocyanin, fluorescein or pyorubin. Since these
workers did not specify the species or pathovar that produced the dark-brown pigment, it
is unknown whether the brown, diffusible pigment observed in the current investigation is
similar to that reported by King 91 91. (1954).

This work also demonstrated that race 2 isolates of 2.9. pv. phaseolicola isolated
from debris in Monduli were more virulent than those isolated from growing bean plants.
These more virulent isolates produced systemic chlorosis and stunting on the usually
resistant differential bean cultivar, Edmund (Fig. 4). These results suggest that virulence of
race 2 isolates may increase during survival in the bean debris under dry environmental
conditions such as in Monduli. It is not known whether this is due to mutation or selection
of a sub-population whose increased virulence is associated with ability to survive in debris.
Spontaneous mutation involving increased virulence in 2.9. pv. phaseolicola and 2.9. pv.
p_i91 has been observed by some researchers (Taylor e_t a_l., 1989). It is possible that
environmental factors may have a greater effect on the selection of some pathogenicity and
virulence genes than the host cultivars grown (Kiyosawa, 1980). Such a phenomenon has
also been suggested to Occur in Xanthomonas camfltris pv. pm; in Japan (Harino, 1981)

but not in dried debris. Although it is difficult to measure precisely the effect on virulence

52
genes that may be caused by environmental changes, frequent surveys of different ecological
areas in Tanzania will always provide information on the effect of the environment on
virulence of halo blight bacteria.

The finding that the ability of 2.9. pv. phaseolicola to survive in bean debris in the
soil and in standing bean plants in the field varied depending on race, geographical location,
depth of placement in the soil and the bean genotype used, indicate that halo blight disease
management practices should take into account such variations. The survival of the
bacterium was favored by soils with a high pH and a long dry period e.g., in Monduli,
where it survived for 6 months in the susceptible bean cultivar, Canadian Wonder. These
results suggest that debris from highly susceptible cultivars constitutes a source of primary
inoculum for halo blight under field conditions in dry areas. These data agree with previous
reports that the survival of 2.9. pv. phaseolicola is favored by dry soils (Natti, 1967, 1970),
although no pH values were given in these studies. This study did not distinguish the effect
of soil moisture and soil pH on survival of halo blight bacteria. Soil moisture affects the
longevity in the soil of many phytopathogenic bacteria by increasing microbial activity
(Natti, 1970). However, the effect of soil pH on the survival of such bacteria has not been
well studied.

Under high soil moisture conditions and low pH such as was the case at Lyamungu,
survival of halo blight bacteria was considerably reduced, especially in the coffee/banana
environment where soil moisture tended to remain high (Table 3). However, halo blight
diseased bean plants of the susceptible cultivar, Canadian Wonder, when left standing in
the field may act as sources of inoculum when beans are grown in pure stand, since in areas
like Lyamungu, two to three bean crops may be planted every year. However, it is unlikely

that standing bean plants left in the field may act as sources of primary inoculum for beans

53

grown in the coffee/banana intercrop environment. This is because copper-based chemicals
used to control coffee rust (Hamileia vastatrix) may further decrease the ability of 2.9. pv.
phaseolicola to survive under such environments. Copper-based chemicals have been
reported to significantly reduce halo blight bacteria populations in bean fields (Legard and
Schwartz, 1987; Saettler and Potter, 1970; Zaumeyer and Thomas, 1957). Data indicate
that resistant cultivars do not seem to support survival of 2.9. pv. phaseolicola (Table 3).
Such cultivars when grown, will reduce the amount of initial inoculum that may survive in
bean debris in the field.

On the basis of this study, it was observed that volunteer bean plants also provide
an alternative survival site for halo blight bacteria in northern Tanzania, bridging the gap
between the long rain (March-June) and the short rain (October-December) bean crops.
The prevalence of race 2 at a higher frequency in northern Tanzania may be related to its
ability to survive longer than race 1 under the same conditions in the field (Table 3). It will
be interesting to study differences in bean seed infection and transmission between races
of halo blight bacteria to verify whether such differences eidst and relate them to race
prevalence in some regions.

Generally, this study has indicated that there is a great deal of pathogenic variation
of 2.9. pv. phaseolicola in northern Tanzania. In addition to the use of resistant varieties,
control strategies should include sanitation such as burning of infected debris in the field
and around backyards after processing the bean crop, especially in dry areas, like Monduli.
This will reduce the amount of inoculum that can survive. Plowing under volunteer bean
plants when they are still very young will allow faster decomposition and therefore reduce

inoculum between crops in wet areas such as Lyamungu.

LITERATURE CITED

Baggett, JR. and WA. Frazier. 1967. Sources of resistance to halo blight in Phaseolus
vulgaris. Plant Dis. Reptr. 51:661-665.

Buruchara, RA and M.A. Pastor-Corrales. 1981. Variation, virulence of Pseudomonas
gig'ngae pv. phaseolicola on bean in Colombia. pp. 341-351. 19: Lozano, J.C. (Ed.).
Proc. Fifth Int. Conf. Plant Pathogenic Bacteria. August 16-23. CIAT, Cali,
Colombia.

CMI. 1965. Description of plant pathogenic fungi and bacteria. No. 45. Commonwealth
Mycological Institute, Kew, Surrey, England.

Coyne, D.P., M.L Schuster and C. Erwin. 1979. Reaction of 2haseolus vulgaris germplasm
to a new virulent strain of Pseudomonas phaseolicola. Ann. Rept. Bean Improv.
Coop. 22:20-21.

Coyne, D.P., M.L Schuster and R. Fast. 1967. Sources of tolerance and reaction of bean
to races of halo blight bacteria. Plant Dis. Reptr. 51:20-24.

Ekpo, E.J.A. 1975. Pathogenic Variation in Common (Xanthomonas phaseoli) and Fuscous
(_X. phaseoli var. fuscans) Bacteria Blights of Bean (Phaseolus vulgaris L). Ph.D.
Dissertation. Michigan State University, East Lansing. 127pp.

Groth, DE. and EJ. Braun. 1989. survival, seed transmision and epiphytic development of
Xanthomonas camm9tris pv. gycines in the North-Central United States. Plant Dis.
73:326-330.

Hagedorn, D.J., J.C. Walker and R.E. Rand. 1974. Wis. HBR40 and Wis. HBR 72 bean
germplasm. Plant Dis. Reptr. 53:178-181.

Hale, C.N. and JD. Taylor. 1973. Races of Pseudomonas phaseolicola causing halo blight
of beans in New Zealand. N. Z. J. Agric. Res. 167:147-149.

Harper, S., N. Zewdie, I.R. Brown and LW Mansfield. 1987. Histological, physiological and
genetical studies of the response of leaves and pods of Phaseolus vulgaris to three

races of Pseudomonas m’ngae pv. phaseolicola and to fi9udomonas gm'ngae pv.
coronafaciens. Physiol. Mol. Plant Path. 31:153-172.

Horino, O. 1981. A survey of geographical distribution of pathogenic groups of
Xanthomonas camfitris pv. m from Japan in 1977 and 1979. Ann. Phytopath.
Soc. Japan 47:50-57.

Jensen, J.H. and J.E. Livingstone. 1944. Variation in symptoms produced by isolates of
Phflomonas medicaginis var. phaseolicola. Phytopathology 34:471-480.

55

King, E.O., M.K. Ward and DE. Raney. 1954. Two simple media for the demonstration
of pyocyanin and fluorescin. J. Lab. Clin. Med. 44:301-307.

Kinyua, GK. and D. Mukunya. 1981. Variability in isolates of Pseudomonas phaseolicola
(Burk) Young, Dye and Wilkie in Kenya and genetic studies on resistance in dry
food beans. pp 352-357. 19: Lozano, J.C. (Ed.). Proc. Fifth Int. Conf. Plant
Pathogenic Bacteria. August 16-23. CIAT; Cali Colombia.

Kiyosawa, S. 1980. On the virulence analysis of pathogen race frequencies. Ann. Phytopath.
Soc. Japan 46:582-593.

Legard, DE. and H.E Schwartz. 1987. Sources and management of Pseudomonfi svringae
pv. phaseolicola and Pseudomonas syg'ngae pv. sm’ngae epiphytes on dry beans in
Colorado. Phytopathology 77:1503-1509.

Lelliot, R.A., E. Billing and AC. Hayward. 1966. A determinative scheme for the
fluorescent plant pathogenic pseudomonads. J. Appl. Bacterial. 29:470-489.

Lindow, SE. 1988. Lack of correlation of i9 vitro antibiosis with antagonism of ice-
nucleation active bacteria of leaf surfaces by non-ice nucleation active bacteria.
Phytopathology 78:444-450.

Lindow, S.E., D.C. Amy and CD. Upper. 1978. Erwini9 herbicfi, a bacterial ice-nucleus
active in increasing frost injury to corn. Phytopathology 68:523-527.

Lindow, S.E., D.C. Amy and CD. Upper. 1982. Bacterial ice-nucleation, a factor in frost
injury to plants. Plant Physiol. 70:1084-1089.

Natti, J.J. 1970. Control of bacterial diseases of bean. Bean Impr. Coop. 13:20-26.

Natti, J.J. 1967. Overwinter survival of Pseudomonas phaseolicola in New York.
Phytopathology 57:343 (Abstr.).

Patel, RN. and J.C. Walker. 1965. Resistance in Phaseolus vulgaris to halo blight.
Phytopathology 55:889-894.

Saettler, A.W., C.R. Cafati and D.M. Weller. 1986. Non-overwintering of Xanthomonas
bean blight bacteria in Michigan. Plant Dis. 70:285-287.

Saettler, A.W., H.S. Potter. 1970. Chemical control of halo blight in field beans. Research
Report No. 98. MSU Agric. Expt. Sta. Extension Bull. 8pp.

Schaad, NW. 1988. Laboratory Guide for Identification of Plant Pathogenic Bacteria.
Second Edition. APS Press. St. Paul. 164pp.

Schuster, M.L, D.P. Coyne and C. Smith. 1979. New strains of halo blight bacterium in
Nebraska. Ann. Rept. Bean Improv. Coop. 22:19-20.

56

Schwartz, HF 1989. Halo blight. pp 175-182. 29: Schwartz, HF and M.A. Pastor-Corrales
(Eds). Bean Production Problems in the Ttopics. Second Edition. CIAT; Cali,
Colombia. 726pp.

Silbernagel, M.J. and M.J. Mills. 1987. Halo blight of Phaseolus vulgaris. pp. 81-83. 19:
Salema, MP. and Minjas, A.N. (Eds). Bean Research Vol. 3. Papers presented at
a Workshop held at Sokoine University of Agriculture, Morogoro, Tanzania.
September 9-12, 1986. Benedictine Publications. N danda, Peramiho. 167pp.

Taylor, J.D., J .R. Bevan, LR. Crute and S.L Reader. 1989. Genetic relationship between
races of Pseudomonas sm'ngae pv. 99 and cultivars of Pisum sativum. Plant Path.
38:364-375.

 

 

Taylor, J.D., D.M. Taverson and J.H.C. Davis. 1987. Diseases of Phaseolus beans - biology,
resistance and control. pp. 63-64. 19: Inst. Hort. Res. Ann. Rept. Wellsborne,
Warwick, England. 1986/87.

Zaumeyer, WJ. and HR. Thomas. 1957. A Monographic Study of Bean Diseases and
Methods of Their Control. U.S.D.A. Agric. Techn. Bull. No 868. 255pp.

CHAPTER3

THE ROLE OF WEEDS IN SURVIVAL OF PSEUDOMONAS SYRINGAE

PV. PHASEOLICOLA IN NORTI-ERN TANZANIA

57

INTRODUCTION

The role of weeds and non-host plant species as alternate sources of inoculum for
disease epiphytotics has been documented for several plant pathogenic bacteria (Goto,
1972). The importance of cruciferous weeds as a reservoir of inoculum of Xanthomonas
campestris pv. camm9tris causal agent of black rot of crucifers has been reported. The
bacterium was observed to spread at least 12 meters from infected weeds to cabbage
transplants in Georgia (Kuan 91 91., 1986; Morris and Knox-Davies, 1980; Schaad and
Dianese, 1981). The bean bacterial brown spot pathogen, Pseudomonas M99 pv.
mi9g99, highly virulent on beans, survived overwinter on leaves of hairy vetch (M
M9) in Wisconsin. A correlation was established between the presence of high epiphytic
populations of bacteria on this weed and outbreaks of bacterial brown spot in adjacent bean
fields. High numbers of bacteria spread from hairy vetch plants to bean plants during
rainstorms early in the summer (Ercolani e_t 91., 1974).

