Weed Control in Rice

Weed Control
in Rice Crops
Suitability of Rhynchosporium alismatis

as a Mycoherbicide for Integrated
Management of Damasonium minus in
Rice Fields
A report for the Rural Industries Research

and Development Corporation
By Farzad Jahromi, Eric Cother and

Gavin Ash
May 2001
RIRDC Publication No 01/39

RIRDC Project No UCS 7A
2001 Rural Industries Research and Development Corporation.

All rights Reserved
ISBN 0 642 58260 2
ISSN 1440 6845
Weed Control in Rice Crops - Suitability of Rhynchosporium alismatis as a Mycoherbicide for Integrated
Management of Damasonium minus in Rice Fields
Publication No 01/39
Project No UCS-7A
The views expressed and the conclusions reached in this publication are those of the author and not necessarily
those of the Rural Industries Research and Development Corporation. RIRDC shall not be responsible in any way
whatsoever to any person who relies in whole, or in part, on the contents in this report.
This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the
Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the Publications
Manager on phone 02 6272 3186.
Researcher Contact Details
Dr Farzad Jahromi
School of Agriculture,
Charles Sturt University,
Wagga Wagga 2678
Phone
Fax
Email
02 69 334 201
02 69 332 812
fjahromi@csu.edu.au
RIRDC Contact Details

Rural Industries Research and Development Corporation
Level 1, AMA House
42 Macquarie Street
BARTON ACT 2600
P O Box 4776
KINGSTON ACT 2600
Phone:
Fax:
Email:
Website:
02 6272 4539
02 6272 5877
rirdc@rirdc.gov.au
http://www.rirdc.gov.au
Published in May 2001