Saprophytic survival of plant pathogenic bacteria on root surfaces was first revealed
by Valleau e_t 91. (1944). In an extensive survey these workers showed that Pseudomonas
@9193 (2. apt-i_nga_e_ pv. Lad) overwintered in close association with the roots of several
crops and weeds. Xanthomonas camfltris pv. vesicatoria is another bacterium which
Valleau 91 91. (1944) revealed to exist in association with roots of wheat under natural
conditions, but they were unable to prove the same for 2. M99 pv. phaseolicola and

Xanthomonas camfltris pv. phaseoli var. sojense. Laub and Stall (1967), and Kishun and

58

59
Sohi (1979), also observed that Xanthomonas camfitris pv. vesicatoria survived on leaves

of weeds, Solarium nigr_um and Phjpalis minima.

 

Smith (1962) detected Xanthomonas camgtris pv. malvacearum in roots and leaves
of various weeds collected from blighted cotton fields by inoculating young cotton seedlings.
Kikumoto and Sakamoto (1969) found that the growth of soft rot E_rw1_'919 was stimulated
in the rhizospheres of weeds which were common in fields of chinese cabbage. These
weeds included _Agrofls; grennans, Portulaca olerace99, Chenopo_dium 91mg, and
Commelina communis. The effect of soil from the rhizospheres of weeds on the growth
of Pseudomonas solanacearum has also been studied (Goto, 1972). The population of this
pathogen in artificially infested soils collected from weed rhizospheres varied considerably
depending on the kind of weed plants. This suggests that differences in quality and
quantity of nutrients secreted from the roots of these plants can influence the pathogen by
stimulating or depressing the rnicroflora or by depressing or stimulating the pathogen.

In many parts of Tanzania, one crop season is well separated from the next by a
dry season of varying duration depending on location and year. Most plant pathogens,
therefore, are faced with a problem of survival between seasons because of the discontinuity
of host crop plant populations. Weeds and wild plants may close this gap and thus
contribute to outbreaks of plant diseases by bridging between seasons, between crops, and
between locations. Weeds also constitute a reservoir of minor diseases which may be
exchanged between them and cultivated crops (Dinoor, 1974). Moreover, there is no way
yet to predict when and how a minor disease of a wild plant or weed will spread into
cultivated crops and become a major problem.

Epiphytic survival of 2.9. pv. phaseolicola was first reported by Ercolani 91 91. (1974).

Their studies indicated that the bacterium multiplied to some extent on hairy vetch (Vicia

60
v_i_llo_sa), a leguminous weed growing near bean fields under Wisconsin conditions.
Recently, Taylor 91 91. (1987) obtained several isolates of 2.9. pv. phaseolicola from
Neonotonia wr_°g_l_it_ii (Arn.) leakey in South America and in Africa, including Tanzania.
While efforts are being made to control halo blight disease in the bean crop, the disease
is not under control in weeds or wild plants. No studies have been conducted in Tanzania
to determine the role of weeds and non-host species in survival and as sources of inoculum
for halo blight bacteria. More information is therefore needed to predict the impact of 2.9.
pv. phaseolicola inoculum that develops and survives on weeds. With this view in mind, a
survey was conducted to investigate the role of weeds in survival and as sources of

inoculum for halo blight bacteria in northern Tanzania, which is the focus of this chapter.

MATERIALS AND METHODS

Sampling and isolation procedures

Isolations of 2.9. pv. phaseolicola were attempted from weed plants with and without
leaf lesions. These weed plants were collected from the bean growing areas of the Arusha
and Kilimanjaro regions, northern Tanzania, from November 1988 to January 1990. Weed
collections were made where halo blight was reported to be severe the previous growing
season. In addition, samples were collected from hedges around backyards of farmers’
houses, roadsides, and on hedges which commonly grow along irrigation channels. Weed
samples were placed in plastic bags and sent to the laboratory where they were kept c0ld
at 5°C until processed within 24-48 hours.

Before processing in the laboratory, samples were separated into leaves, stems, and
roots. Tan to 15 g of each of these weed plant components were weighed and placed in

500 ml flasks containing 100-300 ml of 0.01 M phosphate buffer, pH 7.2, with 0.01 percent

61

Tween 20. Samples were shaken for 30 minutes on a wrist action shaker (Staurt Scientific
Co. Ltd, England) adjusted at medium speed at 22-25°C. The supernatant liquids were
10-fold serially diluted four times in the same buffer and 0.01 ml portions from each
dilution, including the original undiluted liquid, were plated on Medium B of King e_t 91.
(1954) (KMB) supplemented with 100 ,ug cycloheximide per ml to prevent fungal growth.
Inoculated plates were incubated at 24 i 2°C for up to 5 days. Plates were examined
under short- (254 nm) and long-wave (366 nm) ultraviolet light. Fluorescing colonies were
transferred and further purified on KMB using a series of single-colony transfers. Isolation
was also attempted from weed leaves that contained lesions. Pieces of diseased tissue
including small areas of adjoining healthy tissue were excised, surface-sterilized with 2.6%
NaOCl for three minutes, and rinsed twice in sterile glass distilled water. The tissue pieces
were then transferred onto flame-sterilized glass slides and crushed in one or two drops
of sterile distilled water. The resulting suspensions were streaked on KMB-cycloheximide
plates and incubated at 24 :2 2°C. Bacterial colonies resembling 2.9. pv. phaseolicola were
purified as described above.

Storage 'of isolates

After purification, isolates were maintained as very thick suspensions in glycerol:
0.1M phosphate buffer (50:50, v/v) and on nutrient agar slants at 0°C in an incubator (A.
Gallenkamp and Co., London). Isolates were also stored in dry bean leaf powder at room
temperature (22 i 2°C) in darkness. Tan-day old bean seedlings, variety Canadian Wonder,
grown in a screen house were spray-inoculated abaxially with suspensions of pure cultures
of 2.9. pv. phaseolicola isolates. Inoculated plants were maintained in the screen house at

19-27°C. Leaves with typical halo blight symptoms were harvested 10-14 days after

62
inoculation and air-dried at 22 i 4°C. After drying leaves were ground to a fine powder
using sterile mortars and pestles, and the powder containing each bacterial isolate was

stored in sterile glass vials.

Sterilization and gg'nding of samples for detection of Rs. pg. phaseolicola
Samples were surface sterilized using 2.6% NaOCl for 2-3 minutes and rinsed in

sterile glass distilled water. Samples were ground using sterile mortars and pestles, then
mixed with the required volume of sterile 0.01 M phosphate buffer, pH 7.2, unless
otherwise stated. The resulting suspensions were left to stand for 20-30 minutes followed
by ten-fold serial dilutions in the same buffer, and 0.1 ml portions were plated on KMB

or other specified media.

Identification of bacterial isolates

Identification procedures were based on Schaad’s manual (1988) and those given by
Hayward (1985). Isolates were observed for production of green diffusible fluorescent
pigment on KMB (King 91 a_l., 1954) using short- (254 um) and long- (366 nm) wavelength
ultra-violet light. All isolates were tested for levan production, oxidase reaction, pectolytic
activity, arginine dihydrolase activity and tobacco hypersensitivity (LOPAT) (Lelliot 91 91.,
1966). For levan production, 24-hour old bacterial cells grown on KMB were streaked on
NA containing 5% sucrose (w/v) and incubated for up to 5 days. Three replicates were
used for each isolate, and NA plates without sucrose were used as controls. The
appearance of domed mucoid colonies after 3-5 days of incubation was considered positive

for levan production.

63

Oxidase reaction (Kovacs, 1956) was tested by smearing cells from a 24-hour KMB
plate cultures of each isolate with a platinum wire loop on a filter paper containing an
oxidase spot reagent (Difco). Three tests were conducted for each isolate. The appearance
of a dark red color within 60 seconds was considered to be positive. Pectotytic enzymes
activity was detected by pit formation on polypectate containing media, using the procedures
described by Hildebrand and Schroth (1971). Bacterial cells from 24-hour old plate cultures
were spot-inoculated in triplicate on plates of sodium polypectate medium (1 ml 1.5%
alcoholic bromthymol blue, 6 ml 10% CaC12-2H20 and 22 g sodium polypectate in 1000 ml
glass distilled water; 100 ml of 4% molten sterile agar was added after autoclaving. The
pH was adjusted to 5.0 and 8.0 (Olive and McCarter, 1988) using HCl and NaOH).
Inoculated plates were incubated for 6-7 days. High and low pH values were used to
determine the ability of species to grow at the two specific pH values.

Arginine dihydrolase activity was tested for using Thornley’s medium 2A (Thornley,
1960) (g/lzpeptone, 1.0; NaCl, 5.0; KzHPO‘, 0.3; agar, 3.0; phenol red, 0.01; arginine HCl,
10.0); pH 7.2. Tan ml of sterile medium in test tubes were stab-inoculated with bacterial
cells, 24 hours old. After inoculation, the surface of the medium in each test tube was
layered with 3 ml of 3% sterile molten agar (Fahy & Hayward, 1983) and incubated for up
to five days. A color change to red (alkaline) was considered positive. Each bacterial
isolate was replicated three times and test tubes stab-inoculated with a sterile loop wire
were included as negative controls.

The ability of bacterial isolates to induce a hypersensitive reaction (HR) on tobacco
(Nicotiana m L) was tested (Lelliot and Stead, 1987; Klement e_t 91., 1964). Tbbacco
plants were inoculated with suspensions of each bacterial isolate by infiltrating the cell

suspensions into leaves. Infiltration was done abaxially by pressing the end of a sterile 10

64
cc disposable hypodermic syringe against forefinger-supported tobacco leaves and slowly
introducing the bacterial suspensions. Each isolate was replicated twice. 2.9. pv.
phaseolicola isolates 1299A and 1301A supplied by Dr. J .D. Taylor (Institute of
Horticultural Research, Wellesbourne, Warwick, England), and sterile phosphate buffer were
used as positive and negative controls, respectively. Inoculated tobacco plants were kept

in the screen house and evaluated for HR after 24-48 hours.

Pathogenicig tests

All isolates were evaluated for pathogenicity on susceptible Canadian Wonder bean
plants. Screen house-grown bean plants were inoculated when the first trifoliolate leaf was
fully open (10-14 days old) by spraying abaxially with a hand operated plastic atomizer
(Bayer Chemical Co., Nairobi, Kenya) to run-off without water-soaking. Each isolate was
replicated three times using one bean plant per replicate. A known pathogenic 2.9. pv.
phaseolicola isolate (1299A) and sterile phosphate buffer were included as positive and
negative controls, respectively. Inoculated bean plants were incubated in the screen house
and observed daily for symptom development. Pathogenicity was evaluated 10-14 days after
inoculation. Appearance of typical halo blight symptoms confirmed 2.9. pv. phaseolicola

identity.

Ice nucleation activi1y

Isolates were tested for ice nucleation activity (INA) using the single temperature
droplet freezing method of Lindow (1988) and Lindow 91 91. (1978, 1982) at -5°C.
Bacterial suspensions from cultures grown for 2-3 days were prepared in 0.1 M phosphate

buffer, pH 7.0, and adjusted to an optical density of 0.1 at 620 nm using a

65
spectrophotometer. Suspensions were ten-fold serially diluted three times in the same
buffer. INA was determined by placing 30 ten-pl droplets from each of the dilutions,
including the original suspensions, on paraffin-coated aluminum weighing boats floating on
50% ethanol maintained at -5°C in the refrigerator (Lee Refiigeration PLC, Bognor Regis,
West Sussex, England). The freezing of each individual droplet was determined visually at
one minute time intervals up to three minutes. Droplets from blank phosphate buffer

dilutions were included as negative controls.

Carbon source utilization

Stock solutions of mannitol, sorbitol and inositol (Sigma Chemical Co., St. Louis,
M0) were prepared in glass distilled water following the procedures given by Schaad
(1988). Stock solutions were filter sterilized using 0.22 pm pore size membrane filters
(Millipore Product Division, Bedford, MA 01730) attached to a 60 ml hypodermic syringe,
and added at 0.1% (w/v) final concentration to autoclaved and cooled (45-50°C) Dye’s
medium C (g/l of glass-distilled water: NH‘HZPO‘, 0.5; KZHPO‘, 0.5; MgSO,.-7HZO, 0.2;
NaCl, 5.0; yeast extract, 1.0; Bacto-agar, 15.0). Bromocresol purple (1 ml of a 1.6% alcohol
solution) was added before autoclaving. ‘Log phase bacterial cultures grown on KMB were
streaked on the medium in triplicate and plates were incubated for up to 10 days. Growth

was compared to plates containing no added carbon sources.

Bacteriophage tests
Sensitivity to bacteriophage was tested using the methods described by Taylor (1970)

and J .D. Taylor and D.M. Taverson (personal communication) as described in Chapter 2.

Serology

2.9. pv. phaseolicola antiserum was kindly provided by Dr. J .D. Taylor. The
agglutination method of J .D. Taylor and D.M. Taverson (personal communication) described
in Chapter 2 was used. Three tests were conducted for each bacterial isolate, and known

2.9. pv. phaseolicola isolates 1299A and 1375A were included as positive controls.

Race identification

After species identity was confirmed, isolates of 2.9. pv. Phaseolicola obtained from
weeds were subjected to race identification using differential bean cultivars Canadian
Wonder (universal susceptible), Edmund (universal resistant), Red Mexican UI3 (resistant
to race 1) and Tendergreen (resistant to race 3). Both pod and plant leaf reactions were

used for race determination as descn’bed earlier in Chapter 2.

Assay of Neonotonia wightii reproductive parts for Rs. m. phaseolicola infection

Flower buds, open flowers and young flat pods of infected Neonotonia v_ngm plants
were collected from Monduli district, Arusha region, and from Hai district, Kilimanjaro
region, Tanzania, during 1989 and 1990. Samples were collected from reproductive
branches containing severely infected leaves. Plants from each site were marked so that
periodic sampling could be done on the same plants. Samples from each site were placed
in separate plastic bags to allow the correlation of any seed infection with weather
conditions, which were recorded for each site. After collection, samples were sent to the

laboratory and kept at 5°C until processed within 24-28 hours.

67

For each sampling, 10 flower buds and 10 open flowers were weighed separately,
surface-sterilized, rinsed, and ground in 10 ml of sterile phosphate buffer. Aliquots of 0.1
ml were plated on KMB-cycloheximide and incubated as earlier described for 5 days. To
assay for the possibility of 2.9. pv. phaseolicola infection in young flat pods, 10 pods for
each sampling time from each site were weighed, surface sterilized, rinsed, comminuted and
plated as above. Plates were incubated and observations made daily to detect the presence

of fluorescing colonies.

Neonotoniflightii seed infection and transmission assay

Mature dry pods from Neonotonia 11gh_t11 plants collected from selected sites were
placed in plastic bags and sent to the laboratory where they were air-dried. After drying
pods were shelled manually by hand; both seeds and shells were assayed for the presence
of 2.9. pv. phaseolicola. Transmission of 2.9. pv. phaseolicola in seed was assayed using the
method of Schuster and Coyne (1975). N. m seeds were surface-sterilized, rinsed three
times and blotted dry on sterile filter papers in a laminar flow chamber. Drying was
necessary to remove the sticky character of seeds when wet which made them difficult to
handle. After drying, seed lots from each site were transferred into sterile test tubes
containing 10 ml of phosphate bufl‘er with 0.01% T1veen 20. Samples were shaken for 30
minutes on a wrist action shaker, after which seeds were removed from test tubes aseptically
and divided into two lots. One lot was dried aseptically, plated hilum down on KMB-
cycloheximide agar and incubated for 5 days. Plates containing bacterial growth around

seeds were observed under UV-light to detect any fluorescing bacterial colonies.

68

Fluid in which seeds were soaked was diluted and plated on the same medium.
Portions of the seed soakings from each seed lot were also used to inoculate five, 10-15 day
old Canadian Wonder bean seedlings by injection and leaf water-soaking methods. In the
latter, inoculum was introduced into four infiltration sites on primary and fully open first
trifoliolate leaves, and each plant was considered a replicate. Inoculated bean plants were
maintained in the screen house and observed for halo blight symptoms for up to 14 days.
Control plants were inoculated with dilute suspensions of 2.9. pv. phaseolicola (103-10‘
CFU/ml) to simulate the low amount of inoculum which may occur in the soakings, and
negative control plants were inoculated with sterile phosphate buffer.

Another seed lot from each test tube was planted in the screen house. Individual
plants that germinated were examined for the presence of water-soaked lesions on stems
and leaves. Starting from the primary leaf stage, 5 plants were sampled at random at 4 day
intervals for three weeks to further assay for the presence of 2.9. pv. phaseolicola. Sampled
plants were separated into leaves, stems, and roots. The separate plant components were
weighed, surface sterilized, rinsed, and ground in phosphate buffer. The suspensions were
serially diluted and plated on KMB-cycloheximide. Plates were observed for up to 5 days.

Shells from mature pods of 11. m were assayed for presence of 2.9. pv.
phaseolicola. From each site and for each sampling date, 3 g of pod shells were surface-
sterilized, ground and mixed with phosphate bufier. Dilutions of suspensions were plated

and observed for the presence of 2.9. pv. phaseolicola-like colonies for up to 5 days.

_99ction of Neonotonia wightii to races 1, 2 and 3 of Rs. m. phaseolicola

 

Tb test the reaction of N. wightii to the prevailing three races of 2.9. pv.

phaseolicola in the region, uninfected plants were identified from three locations,

69

Lyamungu, Lambo and Monduli, representing different weather conditions. Mature pods-
were harvested, placed in labelled plastic bags and air-dried under laboratory conditions at
22-27°C. Dry pods were then manually shelled by hand and seeds from each site were kept
separate. Tb check for the absence of internal 2.9. pv. phaseolicola infection, seeds from
each location were surface-sterilized, mixed with 10 ml of phosphate buffer containing
0.01% Taveen 20 and shaken for 30 minutes. Seeds were removed from test tubes and the
resulting soakings were diluted and plated as described. Observations were made daily for
the presence of 2.9. pv. phaseolicola-like colonies for 5 days.

One hundred surface-sterilized pathogen-free seeds from each site were planted in
16 cm diameter sterile plastic pots, 20 seeds per pot. The experimental design was a split-
split plot with five replicates, with location as whole plots, plant age as subplots and 2.9.
pv. phaseolicola races as sub-subplots. Three locations, two plant ages (9 and 15 days) and
three isolates for each race were used. Tivo of the race 1 isolates (41, 83) were obtained
from Phaseolus y9lggi9, the third race 1 isolate (1281A) was obtained from 2. coccineus
and was supplied by Dr. J .D. Taylor. All race 2 isolates (7, 42, 106) were obtained from
2. v_9lga_r19, and isolate 106 produced a brown pigment. All race 3 isolates (46, 89, 1301A)
were obtained from 2. m.

Inoculum for each isolate was produced as described above. N. M plants, 9 and
15 days old, were spray-inoculated to run-off without water-soaking, using a hand operated
atomizer. Control plants were spray-inoculated with race 1 isolated from N. _W_ig_h_tl_i plants
and sterile phosphate buffer. Inoculated plants were kept in the screen house and observed

daily for symptom development for 15 days. The experiment was repeated twice.

70
Insects as pgts of beans and N. wightii

Many Neonotonia 9igh_t_ii plants which were infected by 2.9. pv. phaseolicola at
Lyamungu Agricultural Research Station, were highly perforated by leaf chewing insects.
Pod borers were also evident on dry pods. Therefore, efforts were made to determine the
possibility of such insects attacking beans, thus disseminating halo blight bacteria to beans
and _N. w_igh_t9. Dry pods of 9. 91m damaged by insects were collected at Lyamungu
Agricultural Research Station and stored in 500 ml beakers covered with a nylon cloth, in
the laboratory at 20-24°C. After 25-30 days, pods were shelled manually and adult weevils
of an unidentified bruchids were collected and transferred to another container with sound
dry pods of E. 9igl_i_t9 where they were maintained. Weevils were then tested on bean
cultivars Canadian Wonder, Masai Red and Kiburu (local); 91. 9igl_1t_ii seeds were included
as controls.

The method of Giga and Smith (1987) was used to test the susceptibility of bean
seeds to bruchids from 11. 9n_'gl_1t_ii. Seeds were first sterilized by freezing for 10 days to kill
eggs or insects present, then were equilibrated for about 3 weeks at 20 to 24°C. One
hundred seeds of each test cultivar and about 300 seeds of M 91_g199 seeds were placed in
sterile 250 ml flasks. Five weevils (2 males, 3 females) were placed in each flask and
covered with finely perforated nylon cloth for aeration. Each test sample was replicated
three times and kept in the laboratory at temperatures mentioned above, and observed
periodically. After the introduced insects had died, samples were kept for a period of 28
days to allow any hatched eggs to develop. When emergence holes were observed on the
control samples, the number of adult bruchids in each test sample were recorded. Samples
were considered susceptible when the weevil count was higher than the original number in

each flask. For leaf chewing insects, highly perforated plants were frequently visited at

71
different times of the day. The aim of such visits was to collect insects found on these
plants and cage them on bean plants to find out possibilities that such insects could be bean
pests, thus disseminate halo blight bacteria between N, 9igl99 and beans. Visits were

conducted for a period of four months beginning September, 1989.

RESULTS

A total of 17 weed species belonging to 10 families were collected and assayed for
the presence of 2.9. pv. phaseolicola for the period of November 1988 to January 1990
(Table 1). Only one weed species, Neonotonia 9ig9t9 (formerly M 9ig9t9), belonging
to the family Leguminosae was found to be a natural host of halo blight bacteria.
Symptoms were restricted to leaves. Characteristic leaf symptoms initially appeared as small,
water-soaked spots on the undersides of young leaves. Later, a greenish-yellow halo zone
developed around the water-soaked area (Fig. 1). Symptoms were more evident on young
leaves a few weeks after the onset of the short rains (October to December). During the
dry season, halo development around necrotic areas was very restricted, and sometimes did
not occur.

2.9. pv. phaseolicola was isolated from _N_. 9igl_119 samples collected in Kilimanjaro
and Arusha regions throughout the year because the weed survives dry seasons, mainly on
fences (Fig. 2), hedge rows, by the roadsides, on ditch banks near bean fields and in corners
of the farmers’ fields. In some fields, halo blight symptoms occurred only on the foliage
and pods of bean plants grown near 9. 9igl9ii. They occurred especially on the leeward
side of the hedges, indicating that 2.9. pv. phaseolicola was disseminated from infected

weeds to the bean crop during rains.

72

Table 1. Weed families and species assayed for presence of Pseudomonas m’ngae pv.
phaseolicola in northern Tanzania.

 

Family/Weed species Location Symptoms

 

1. Asteraceae (Compositae)

Bidens pilosa L Kilacha/Lyamungu/ Leaf necrosis &
Monduli chlorosis
Bothriocline 1&9 N.E.Br: Lyamungu Marginal necrosis
(Erlangea laxa)(N.E.Br.)S.Moore oi stems & hairs
Conm sumatrensis (Retz) , Miwaleni/Lyamungu None
EH. Walker
Galinsoga parviflora Cav. Lyamungu/Monduli/ None
Sanya J uu

2. Boraginaceae

 

Ti‘ichodesma z_eylanicum Lambo/Monduli Necrotic lesions
(Burn.f.) R.Br. Kirua/Iangeru on leaves
3. Commelinaceae I
Commelina benghalensis L. Lambo/Monduli None
4. Cyperaceae (Sedges)
_C_3yp9rus lotundus L Lyamungu Stem & leaf
necrosis, water
soaking chlorosis

5. Leguminosae
Neonotonia wightii (Wight & Arn.) Lambo/Monduli/Sanya Juu/ Leaf necrotic
Lackey: (Glycine wightii) Lyamungu/Tengeru lesions, water
soaking & halo

Pueraria sp. Lyamungu Necrotic lesions
on leaves

6. Malvaceae
m 9lb_a L Lambo Stem & leaf
necrosis, water

soaldngchlorus's

Table 1. (cont)

73

 

 

Family/Weed species Location Symptoms
7. Nyctaginaceae
Boerhavia erecta L Lambo/Miwaleni None
8. Poaceae (Grasses)
Panicum sp. Lambo/Selian None
Setaria sp. Kibosho/Lyamungu Leaf necrotic
lesions
9. Polygonaceae
mgonum sinuatum (Meisn.) Lambo/Monduli None
Dammer
10. Solanaceae
Nicandra phy9alodes (L) Gaertn Lambo/Narumu/Sanya J uu/ Leaf necrosis &
Monduli/Ibngeru halo
Solarium incunum L Sanya Juu/Lyamungu Leaf necrotic

 

 

lsions, chlorosis

74

 

Figure 1. Typical halo blight symptoms on Neonotonia wightii leaves.