Printed on environmentally friendly paper by Canprint
ii
Foreword
Weed control in many crops, especially rice, has typically been focused on the use of chemical
herbicides. These have provided quick and convenient control for years, if not decades. However, as
the use of single mode of action chemicals has increased, so too has the incidence of weed populations
developing resistance to that chemical.
Herbicide resistance is now a major concern to the farming industry and integrated weed management
is becoming more important in all cropping systems. Integrated weed management is a tool box of
options that can be tailored to individual farm, cropping and weed situations.
Biological control is one of these management options. As populations of aquatic monocot weeds
resistant to registered herbicides appear, there is increasing interest in the use of naturally occurring
pathogens of weeds as potential biological control agents.
This report describes on-going research into the use of one such pathogen for control of an important
aquatic weed in rice.
This project was funded by industry revenue, which is matched, by funds provided by the Federal
Government.
This report, a new addition to RIRDCs diverse range of over 600 publications, forms part of our Rice
R & D program, which aims to improve the profitability and sustainability of the Australian rice
industry through the organisation, funding and management of a research development and extension
program that is both market and stakeholder driven.
Most of our publications are available for viewing, downloading or purchasing online through our
website:
downloads at www.rirdc.gov.au/reports/Index.htm
purchases at www.rirdc.gov.au/eshop
Peter Core
Managing Director
Rural Industries Research and Development Corporation
iii
Acknowledgments
The thesis by Farzad Jahromi, on which this report is based, was submitted for the degree of Doctor of
Philosophy at Charles Sturt University. It would not have been feasible without the help and support
of a number of people.
Dr Jahromi most sincerely thanks:
his wife Aurora Rodriguez who has been patient and supportive beyond belief.
A Postgraduate Scholarship from Rice Research Committee and Rural Industry Research &
Development Corporation.
Dr. Eric Cother and Dr. Gavin Ash for their contribution as supervisors of this project. They have
been extremely patient, friendly and useful in providing valuable and strategic guidance
throughout my work.
Professor Jim Pratley for his crucial advice at the start of my project and his continuous support
and interest in my work.
Mr. Remy van de Ven for his extremely valuable contribution for designing experiments and
statistical advice for data analyses.
A special stipend from Charles Sturt University.
Dr. Lawrie Lewin for information on rice crops and organising other research requirements.
iv
Contents
Foreword ...............................................................................................................................iii
Acknowledgments .................................................................................................................iv
Executive Summary............................................................................................................. viii
1. Introduction ....................................................................................................................... 1
2. Literature review............................................................................................................... 2
2.1 Rice (Oryza sativa)..................................................................................................... 2
2.2 The Australian rice industry........................................................................................ 2
2.3 Weed management in rice ......................................................................................... 3
2.4 Losses due to weed competition ................................................................................ 3
2.5 Weed control.............................................................................................................. 4
2.6 Problems associated with chemical herbicides........................................................... 4
2.7 Integrated weed management systems...................................................................... 5
2.8 Biological weed control .............................................................................................. 7
2.8.1 Definitions and terminology ................................................................................ 7
2.8.2 History and general principles ............................................................................ 8
2.8.3 Biological control of weeds with pathogens ........................................................ 9
2.8.4 Biocontrol strategies using pathogenic fungi..................................................... 10
2.8.4.1 Approach ............................................................................................. 10
2.8.4.2 Augmentative approach ....................................................................... 11
2.8.4.3 Inundative approach ............................................................................ 11
2.9 The host target and its pathogen................................................................................ 13
2.9.1 The weed- Damasonium minus (starfruit)............................................................ 13
2.9.2 The fungus (Rhynchosporium alismatis) ............................................................. 13
2.9.2.1 Taxonomy............................................................................................ 13
2.9.2.2 Distribution and host range .................................................................. 14
3. Environmental and Cultural Factors Affecting Fungal Growth, Sporulation and Conidial
Infectivity......................................................................................................................... 16
3.1 Introduction .............................................................................................................. 16
3.2 Materials and methods............................................................................................. 16
3.2.1 Fungal isolation and storage............................................................................. 16
3.2.2 Sporulation experiments................................................................................... 16
3.2.2.1 Effect of media..................................................................................... 16
3.2.2.2 Effect of light cycle............................................................................... 17
3.2.2.3 Effect of temperature ........................................................................... 17
3.2.3 Conidial germination......................................................................................... 17
3.2.3.1 Effect of media..................................................................................... 17
3.2.3.3 Effect of temperature during conidial production .................................. 18
3.2.3.4 Effect of water depth ............................................................................ 18
3.2.4 Leaf disc bioassays for lesion development...................................................... 18
3.2.4.1 Effect of media..................................................................................... 19
3.2.5 Data analysis.................................................................................................... 19
3.4 Results..................................................................................................................... 19
3.4.1 Sporulation experiments................................................................................... 19
3.4.1.1 Effect of media..................................................................................... 19
3.4.1.2 Effect of light........................................................................................ 20
3.4.2 Conidial germination experiments..................................................................... 20
3.4.2.1 Effect of media..................................................................................... 20
v
3.4.2.4 Effect of water depth ............................................................................ 21