75

 

Figure 2. Halo blight-infected Neonotonia wightii plants surviving the dry season on a fence
at Lyamungu Agricultural Research Station, Tanzania.

76

Identification of bacterial isolates

Twenty seven isolates of fluorescent pseudomonads, of which 22 were obtained from
H. 9igl_it_ii and 5 from Bothriocline 199, were characterized. All isolates were fluorescent
under both short- (254 nm) and long- (366 nm) wavelength ultraviolet light. All 22 isolates
from 2. 9igh_t9 were positive for levan sucrase; negative for oxidase, arginine dihydrolase,
INA, and pectolytic activity; and produced a rapid hypersensitive reaction on tobacco leaves.
These isolates also did not utilize mannitol, sorbitol, and inositol as carbon sources, they
were sensitive to bacteriOphages 11P, 12F and 48P as evidenced by plaque formation, and
agglutinated with 2.9. pv. phaseolicola antiserum, confirming 2.9. pv. phaseolicola identity.
Bacterial isolates from 2. lax_a were positive for INA, oxidase, arginine dihydrolase, and 2
of the five isolates utilized the three carbon sources. The five isolates were negative for
levan production and, pectolytic activity; they did not produce a hypersensitive reaction on
tobacco, and were insensitive to the three bacteriophages used. These data indicated that

the five isolates from 2. laxa were 2. fluorescens.

Pathogenicig tests

Identity of the 27 bacterial isolates was confirmed by pathogenicity tests. All isolates
from E. w_ig11t_ii were pathogenic on susceptible Canadian Wonder bean plants, whereas
those from 2. 1999 were not. Water-soaked lesions developed within 3-4 days of inoculation
and typical halo blight symptoms developed 7-10 days after inoculation. Systemic chlorosis

was also evident in young leaves.

77

Race identification

Leaf and pod reactions of bean differential cultivars to the 22 bacterial isolates
obtained from _N_. _wjg9t_ii are shown in Table 2. On the basis of these results all isolates
were typical of race 1 as shown by a resistant reaction on Red Mexican UI-3. All
compatible reactions involved the development of large water-soaked lesions, especially on
Tandergreen, and the production of bacterial ooze at the point of inoculation on the stem.
Inoculation by leaf spraying and pod inoculation gave similar results. Rapid hypersensitive
reactions were observed in leaves and pods of the cultivar Edmund as indicated by tissue
browning at the inoculation point. The virulence of 10 race 1 isolates from 191. 9igh_t9 was
compared with 10 race 1 isolates from bean sources on the differential varieties. There was
no difference in incubation period and symptom development between the two groups of

race 1 isolates.

Assay of N. wightii reproductive parts for infection by Rs. py. phaseolicola

Throughout 7 months (August to December, 1989 and January to February, 1990)
of testing, no flower bud, open flower, or flat pod samples were found infected with 2.9.
pv. phaseolicola, despite severe foliage infection of sampled plants. Similar results were
obtained for all three locations studied, which also represented different weather conditions
(Chapter 2). Even random samples of reproductive parts collected within 5-10 cm of
highly diseased shoots were free of any detectable levels of the pathogen. Moreover,
symptoms were not seen on any reproductive part examined for samples collected from all
locations. It was also observed that pods on E. m plants were very hairy, a factor
which may prevent the inoculum from reaching the pod surface, but this was not true for

flower buds or parts of open flowers. The restriction of symptoms to foliage and lack of

78

Table 2. Race identification of Pseudomonas 9m'ngae pv. phaseolicola strains isolated from
Neonotonia wightii in northern Tanzania.
Bean differential cultivar“

 

 

 

CW RM TG ED
No. of
Location isolates Leaf/Pod Leaf/Pod Leaf/Pod Leaf/Pod Race
Lambo 3 +/+b -/- +/+ -/- 1
Lyamungu 11 +/+ -/- +/+ -/- 1
Machame 1 +/+ -/- +/+ -/- 1
Monduli 6 +/+ -/- +/+ -/- 1
Tangeru 1 +/+ -/- +/+ -/— 1

 

'lCW = Canadian Wonder RM = Red Mexican UI-3
TG = Tandergreen ED = Edmund

b+ = Susceptible reaction; - = Resistant reaction

Data were compiled from two experiments

 

 

b1

79
detection of halo blight bacteria in flower buds, open flowers, and flat pods indicate that
2.9. pv. phaseolicola was not able to move systemically in the vascular tissue to these

reproductive tissues under difi'erent weather conditions that existed in these locations.

Neonotonia wightii seed infection 9nd transmission assay

To determine whether halo blight bacteria were present in mature seeds of E.
91gl9ii, samples were taken monthly from three locations (Lambo, Lyamungu, and Monduli),
representing differing weather conditions. All mature pods harvested from 1!. 9igh_t9 plants
with severe foliage symptoms of halo blight were symptomless. Seeds from these pods were
also symptomless. Direct seed plating, plating of seed washings, and of shell pod
suspensions all failed to detect the pathogen. Canadian Wonder bean plants inoculated
with seed soakings did not develop visible halo blight symptoms unlike positive control
plants which showed typical halo blight symptoms 12 days after inoculation. However, seeds
at all three locations studied were infected with other bacteria. The dominant colony color
of such bacteria was creamy-white; a few colonies were orange colored. The percentages
of seed infected with these unidentified bacteria for each sampling period at different
locations are shown in Table 3. None of the colonies observed fluoresced under UV-light
and none were tested for pathogenicity.

In the screen house experiments, a total of 2860 seedlings were assayed for infection
by Psp. None of the seedlings of E. ml99 from seeds collected from plants naturally
infected with halo blight bacteria showed any visible symptoms on stems or leaves. The

bacterium was also never recovered from root, stem, and leaf samples from such seedlings.

Table 3. Percentage of Neonotonia wightii seed infected with Pseudomonas w’ngae pv.
phaseolicola and other bacteria in northern Tanzania.

% Seed infection at'

 

 

 

Mean # Lambo Lyamungu Monduli
Sampling of seeds
date assayed Pspb Others Psp Others Psp Others
August 1989 652 0.0 55.7 0.0 75.9 0.0 62.2
Sept 1989 596 0.0 20.1 0.0 52.7 0.0 13.4
Nov 1989 990 0.0 37.0 0.0 31.1 0.0 64.2
Dec 1989 495 0.0 41.3 0.0 39.4 0.0 85.0

 

aFor each sampling date, values are means of three replicates. Seeds were surface sterilized
in 2.6% NaOCl for three minutes and plated on KMB supplemented with cycloheximide
(100 mg/ml), and incubated for 5 days at 24 i 2°C.

I’Psp = Pseudomonas 9yg'ngae pv. phaseolicola

81

 

Table 4 presents the reaction of 111. 9igl9ii seedlings to races 1, 2, and 3 of 2.9. pv.
phaseolicola under screen house conditions. Pathogen-free seeds used for the study were
collected from three locations in northern Tanzania, representing different weather
conditions. All three races produced susceptible reactions in plants inoculated at 9 and 15
days old. The development of halo blight symptoms was observed daily after foliage
inoculation. There was no difference in incubation period and the kind of symptoms
produced on these seedlings for all three races. Typical halo blight symptoms appeared 8
days after inoculation at 19-27°C; results were consistent for the two experiments conducted.
Stunting of plants, chlorosis of young leaves, and presence of yellowish-green halos around
water-soaked and necrotic lesions were evidence of toxin production by isolates of all three

races tested.

Susceptibilig of three bean cultivars to insect fits of N. wim
An unidentified bruchid from 2. wigl_itii did not attack the bean cultivars: Canadian

Wonder, Masai Red, and Kiburu (a local landrace). In contrast, there was a 60% increase
in insect count on E. 9igh_t9 (Table 5). This may indicate that female bruchid weevils were
either not able to lay eggs on the tested bean cultivars, or eggs were laid and failed to
hatch. Despite frequent visits to insect-damaged and halo blight-infected 91. 9ig199 plants
to search for leaf chewing insects, no such insects were observed on the plants. It was,
therefore, suspected that the damage may be caused by insects which could be nocturnal

feeders.

Table 4. Reaction of Neonotonia wightii to races 1, 2, and 3 of Pseudomonas _syg'ngae pv.
phaseolicola from northern Tanzania.

Seedling reaction for seed from
Plant age' Isolate

 

 

(days) numbers Race Lambo Lyamungu Monduli
9 41,83,1281A 1 +b + +
7,42,106 2 + + +
46,89,1301A 3 + + +
15 41,83,1281A 1 + + +
7,42,106 2 + + +
46,89,1301A 3 + + +

 

8Age at which Neonotonia wightii seedlings were inoculated. Inoculation was done as
explained in the text. Two experiments were conducted for each plant age and each
treatment was replicated five times, each replication consisting of 20 plants.

b+ = Susceptible reaction.

Table 5. Susceptibility of seed of three bean cultivars grown in northern Tanzania, and
Neonotonia wightii seed, to bruchid weevils.

 

No. of insects No. of newly emerged
Cultivar introduced' insects after 28 daysb
Canadian Wonder 5 0
Masai Red 5 0
Kiburu 5 0
li- mm 5 3

 

°Five insects (2 males/3 females) were introduced for each treatment.

bInsect counts are means of three replicates. Each replicate consisted of 100 bean seeds
and 300 191. wightii seeds. Counts were taken 28 days after the introduced insects had died.

DISCUSSION

It has generally been considered that halo blight infected bean seed is the most
important source of 2. 9. pv. phaseolicola inoculum, especially given that pathogen-free seed
programs are lacking in Tanzania, and farmers retain portions of their harvest for the next
planting (Gondwe, 1987). The current study indicated that at least one weed species, 19.
9igh_t9 can also serve as a perennial reservoir of 2. 9. pv. phaseolicola under northern
Tanzania conditions. Halo blight bacteria were detected in this weed throughout the year,
even during dry periods when beans were not growing in the field. These results suggest
that 9. 9igl9ii provides a suitable habitat for long term survival and multiplication of 2. 9.
phaseolicola over much of the year.

The pathogen was readily disseminated from the weed to neighboring bean plants
under different weather conditions. These findings may account for occasional outbreaks
of bean halo blight which have been reported in the Kilimanjaro region. Since some of the
farmers’ bean crops were free of halo blight disease, it is suggested that infected seed is
not always the source of initial inoculum in the region. Furthermore, halo blight was
observed only at the edges of the fields in some areas in the region, another indication that
E. 91_'gl9ii is an important source of supplementary initial inoculum for this pathogen. B.
Gondwe, (personal communication) and farmers in northern Tanzania have noted that
halo blight outbreaks were few in the region in years of little rain. Such observations have
led farmers in Monduli to believe that halo blight is not a disease but a stress caused by
excess rainfall. Gondwe (1985) speculated that occasional outbreaks of halo blight disease
in the Kilimanjaro region were due to the occurrence of a new virulent race, probably race
3. However, in an extensive survey of the prevailing races in northern Tanzania, race 3 was

not recovered in the Kilimanjaro region (Chapter 2).

84

 

and
volu
area
C0111
zero
furtl.
repo
(Star

in dr

Do: all;

“Edam,

85

Since N. 9igl9ii is a common weed on fences and hedges around farmers’ backyards,
and most farmers thresh their bean crop in backyards, infected bean debris and diseased
volunteer plants may be a source of primary inoculum for N. 9igl99 weeds around such
areas. In addition, the bacterium is also likely to be spread from prunings from hedges
containing diseased weeds. N. 9igl_1_t9 is also used as a feed for animals kept under the
zero grazing system commonly employed in the Kilimanjaro region, a factor which may
further assist in disseminating the bacterium. Moreover, halo blight bacteria have been
reported to survive for 9 months after passage through sheep under Australian conditions
(Starr and Kercher, 1969). Such may also be the case in the Kilimanjaro region, especially
in dry areas.

The finding that only race 1 of 2. 9. pv. phaseolicola was recovered from N. 9111119
throughout the period of this study is surprising, but agrees with those reported by CIAT
(1987) and in general by those reported by Taylor 91 91. (1987). Since both bean and N.
9igl9ii are susceptible to race 1 of the pathogen, the original host of this race is not clear.
Screen house experiments revealed that races 2 and 3 could also cause disease in N. 9ng9t9
when inoculated artificially. However, there are only two reports of race 2 isolates from
the same weed under natural conditions, one from Rwanda and another from Tanzania
(Taylor e_t 91., 1987 ). Thus, the general trend seems to be clear, that under natural
conditions susceptibility of N. 9igl9ii seems to be restricted to race 1. It is unknown why
race 2 was not recovered from N. 91g11t_ii in this study, despite its high frequency on bean
in northern Tanzania (Chapter 2). One possible reason for lack of natural infection may
be that the minimum bacterial population required for disease occurrence in the weed was
not attained. Infectivity titration and host-pathogen relationship studies are needed to

understand the relationship between races of 2. 9. pv. phaseolicola and N. wightii.

Based on the lack of seed infection in this weed, we can hypothesize that under

field conditions a nearby infected bean crop acts as a source of initial inoculum for N.
911199 plants. Once diseased, the weed remains so until the next season when it acts as
a source of inoculum for a pathogen-free bean crop that may be grown nearby, or as a
supplementary source of inoculum for the bean crop. The restriction of halo blight disease
symptoms on foliage, and the failure to detect any internal infection in flower buds,
blossoms, young pods, mature dry seeds and pod shells, indicate that the pathogen does not
move systemically in the vascular tissue of this weed to the seed in the pod. The mechanism
by which 2. 9. pv. phaseolicola is excluded from vascular transmission to the seed of this
weed is unknown. Also, it is not known why other bacterial contaminants occurred in N.
99119 seed at very high numbers, nor how they gained access to the seeds. It is
epidemiologically significant that 2.9. pv. phaseolicola is not seed-borne in N. @1911. This
eliminates the danger of N, @1911 spreading the halo blight pathogen into new areas
through seed. Although this danger remains with infected bean seed, procedures of
producing bean seed relatively free from halo blight bacteria are feasible (Schwartz, 1989).
Insect pests attacking N. 91199 pods and seed do not appear to be bean pests.
However, the possibility of cross attack by leaf chewing insects can not be eliminated.
Moreover, contaminated insect pests from diseased N. 91gl99 might, under high moisture
conditions such as those at Lyamungu, deposit halo blight bacteria when walking on bean
leaves. This possibility increases when an insect is a common pest of both beans and N.
913m. More experiments are needed to establish the possibility that insect pests may serve

as vectors of 2. 9. pv. phaseolicola between the two hosts.

87

Attempts to control halo blight disease in northern Tanzania and in other areas
must take into consideration infected N. @199 plants as reservoirs of the pathogen.
Control measures should aim at disengaging the source of inoculum from the new bean
crop by use of pathogen-free seed and sanitation measures. Because survival of N. @191
plants occurs primarily outside cultivated fields, destruction of such plants around bean
fields would prevent a disease-free bean crop from infection by inoculum from weeds by
shifting the source of inoculum further away from the bean fields. These disease control

measures could also be supported by the introduction of halo blight-resistant bean varieties.

LITERATURE CITED

CIAT 1987. Annual Report. Bean Program Working Document No. 39. CIA'I; Cali,
Colombia. 352pp.

Dinoor, A. 1974. Role of wild and cultivated plants in the epidemiology of plant diseases
in Israel. Ann. Rev. Phytopath. 12:413-436.

Ekpo, E.J.A. 1975. Pathogenic Variation in Common (Xanthomonas phaseoli) and fuscous
(5. phaseoli var. fuscans) Bacterial Blights of Bean (Phaseolus vulgaris L.). Ph.D.
Dissertation. Michigan State University, East Lansing. 127pp.

 

Ercolani, G.L, DJ. Hagedorn, A. Kelman and R.E. Rand. 1974. Epiphytic survival of
Pseudomonas sygingae on hairy vetch in relation to epidemiology of bacterial
brownspot of bean in Wisconsin. Phytopathology 64:1330—1339.

Giga, DP. and RH. Smith. 1987. Egg production and development of Callosobruchus
rhodesianus (Pic) and Callosobruchgg maculatus (E) (Coleoptera: Bruchidae) on
several commodities at two different temperatures. J. Stored Prod. Res. 23:9-15.

Gondwe, B. 1987. Halo blight of Phaseolus beans. pp. 70-74. 19: Salema, M.P. and AN.
Minjas (eds). Bean Research 1. Proceedings of the Fifth Bean Research Workshop
held at Sokoine University of Agriculture, Morogoro, Tanzania, September 9-12,
1986. Benedictine Publications, Ndanda, Peramiho. 167pp.

Gondwe, B. 1985. Screening beans for resistance to halo blight bacteria. pp. 93-97. I_n:
Salema, M.P. and Minjas, AN. (eds). Proceedings of the Fourth Workshop on
Bean Research in Tanzania, held at Sokoine University of Agriculture, Morogoro,
Tanzania, September 5-7, 1985. Washington State University, Pullman. 185pp.

Goto, M. 1972. The significance of the vegetation for the survival of plant pathogenic
bacteria. pp. 39-53. 19: Maas-Geesteranus, H.P. (ed). Proc. Third Int. Conf. on
Plant Pathogenic Bacteria. Center for Agricultural Publishing and Documentation.
Wageningen, The Netherlands. 365pp.

Hayward, AC. 1985. Characterization of phytopathogenic pseudomonads: an approach to
minimal standards for description. pp. 775-782. I_n: Civerelo, E.L., A. Collmer, R.E.
Davis, and AG. Gillaspie (eds). Plant Pathogenic Bacteria. Proc. Sixth Int. Conf.
on Plant Pathogenic Bacteria. Maryland, June 2-7. Martinus Nijhoff Publishers,
Dordrecht. 1050pp.

Hildebrand, DC. and M.N., Schroth. 1971. Identification of fluorescent pseudomonads. pp.

281-287. I_n: Maas-Geesteranus, H.P.M. (ed). Proc. Third Int. Conf. on Plant
Pathogenic Bacteria. University of Toronto Press, Canada. 365pp.

88

89

Kikumoto, 'II and M. Sakamoto. 1969. Ecological studies on the soft rot bacteria of
vegetables. VII. The preferential stimulation of the soft rot bacteria in the
rhizosphere of crop plants and weeds. Ann. Phytopath. Soc. Japan 35:36-40.

King, E.O., M.K. Ward, and DE. Raney. 1954. Two simple media for the demonstration
. of pyocyanin and fluorescin. J. Lab. Clin. Med. 44:301-307.

Kishun, R. and HS. Sohi. 1979. Two new hosts for Xanthomonas vesicatoria (Doidge)
Dowson. Curr. Sci. 48:485-486.

Klement, A., 6.1.. Farkas, and Lovrekovich. 1964. Hypersensitive reaction induced by
phytopathogenic bacteria in the tobacco leaf. Phytopathology 54:474-477.

Kovacs, N. 1956. Identification of Pseudomonas pyocyanin by the oxidase reaction. Nature
178:703.

Kuan, 'IIL., G.V. Minsavage, and NW. Schaad. 1986. Aerial dispersal of Xanthomonas
campestris pv. campestris from naturally infected Brassica campestris. Plant Dis.
70:409-413.

Laub, CA. and R.E. Stall. 1967. An evaluation of Solanum nig1_'9m and Phfialis minima as
suscepts of _Xgnthomonag vesicatoria. Plant Dis. Reptr. 51:659-661.

Lelliot, RA. and DE. Stead. 1987. Methods in Plant Pathology, Vol. 2. Methods for
Diagnosis of Bacterial Diseases of Plants. Blackwell Scientific Publications. Oxford.
216pp.

Lelliot, R.A., E. Billing, and AC. Hayward. 1966. A determinative scheme for the
fluorescent plant pathogenic pseudomonads. J. Appl. Bacteriol. 29:470—489.

Lindow, SE. 1988. Lack of correlation of i9 vitro antibiosis with antagonism of ice
nucleation active bacteria of leaf surfaces by non-ice nucleation active bacteria.
Phytopathology 78:444-450.

Lindow, S.E., D.C. Amy, and CD. Upper. 1978. Erwinia herbicola a bacterial ice-nucleus
active in increasing frost injury to corn. Phytopathology 68:523-527.

 

 

Lindow, S.E., D.C. Amy, and DC. Upper. 1982. Bacterial ice nucleation, a factor in frost
injury to plants. Plant Physiol. 70:1084-1089.

Morris, M.J. and RS. Knox-Davies. 1980. Raphanus raphanistrum as a weed host of
pathogens of cultivated cruciferae in the Western Cape Province of South Africa.
Phytophylactica 12:53-55.

Olive, J .W. and SM. McCarter. 1988. Occurrence and nature of ice nucleation active
strains of Pseudomonas gm'ngae on apple and peach trees in Georgia. Plant Dis.
72:837-843.

90

Schaad, N .W. 1988. Laboratory Guide for Identification of Plant Pathogenic Bacteria,
Second Edition. APS Press. St. Paul. 164pp.

Schaad, NW. and J.C. Dianese. 1981. Cruciferous weeds as sources of inoculum of
Xanthomonas camfltris in black rot of crucifers. Phytopathology 71:1215-1220.

Schuster, M.L. and DP. Coyne. 1975. Detection of bacteria in bean seed. Ann. Rept. Bean
Improv. Coop. 18:71.

Schwartz, HR 1989. Halo blight. pp. 285-301. 19: Schwartz, HE and M.A. Pastor-Corrales
(Eds). Bean Production Problems in the Tropics. 2nd Edition. CIA'I; Cali,
Colombia. 726pp.

Smith, TE. 1962. A variant culture of Xanthomonas malvacearum obtained from weed
roots. Phytopathology 52:1313-1314.

Starr, G.H. and CJ. Kercher. 1969. Passage of Pseudomonas phaseolicola in bean plants
through sheep. Phytopathology 59:1976 (Abstr.).

Taylor, JD. 1970. Bacteriophage and serological methods for the identification of
Pseudomonas phaseolicola (Burkh.) Dowson. Ann. Appl. Biol. 66:387-395.

Taylor, J.D., D.M. Teverson, and J.H.C. Davis. 1987. Diseases of Phaseolus beans - biology,
resistance and control. Institute of Horticultural Research Annual Report.
Wellsborne, Warwick, England, 1986/1987. pp. 63-64.

Thornley, M.J. 1960. The differentiation of Pseudomonas from other Gram-negative bacteria
on the basis of arginine metabolism. J. Appl. Bacteriol. 23:37-52.

Valleau, W.D., EM. Johnson, and S. Diachum. 1944. Root infection of crop plants and
weeds by tobacco leaf spot bacteria. Phytopathology 34:163-174.

CHAPTER4

POPULATION DYNAMICS OF PSEUDOMONAS SYRINGAE PV. PHASEOLICOLA

IN TWO BEAN CROPPING SYSTEMS IN NORTHERN TANZANIA.

91

INTRODUCTION

The majority of the bean crop in Africa and other tropical areas is produced by
small farmers in complex associations with other crop species, notably maize. Depending
on the country, companion crops may include bananas, cassava, sweet potatoes, yarns, peas,
and coffee. Such crop associations on the same land and at the same time are referred to
as intercrops (CIAT; 1986a). The bean cultivars used in these cropping systems (landraces)
usually represent years of natural evolution as the farmers have selected them for survival
and production. The term landrace refers to a heterogeneous mixture of a predominantly
self-pollinating species which is maintained by a subsistence farmer. Very low levels of
hybridization may occur causing genetic recombination. The occurrence of such genetic
drift and individual farmer selection serve to make each population genetically unique
(Martin, 1984). Some of these landraces include a great mixture of seed types, while others
may have a high proportion comprising generally similar seed types though of diverse
phenotype, otherwise. Bean landraces can easily be separated into their components by
planting progeny rows from single plants, and when this is done, differences in grth habit,
disease resistance and yield potential between them are readily apparent. Some components
are often so obviously inferior that it is difficult to understand why they have survived.
Occasionally, one component may appear so outstanding in comparison with others that it
is equally difficult to understand why its selection as a more or less pure line has been
avoided (Leakey, 1970). Landraces are generally good competitors with weeds and other

associated crop species (CIAT: 1986b, Leakey, 1970).

92

93

Until the 1960’s there has been a tendency in East Africa and Africa in general, for
agricultural workers to discourage intercropping and instead to promote monocropping.
However, small farmers have persistently refused to abandon the system of intercropping
(van Rheenen, gt 91., 1981). Moreover, during the last ten years, there has been an
increasing awareness that the impact of the so-called green revolution on small holder
farming in developing countries has remained rather limited. Hence, research has directed
more attention to the analysis and subsequent improvement of traditional cropping systems
including the more efficient use of limited resources. Research results so far clearly
indicate that traditional intercropping systems are better adapted to the ecological, socio-
economic and socio-cultural conditions of tropical agriculture than monocropping systems
(Burdon, 1978; CIAT; 1986b; Norman and Baker, 1974; Steiner, 1984).

Various researchers have pointed out that plant disease epidemics are favored by
morphologically and genetically uniform crops grown on large areas of land (Browning,
1957; Browning and Frey, 1969; National Academy of Science, 1972). By contrast, a
combination of genetically different crops grown together in the same field does not provide
the uniform substrate needed by the pathogen to multiply and acquire epidemic proportions
(Borlaug, 1958; Browning and Frey, 1969; Burdon, 1978; CIAT; 1986b). Accumulating
evidence also indicates that a reasonable proportion of the plant population will prove
resistant to any host-specific pathogens that appear. Intercropping of bean with maize in
the tropics has been reported to reduce disease severity in the bean crop (Katunzi, _et 91.,
1987; Kikoka and Teri, 1988; Mukiibi, 1976; Msuku and Edje, 1982; van Rheenen, e_t a_l.,
1981). Insect infestation has also been reported to be reduced (Altieri, e_t a_l., 1978).
However, this is not always the case. Much depends on other factors such as the host

range of the pathogen, the relative sowing time and the spatial geometry of the associated

94
crops (CIAT; 1986b). In addition, the behavior of pathogen populations in intercrop
systems is a complex function of the reproductive rate of the pathogen and the rate of
propagule transfer between host components (Burdon, 1978; CIAT; 1986b).

For example, anthracnose (Colletotrichum lindemuthianum) may be more severe on
beans associated with maize than on beans in the monocrop system due to increased
humidity and lower temperature in the canopy. Angular leaf spot (Phaeoisariogis M)
may also be more severe on beans intercropped with maize than on beans intercropped
with cassava (CIAT; 1986b; Moreno, 1977). Bacterial diseases (Xanthomonas campestris
pv. M and E. s. pv. phaseolicola) were reported to be less severe in beans in
associations with maize than in a sole cropping system (Katunzi, e_t 91, 1987; Msuku and
Edje, 1982; van Rheenen, e_t 91., 1981; Vermeulen, 1982).

Several mechanisms of disease reduction in the intercrop have been suggested
(Burdon, 1978; Manners, 1982). These include: 1) The replacement of a high proportion
of genotypes that in a pure stand would be susceptible plants by resistant ones reduces the
amount of susceptible tissue per unit area of crop, and hence the amount of infection in
the succeeding generation, 2) The average distance which the inoculum has to travel to
reach a susceptible plant surface is increased, 3) The movement of inoculum from one
susceptible plant to another may be hindered by the presence of intervening resistant
plants and, 4) Cross protection may occur, i.e., pre-inoculation with a race to which the line
is resistant may protect it from a race to which it is normally susceptible. However, the
relative contribution made by each of these mechanisms is far from clear.

Research work on the effect of intercropping bean with maize in reducing disease
severity has been based on visual disease ratings. However, no work has been done to

relate visual disease severity ratings with changes in pathogen populations in the bean

95
monocrop and bean-maize intercrop systems. Population dynamics of plant bacterial
pathogens may provide an estimate of the degree to which a given agricultural system is
ecologically suited to a particular pathogen, and has been used to describe various bacteria-
plant interactions (O’Brien and Lindow, 1988). The objective of the current investigations
was, therefore, to examine the population dynamies of halo blight bacteria in the two bean

cropping systems and relate bacterial populations to disease severity ratings.

MATERIALS AND METHODS

Location

Field experiments of bean (Phaseolus M and maize (_Zg 99m) in
monoculture and in association were conducted in experimental fields at Lambo and
Lyamungu at altitudes of 1020m and 1268m, respectively, and at latitude 3°14’S and
longitude 37°15’E, in 1989 during the long rainy season (March to June). The two
locations were about 13km apart on the southern side of mount Kilimanjaro. Both
locations were in Hai district, Kilimanjaro region, and were planted with maize the previous

two seasons.

mfimental desig9
A randomized complete block design (RCBD) with three replications per treatment

was used. Maize cultivar Kilima and bean cultivar Canadian Wonder were used. Pathogen-
free seeds were kindly provided by B. Gondwe (Agricultural Research Institute, Lyamungu,
Moshi, Tanzania). The three treatment combinations included i) Monoculture beans with
intra- and inter-row spacings of 75 cm and 20 cm, respectively, 2 plants per hill; ii)

Monoculture maize with a spacing of 75 cm x 25 cm, 2 plants per hill; and iii) Beans in

96
association with maize, i.e., single rows of beans alternating with single rows of maize; with
spacings as indicated for i) and ii). The two crops were sown simultaneously. Each plot
measured 4m x 3m. Planting and all other agricultural practices were accomplished by
manual labor, and were carried out as recommended for the region. Before planting,
samples of soil, weeds and maize debris were taken using the stratified sampling design of
Delp gt a_l. (1986). Samples were placed in plastic bags and sent to the laboratory for assay

of E. s. pv. phaseolicola as described in Chapter 3.

Bacterial isolate

Race 1 isolate #6, used in this study, was obtained from 19. M9; and reisolated
from artificially inoculated bean cultivar Canadian Wonder. Inoculum was prepared from
48-hour-old cultures grown on KMB (King, e_t a_l, 1954) plates by rinsing the bacteria off
the agar surface and suspending them in glass distilled water. Concentrations were adjusted

turbidimetrically to contain about 107-108 cells/ml.

Inoculation of plants

Plants were spray-inoculated with bacterial suspensions on both sides of the leaves
to run-off, with some slight water-soaking, using a hand-operated atomizer. Both bean and
maize plants were inoculated so that the behavior of the pathogen on maize plants could
also be studied. Plants were inoculated when 19 days old, when the second trifoliolate leaf

was fully open on beans, and when the fifth leaf was just unfolding on maize plants.

Sampling of leaves

Leaf samples were collected between 8 to 10 a.m. before inoculation, immediately
after inoculation when leaves had dried and thereafter, at 3-day intervals for up to 15 days.
For each replicate, 6-8 leaves were sampled at random and were immediately transported

on ice in plastic bags to the laboratory and kept at 5°C until processed within 24-36 hours.

Bacterial mpulation trends

Epiphytic populations of _12. s. pv. phaseolicola were estimated using procedures of
Smidt and Vidaver (1986). Leaves were shaken in sterile 0.01M phosphate buffer
containing 0.01% Tween 20 on a wrist action shaker adjusted at medium speed for 30
minutes at 20-22°C. The amount of sterile phosphate buffer varied with each sample to
ensure that sufficient volume was used to completely immerse the samples. After
appropriate 10-fold serial dilutions, suspensions were plated in triplicate on KMB
supplemented with IOOug cycloheximide per ml. Bacterial colonies were counted after 4-
5 days of incubation at 24 j; 2°C. After washing, leaves were blotted dry and traced on
paper and the area measured using a LI-3100 area meter (LI-COR, Inc., Lincoln, Nebraska,
U.S.A.). Colony counts were expressed as numbers per cm2 of leaf. Only symptomless
leaves were used for determination of epiphytic populations.

Internal populations of 2. s. pv. phaseolicola were estimated from surfaCe-sterilized
leaves (3 minutes in 2.6% NaOCl). Ten 2-cm diameter disks were punched out with a
sterile cork borer, placed on sterile filter papers, weighed, and homogenized in 3-5 ml of
phosphate buffer. Homogenates were prepared, serially diluted and plated as described

earlier. Bacterial counts were expressed as numbers per gram of fresh tissue.

98

O

For each sampling, 10 randomly chosen _P_. g. pv. phaseolicola colonies were tested
for pathogenicity on Canadian Wonder bean seedlings by the three methods described in

Chapter 3.

Evaluation of disease reaction

Bean plants in monoculture and in association with maize were evaluated for disease
reactions on foliage and pods. Halo blight severity was evaluated using the CIAT scale of
1-9 (CIAT; 1987), where 1 represented absence of symptoms and 9 represented very severe
disease. Leaf disease reaction was estimated at 10 and 25 days after inoculation, whereas
pod symptoms were rated at physiological maturity. For each disease rating date, the
canopy of each individual plant was visually inspected row-wise in every plot and a rating
representing the whole plot was assigned. Maize plants were also observed for any

presence of symptoms.

Leaf wetness studies

To understand the effect of the two cropping systems on moisture retention in the
canopy, the period required for bean and maize leaves to dry after rain was determined.
Leaves were examined for the presence of free moisture at 30-minute intervals following

the cessation of rain at Lyamungu for 10 different days of rainfall.

Bean seed infection assay

After drying, bean plants were harvested manually. Pods from each plot were kept
separate and allowed to dry under the shade. Following hand-shelling, a random sample

of 600 bean seeds was drawn from each replicate using a wooden board with 100 holes.

99
After surface-sterilization, seeds were aseptically plated, hilum down, on KMB-cycloheximide
agar, 6 seeds per 10 cm-diameter glass petri dish and incubated for 5 days. To identify 2.
s. pv. phaseolicola-infected seed, plates were observed under UV-light in the darkness.
Pure cultures of the pathogen were made from infected seeds by a series of single colony
transfers on KMB agar plates and tested for pathogenicity as described earlier (Chapter 3).
Two separate determinations of seed infection were made for the same plot using

the same sample size.

Weather data
Rainfall and temperature data for the Lambo estate and Lyamungu, were obtained
from the Agrometeorology Section at the Agricultural Research Institute, Lyamungu, Moshi,

Tanzania.

Data analysis

Statistical computations were made using the MSTAT-C statistical package
(Michigan State University). Bacterial populations were log-transformed to determine the
effect of each cropping system on population size. Significant difierences between cropping
systems were estimated using Student’s t-test. Data for percentage bean seed infection

were arcsine transformed before analysis (Little and Hills, 1978).

RESULTS

Population dmamics at Lambo

Populations of 13. s. pv. phaseolicola on and in bean leaves in the bean monocrop

and in bean/maize intercrop systems at Lambo are shown in Fig. 1. From 0 to 6 days after

100

inoculation surface populations were almost the same in both cropping systems. However,
at 9 days after inoculation surface populations were higher in the bean/maize intercrop
system than in the bean monocrop system, and were significantly higher (2 = 0.05) in the
intercrop than in the monocrop system at 12 and 15 days after inoculation. A similar
pattern was observed for internal populations of Ls. pv. phaseolicola in bean leaves.
However, internal populations were much higher than surface populations throughout the
monitoring period (Fig. 1).

Internal populations were higher in the bean/maize intercrop than in the monocrop
starting three days after inoculation. The differences were statistically significant (2 = 0.05)
at 12 and 15 days after inoculation when the maize leaf canopy started shading the bean
canopy. Internal bacterial populations reached a maximum of 7.8 and 9.6 log10 units/g fresh
weight at 9 and 12 days after inoculation in the bean monocrop and bean/maize intercrop
systems, respectively, after which there was a gradual decline in fig. pv. phaseolicola

populations. Such was also the case for surface populations (Fig. 1)

P_op991tion dyngmics gt Lva_mu_ng9

Differences in bacteria multiplication among the two cropping systems also were
apparent at Lyamungu, but were less pronounced than at Lambo. The internal population
of fig. pv. phaseolicola on and in bean leaves tended to be greater in the bean/maize inter-
cropping than in the bean monocropping system throughout the sampling period (Fig. 2).
However, internal populations were not significantly affected (_13 = 0.05) by cropping
systems except at the 15 day sampling period. Surface populations of Lg. pv. phaseolicola
increased slightly during the first 6 days and then increased rapidly in both cropping systems

reaching a maximum level of 4.7 and 5.6 log1o units/cm2 leaf area at 9 days in the bean

101

Figure 1. Surface and internal population dynamics of Pseudomonas s99'ngae pv.
phaseolicola race 1 on and in bean foliage in two bean cropping systems at Lambo,
Kilimanjaro region, northern Tanzania. Bars indicate standard errors of the means.

LOG CPU/0M2 LEAF AREA (SURFACE)

 

._o

 

 

 

 

 

 

 

m 4 \ lml 1% .
i I
m . 4
.a 1 w
M . _zammomom. .zammezOes ..
. _zammz>rt IIIII .
z_ozoofio.U wcmmpom 2.020032v
O b I O O
m w m m a.» «m

O><m 23M.» _ZOOCP>._._OZ

3

N

LOG CFU/G FRESH WEIGHT (INTERNAL)

102

103

Figure 2. Surface and internal p0pulation dynamics of Pseudomonas sy9'ngae pv.
phaseolicola race 1 on and in bean foliage in two bean cropping systems at Lyamungu,
Kilimanjaro region, northern Tanzania. Bars indicate standard errors of the means.

 

LOG CPU/0M2 LEAF AREA (SURFACE)

 

 

 

 

 

 

 

8 3

m: ,m

a. -m

a. L.

m . _zammomgU _zqmmomQu - m
sin...» mcmfiom oil...

.. _zammzz. gOZOOmQu 7820030“. ..

T 5 O O

o d a d u 4 d o
o a m o a a

O><m >m._.mm _ZOOCCKZOZ

LOG CFU/G FRESH WEIGHT (INTERNAL)

104

105
monocrop and bean/maize intercrop systems, respectively. Thereafter, surface populations
declined sharply; monocrop populations reached a level significantly lower than intercrop
populations at 12 days, and then suddenly increased to a level not significantly different (2
= 0.05) from that of the intercrop system at 15 days after inoculation. While internal
population patterns were similar in both systems, maximum bacterial populations generally
tended to be lower in the bean monocrop than in the bean/maize intercrop system. At 15
days after inoculation the difference was about 100-fold and was significantly different (9
= 0.05). Internal populations reached a maximum of 8.3 log10 units/g fresh weight at 12
days after inoculation in the bean monocropping system, then declined at 15 days. Bacterial
populations in bean leaves in the intercrop system increased to 9 log1o units at 15 days.
In both cropping systems internal Ls. pv. phaseolicola populations increased steadily except

at 15 days after inoculation.

Populations of Rs. pv. phaseolicola on and in maize

In contrast to behavior on and in bean leaves, internal populations of Ls. pv.
phaseolicola in maize plants grown alone or in association with bean, decreased throughout
the sampling period both at Lambo and at Lyamungu (Figs. 3 and 4). At Lambo, practically
no internal bacteria were detected in maize leaves 6 days after inoculation in both cropping
systems. In the maize monocropping system surface Ls. pv. phaseolicola populations also
declined rapidly and were not detected 6 days after inoculation. However, in the
bean/maize intercrop system, surface populations on maize leaves increased slightly 3 days
after inoculation, then maintained a fairly constant population of 2.8 log1o units/cm2 leaf

area until 9 days and then declined to undetectable concentrations at 15 days.

106

Figure 3. Surface and internal population dynamics of Pseudomonas _svg'ngae pv.
phaseolicola race 1 on and in maize foliage in monocrop and intercrop systems at Lambo,
Kilimanjaro region,northern Tanzania. Bars indicate standard errors of the means.

LOG CFU/CM2 LEAF AREA (SURFACE)

 

 

 

 

 

 

 

 

 

a 8
. _zammomom. .
Ell-I"
_zqmmz>r
m 2.058%
n .P I rm
_zammomQu
- mcmm>om oil... I
26200an
a. o s .m
L w
a. L
v \mnniimiin /
N- Ill W rm
I
i I 1
I
I
/
o . J I o
o w m o E a

O><m >_u._.mm _ZOOCFPdOZ

LOG CFU/G FRESH WEIGHT (INTERNAL)

107

108

Figure 4. Surface and internal population dynamics of Pseudomonas m'ngae pv.
phaseolicola race 1 on and in maize foliage in monocrop and intercrop systems at
Lyamungu, Kilimanjaro region, northern Tanzania. Bars indicate standard errors of the
means.

 

LOG CFU/CM2 LEAF AREA (SURFACE)

 

do do

Edd—£030."v
Till}

m .. _zammzz. szozoowon . m

 

L _ZHMDODOU u
' | | I .
mcmfi>0m
a . £028an .. m
o) 6

 

 

 

 

 

o w m o S.» a
o><m >3mm _zooc_.>:oz

LOG CFU/G FRESH WEIGHT (INTERNAL)

109

110

At Lyamungu, population dynamies of fig. pv. phaseolicola were similar to those at
Lambo. Internal populations in the maize monocrop decreased rapidly and were not
detected after 6 days (Fig. 4). Internal populations in the bean/maize intercropping system
declined more slowly, but were not detected after 9 days. On the other hand, surface
p0pulations increased to a maximum of about 3.4 log1o units/cmz at 3 days after inoculation
and then declined (Fig. 4). The bacterium was not detected on maize after 9 days in the
maize monocrop but remained at ca. 2 log10 units/cm2 through day 12 in the bean/maize
intercrop system.

In all cases colony samples of Rs. pv. phaseolicola which were taken randomly and
tested for pathogenicity on Canadian Wonder bean seedlings were pathogenic and produced

typical halo blight symptoms within 8-10 days after inoculation.

Leaf and M disease ratings

Foliage disease severity was evaluatedtwice, at 10 and 25 days after inoculation, and
pod disease severity was rated once at pod physiological maturity. The mean disease
severity ratings plus/minus their standard errors for the two locations are shown in Table
1. Halo blight severity at Lambo was similar for the two cropping systems. However, at
Lyamungu, halo blight was slightly, but not significantly, more severe (_13 = 0.05) in the bean
monocrop system than in the bean/maize intercrop system. In contrast, however, pod
disease severity at both locations was more severe in the bean/maize association than in the

monocropping system, although differences were not significant. At Lambo, the proportion

111

Table 1. Halo blight disease severity ratings for foliage and pods of Canadian Wonder bean
in two bean cropping systems in northern Tanzania.

 

Mean disease severig ratings"

Location Plant part _D_aEZ Monocrop Intercrop
Lambo Leaves 10 4.3 i 0.2 4.2 t 0.2
25 6.8 :t 0.4 6.8 i 0.4

Podsz 3.8 t 0.2 5.0 i 0.3

Lyamungu Leaves 10 4.0 i 0.0 3.8 i 0.3
25 7.2 i 0.2 6.2 :L- 0.2

Podsz 4.8 t. 0.3 6.0 i- 0.3

 

xValues are means of three replicates : standard errors of the means. Disease severity
indices ranged from 1 = no disease to 9 = very severe disease, with 50% or more of leaf
area covered with lesions.

YDays after inoculation.

zDisease severity rating on pods was done at pod physiological maturity with a scale of 1-
9 where 1 = no disease and 9 = very severe disease, with 50% or more of pod area
covered with lesions.

112
of pod area covered with lesions in the intercropping system exceeded that in the monocrop

system by 24% while at Lyamungu the difference was 20%.

Retention of moisture on leaves

Leaves of beans growing in the intercrop system required 2.8 hours to dry after rain
as compared with 2.0 hours for the bean monocrop, a difference of 40% (Table 2). Such
differences were not observed for maize in the two cropping systems. Maize leaves

required about 2 hours longer to dry than bean leaves.

Climatological data

Rainfall totals for April and May at Lambo were 242.1 and 201.2 mm, respectively
(Fig. 5). However, the 53 year rainfall averages were 209.0 mm for April and 161.7 mm
for May. Temperatures at the same location in April ranged from 17.1 to 27.4°C, which
almost agreed with the range of 53 years, which was 18.1 to 27.1°C. In May however,
temperatures ranged from 17.4 to 27.7°C as compared to 17.6 to 249°C, the average range
of 53 years. On the other hand, Lyamungu received 303.1 and 810.6 mm of rain for April
and May, rCSpectively (Fig. 6). The means of 53 year rainfalls for the same months were
502.3 and 414.1 mm, respectively. TCmperatures at Lyamungu were cooler than those at
Lambo and ranged from 15.2 to 23.9°C for April and 15.1 to 222°C for May. The 53 year
average ranged from 15.8 to 246°C for April and 13.0 to 222°C for May. Rainfall and
temperatures at both locations decreased progressively towards the end of the growing
season. Population dynamics studies of Ls, pv. phaseolicola were conducted during the

period of April and May.

113

Table 2. Time required for visible moisture on bean and maize leaves to dry following
cessation of rain at Lyamungu, Kilimanjaro region, northern Tanzania.

 

Cropping sgtem Qrgp Time (hours)z

MonocrOp Bean 2.0 t 0.2
Maize 4.3 t 0.3

Intercrop Bean 2.8 :1: 0.2
Maize 4.3 d: 0.2

 

zFor 10 rain instances in a field with 3 replications (= 30 readings per value) plus/minus
standard errors of the means.

114

Figure 5. Monthly rainfall and mean monthly maximum and minimum temperatures from
March to July at Lambo for 1989 and for 53 years (1935-1988). Numbers above each bar
represent number of rain days during the indicated month.

RAINFALL (MM)

 

600 :-

500 ~-

400 ~-

I

300 .

200 ~-

100 ‘-

 

 

"30
«25 8
‘3...
LL!
”-20 n:
,...-_ a
k”.—O——"'.\\\ <
‘x T‘ “15 Ibl-J
\
17 ‘ 5.
7 IE
2‘ «~10
18
9 -5
lޤl 8
F8 0

 

 

 

MAR APR 'MAY JUN ‘JUL
MONTHS

RAINFALL [3 1939
MEAN OF 53 YRS.

TEMPERATURE °—-° MEAN MAXIMUM 1989

'- -° MEAN MINIMUM 1989

.—-. MEAN MAXIMUM
OF 53 YRS.

- -‘ MEAN MINIMUM
OF 53 YRS.

115

116

Figure 6. Monthly rainfall and mean monthly maximum and minimum temperatures from
March to July at Lyamungu for 1989 and for 53 years (1935-1988). Numbers above each
bar represent number of rain days during the indicated month.

RAINFALL (MM)

 

       

 

  
    

 

 

 

28
800 -- ...
700 +- 2—35
600 «L 2-30
500 ~- \23 2-25
\
4oo -- N 2—20
A A
22 A
3002- r:: § 9 (~15
.— § \
A A
200 )_ Q S ~10
A A
2 A A
100» \ § - 5
A \ “is
2 A A (A
\ .

 

 

 

 

MART APR‘ MAW JUN ‘ JUL '
MONTHS

RAINFALL El 1939
MEAN OF 53 YRS.

TEMPERATURE -—-- MEAN MAXIMUM 1989
°—-° MEAN MINIMUM 1989

a- -A MEAN MAXIMUM
OF 53 YRS.

‘- -‘ MEAN MINIMUM
OF 53 YRS.

117

TEMPERATURE (°C)

118 .

Egan seed infection

The percentage of bean seed infected with Ls. pv. phaseolicola was lower for bean
grown in pure stands than when grown in association with maize, but the differences were
not statistically significant (2 = 0.05) except for experiment 2 at Lambo (Table 3). In
general, experiments at Lambo and at Lyamungu showed similar trends of seed infection.
High levels of seed infected with halo blight bacteria in the bean/maize intercrop system
were consistent with the high disease severity ratings on pods at both locations. Randomly
selected pure cultures of fig. pv. phaseolicola colonies from infected bean seed were all

pathogenic when tested on Canadian Wonder bean seedlings.

DISCUSSION

When considered together, the data from this study suggest that intercropping beans
with maize ecologically favors multiplication of fig. pv. phaseolicola in bean leaves.
Although there were no significant differences in foliage disease severity between the two
cropping systems, the increased inoculum density in the bean/maize intercrop system
resulted in more severe disease on pods and in a higher percentage of seed infection than
when beans were grown in a pure stand. Results of this study further suggest that disease
severity on foliage may not always be adequate to assess the advantage of intercrop systems
in reducing disease because high numbers of infected seed may be produced in such
systems.

The mechanism for increased _P_.s. pv. phaseolicola multiplication in the intercrop
system seems to involve increased moisture retention in the canopy as indicated by the
fact that maize leaves took longer time to dry after cessation of rain than bean leaves. In

addition, bean in association with maize retained moisture longer than bean grown alone.

119

Table 3. Percentage of Canadian Wonder bean seed infected with Pseudomonas sy9'ngae
pv. phaseolicola in two bean cropping systems in northern Tanzania.

 

Seed infection (%)Y

Location Expgriment Monocro Intercrop

Lambo 1 16.8 t 82 a 28.3 :I: 7 a
2 11.8:8a 38.11-5b

Lyamungu 1 12.3 1- 2 a 20.6 i 10 a
2 12.4:8a 32.2:9a

 

YValues are means of three replicates : standard errors of the means.

2Within each experiment, means followed by the same letter are not significantly different
(2 = 0.05) by Student’s t-test. Data were arcsine transformed before analysis.

120

This effect may be magnified when beans are planted late in the maize crop as is sometimes
the case in northern Tanzania and in other areas of the country because of the greater
canopy development. Bacterial populations, especially epiphytic populations, have been
reported to increase when plants are wet (Leben and Daft, 1967; Mulrean and Schroth,
.1982; Smitley and McCarter, 1982). Increased moisture retention in the intercrop system
may also result in a lowered temperature in the canopy. Therefore, in areas where high
temperature and reduced moisture are limiting factors for halo blight development, such as
at Lambo, intercropping bean with maize may increase halo blight severity, especially on
pods. This is of practical significance in countries such as Tanzania, where programs to
produce pathogen-free seed are lacking. Data from this work also indicate that maize
leaves do not support high populations of halo blight bacteria. This reduces the risk of the
maize canopy providing additional inoculum for the bean crop when the two are grown in
association.

Comparison of data obtained from the two sites indicate that Ls. pv. phaseolicola
multiplied to higher numbers at Lambo than at Lyamungu (Figs. 1 and 2). It is likely that
differences in population dynamics were caused by differences in temperature.
Temperatures were about 4 C lower at Lyamungu than at Lambo during the assay period.
When combined with canopy effects, temperature differences between the two locations
could even be increased beyond 4 C. Differences in the amount of rainfall may also
account for differences in population dynamics between the two locations. Although
multiplication of phytOpathogenic bacteria on plant surfaces has been reported to increase
after rain (Hirano and Upper, 1983), when excessive rainfall, such as was the case at
Lyamungu, is combined with low temperature, a net negative effect on population size may

occur. We can therefore hypothesize that when initial inoculum is not a limiting factor,

121

temperature and moisture availability in the associated crops become key elements in
determining the balance between increasing and reducing disease severity in such cropping
systems. However, when initial inoculum comes from outside the cropping system, the
maize crop would act as a barrier and therefore, reduce the amount of initial inoculum
available for the bean crop (Burdon, 1978; Msuku and Edje, 1982). But this may not be
the case for seed-borne pathogens such as halo blight bacteria. The original source of
initial inoculum, therefore, is another factor governing the balance between increased and
reduced disease severity in intercrop systems.

Earlier studies (Msuku and Edje, 1982; van Rheenen, pt 91., 1981; Vermeulen, 1982)
have indicated that halo blight is generally less severe when beans are grown in association
with maize than in pure stands. Some of these studies, however, have involved only
observations in agronomy experiments (Msuku and Edje, 1982; Stoetzer and Waite, 1984;
van Rheenen, e_t 91., 1981), depending therefore, on natural infection which does not insure
uniform initial inoculum in the two bean cropping systems. Moreover, population dynamics,
percentage seed infection and moisture retention difierences have not been considered in
previous studies as quantitative factors to measure ecological fitness of the pathogen in
such systems. By using these parameters to quantify disease and introducing a uniform
amount of initial inoculum in the two bean cropping systems, results of the current study
generally do not agree with previous reports, because the environment in the bean/maize
intercrop system ecologically favored increased multiplication of halo blight bacteria as well
as increased seed infection.

Contradictory results of the effect of intercropping beans with maize on disease
severity have also been reported for bean anthracnose (CIAT; 1986b; CIAT; 1989; Msuku

and Edje, 1982) and on bean rust (Msuku and Edje, 1982). As van Rheenen e_t, 91. (1981)

122
suggested, the environmental differences between monocropping and intercropping, and
their influence on disease are far too complex for generalization and are not well
understood. In order to understand better the interaction of plant pathogens and crops
grown in association, we need to quantify the interacting parts of the systems. This will
involve simultaneous measurement of parameters related to the pathogen, which include
population size, disease incidence and severity, and weather variables such as moisture and

temperature (Hirano and Upper, 1983).

LITERATURE CITED

Altieri, M.A., C.A. Francis, A. van Schoonhoven and J .D. Doll. 1978. A review of insect
prevalence in maize (299 may L.) and bean (ghaseolus vulgaris L.) in polycultural
systems. Field Crops Res. 1:33-49.

Borlaug, M.P.. 1958. The use of multiline or composite varieties to control airborne
epidemic disease of self pollinated crop plants. pp. 12-26. 91: Proceedings of the
First International Wheat Genetics Held at the University of Manitoba, Winnipeg,
Manitoba, Canada. August 11-15. The Public Press Ltd., Winnipeg. 268pp.

Browning, LA. 1957. Studies of the effect of yield blends of oat varieties on stem rust
losses. Phytopathology 47:4-5 (Abstr.).

Browning, IA. and KJ. Frey. 1969. Multiline cultivars as means of disease control. Ann.
Rev. Phytopath. 7:355-382.

Burdon, J.H. 1978. Mechanisms of plant disease control in heterogeneous plant populations
- an ecologists view. pp. 193-200. I_n: Scott, RR. and A. Bambridge (Eds) Plant
Disease Epidemiology. Blackwell Scientific Publications. Oxford. 329pp.

CIAT 1986a. Bean Production Systems in Africa. CIAT; Cali, Colombia. 16pp.
CIAT 1986b. Principles of Intercropping with Beans. CIAT; Cali, Colombia. 40pp.

CIAT 1987. Standard Systems for the Evaluation of Bean Germplasm. CIAT; Cali,
Colombia. 54pp.

CIAT 1989. Annual Report. Bean Program. Working Document No. 68. CIAT; Cali,
Colombia. 397pp.

Delp, B.R., L.J. Stowell and J .J . Marois. 1986. Evaluation of field sampling techniques for
estimation of disease incidence. Phytopathology 76:1299-1305.

Hirano, SS. and CD. Upper. 1983. Ecology and epidemiology of foliar bacterial plant
pathogens. Ann. Rev. Phytopath. 21:243-269.

Katunzi, A.L., J.M. Thri, KP. Sibuga and RN. Misangu. 1987. The efl'ect of intercropping
and fertilizer application on bean diseases and yield. pp. 66-73. E: Salema, M.P. and
AN. Minjas (Eds). Bean Research. Vol. 2. Papers Presented at a Workshop Held
at Sokoine University of Agriculture, Morogoro, Tanzania, October 1-3. Benedictine
Publications, Ndanda, Peramiho. 261pp.

Kikoka, LP. and J.M. Tbri. 1988. Efl‘ect of cropping systems on bean disease and yield. pp.
40-44. 19: Salema, M.P. and AN. Minjas (Eds). Bean Research. Vol. 3 Papers
Presented at a Workshop Held at Sokoine University of Agriculture, Morogoro,
Tanzania, September 28-30. Benedictine Publications, Ndanda, Peramiho. 173pp.

123

124

King, E.O., M.K. Ward and DE. Raney. 1954. Two simple media for the demonstration of
pyocyanin and fluorescin. J. Lab. Clin. Med. 44:301-307.

Leakey, C.L.A. 1970. Crop Improvement in East Africa. Commonwealth Agricultural
Bureaux. Farnham Royal, England. 280pp.

Leben, C. and G. C. Daft. 1967. P0pulation variation of epiphytic bacteria. Can. J.
Microbiol. 13:1151-1156.

Little, TM. and FJ. Hills. 1978. Agricultural Experimentation: Design and Analysis. John
Wiley and Sons. New York. 350pp.

Manners, J.G. 1982. Principles of Plant Pathology. Cambridge University Press. London.
264pp.

Martin, GB. 1984. Genetic Diversity of Bean Landraces in Northern Malawi. MS. Thesis.
Michigan State University, East Lansing. 152pp.

Moreno, R. and L. Mora. 1984. Effect of cropping pattern and soil management of plant
diseases. 11. Bean rust epidemiology. Turrialba 34:41-45.

Msuku, W.A.B. and O.TI Edje. 1982. Effect of mixed cropping of maize and bean on bean
diseases. Ann. Rept. Bean Improv. Coop. 25:16-18.

Mukiibi, J. 1976. Possible relationship between intercropping and plant disease problems in
Uganda. pp. 116-117. I_n: Monyo, J.H., A.D.R. Ker and M.Marilyn (Eds).
Intercropping in Semi Arid Areas. Proc. First Symp. Intercrop. Semi Arid Areas
Held at Morogoro, Tanzania. May 10-12, 1976. 45pp.

Mulrean, EN. and M.N. Schroth. 1982. Ecology of Xanthomonas camgtris pv. juglandis
on Persian (English) walnuts. Phytopathology 72:434-438.

National Academy of Sciences. 1972. Genetic Vulnerability of Major Crops. National
Academy of Science. Washington, DC. 307pp.

Norman, D.W. and EFT. Baker. 1975. Cropping systems in northern Nigeria. pp 334-361.
I_n: IRRI (Ed). Proc. Cr0pping Systems Workshop. Los Bar'ios, Philippines. March,
1975.

O’Brien, RD. and SE. Lindow. 1988. Effect of plant species and environmental conditions
on epiphytic population size of Pseudomonas sm'ngae and other bacteria.
Phytopathology 79:619-627.

Smidt, M. and AK. Vidaver. 1986. Population dynamics of Clavibacter michiganense subsp.
nebraskense in field-grown dent corn and pop-corn. Plant Dis. 70:1031-1036.

125

Smitley, DR. and SM. McCarter. 1982. Spread of Pseudomonas m’ngae pv. tomato and
role of epiphytic populations and environmental conditions in disease development.
Plant Dis. 66:713-717.

Steiner, KG. 1984. Intercropping in Tropical Small-holder Agriculture with Special
Reference to West Africa. Second Edition. Deutsche Gesellschaft fiir Technische
Zusammenarbeit. Eschborn. Germany. 304pp.

Stoetzer, H.A.I. and B.H. Waite. 1984. Evaluation of bean diseases in the grain legume
project trials, Katumani Research Station. East African Agric. and Forestry Journal
44:333-353 (Special issue).

van Rheenen, H.A., O.E. Hasselbach and S.G.S. Muigai. 1981. The effect of growing beans
together with maize on the incidence of bean diseases and pests. Neth. J. Plant
Path. 7:193-199.

Vermeulen, J. 1982. Screening Thn Bean Cultivars (Phaseolus vulgaris L.) to Halo Blight
Under Six Different Planting systems. MS. Thesis (Plant Breeding). Dutch
Agricultural University, Wageningen and CIAT; Cali, Colombia (Abstr.).

TATE UNIV. RRRIES

LIB
IIIIIII IIIIHIIII I
443

300876 17

ICHISRN s
IHIIIIIIIII
3129