Leaf disc bioassay for lesion development ....................................................... 21
3.4.3.1 Effect of media..................................................................................... 21
3.5 Discussion ............................................................................................................... 21
4. Factors affecting disease development in starfruit......................................................... 23
4.1 Introduction .............................................................................................................. 23
4.2 Leaf disc bioassays.................................................................................................. 23
4.2.1 Materials and methods ..................................................................................... 23
4.2.1.1 Isolate virulence................................................................................... 24
4.2.1.2 Additives .............................................................................................. 24
4.2.1.3 Conidial density ................................................................................... 24
4.2.1.4 Effect of post inoculation temperature.................................................. 24
4.2.2 Results ............................................................................................................. 24
4.2.2.3 Isolate virulence................................................................................... 24
4.2.2.4 Additives .............................................................................................. 25
4.2.2.6 Effect of post inoculation temperature.................................................. 25
4.3 Glasshouse experiments.......................................................................................... 25
4.3.1 General ............................................................................................................ 25
4.3.2 Effect of plant growth stage at constant water level on susceptibility to disease27
4.3.3 Results ............................................................................................................. 27
4.3.4 Factors affecting disease development on the floating leaf stage ..................... 28
4.3.4.2 Effect of dew period ............................................................................. 28
4.3.4.3 Effect of a nutrient-humectant additive................................................. 28
4.3.4.4 Effect of water level in disease development ....................................... 28
4.3.4.5 Effect of isolates .................................................................................. 29
4.3.5 Results ............................................................................................................. 29
4.3.5.2 Dew period .......................................................................................... 29
4.3.5.3 Effect of nutrient-humectant ................................................................. 29
4.3.5.4 Effect of water level ............................................................................. 29
4.3.5.5 Effect of isolate .................................................................................... 29
4.3.6 Effect of plant growth stage after lowering the water level to the soil level on
susceptibility to disease.................................................................................... 29
4.3.7 Results ............................................................................................................. 29
4.3.8 Factors affecting disease development at the juvenile stage ............................ 30
4.3.8.2 Effect of a nutrient-humectant additive................................................. 30
4.3.8.3 Isolates ................................................................................................ 30
4.3.8.4 Dew period .......................................................................................... 30
4.3.8.5 Maximum age of juvenile plants........................................................... 31
4.3.8.6 Growth suppression over time ............................................................. 31
4.3.9 Results ............................................................................................................. 31
4.3.9.2 Effect of a nutrient-humectant .............................................................. 31
4.3.9.3 Isolate.................................................................................................. 31
4.3.9.4 Dew period .......................................................................................... 31
4.3.9.5 Maximum age of juvenile plants........................................................... 31
4.3.9.6 Growth suppression over time ............................................................. 32
4.4 Isolation from inoculated juvenile plants ................................................................... 32
4.5 Discussion ............................................................................................................... 32
5. Microscopy..................................................................................................................... 34
5.1 Introduction ................................................................................................................ 34
3.4.3
vi
5.2.1
Effect of temperature on growth and appressorium formation on cellophane paper

34
5.2.2 Conidial germination, growth and appressorium formation on starfruit floating
leaves............................................................................................................... 35
5.2.3 Effect of temperature on growth and appressorium formation on floating leaves35
5.2.4 Conidial germination and appressorium formation on juvenile leaves............... 35
5.2.5 Scanning electron microscopy.......................................................................... 35
5.3 Data analysis ........................................................................................................... 36
5.4 Results..................................................................................................................... 36
5.4.1 Effect of temperature on growth and appressorium formation on cellophane paper
36
5.4.2 Conidial germination and appressorium formation on floating leaves ............... 36
5.4.3 Effect of temperature on growth and appressorium formation on floating leaves36
5.4.4 Conidial germination and appressorium formation on juvenile leaves............... 37
5.4.5 Scanning electron microscopy.......................................................................... 37
5.5 Discussion ............................................................................................................... 37
6. Fungal interaction with chemical herbicides ................................................................... 39
6.1 Introduction .............................................................................................................. 39
6.2 Effect of herbicides on conidial germination ............................................................. 39
6.2.1 Materials and methods ..................................................................................... 39
6.2.2 Results ............................................................................................................. 40
6.3 Glasshouse experiments.......................................................................................... 40
6.3.1 Fungal interaction with MCPA .......................................................................... 40
6.3.1.1 Fungal interaction with MCPA at floating leaf stage ............................. 41
6.3.1.2 Fungal interaction with MCPA at the juvenile stage ............................. 41
6.3.2 Fungal interaction with Londax ....................................................................... 42
6.4 Results..................................................................................................................... 42
6.4.1 Fungal Interaction with MCPA .......................................................................... 42
6.4.1.1 Fungal interaction with MCPA at the floating leaf stage ....................... 42
6.4.1.2 Fungal interaction with MCPA at juvenile stage ................................... 42
3.2.2.2 Effect of light cycle................................................................................. 42
6.4.2 Fungal interaction with Londax ....................................................................... 42
6.5 Discussion ............................................................................................................... 43
7. Genetic variability between starfruit populations............................................................. 44
7.1 Introduction .............................................................................................................. 44
7.2.1 Plant materials.................................................................................................. 44
7.2.2 DNA extraction ................................................................................................. 45
7.2.3 DNA quantification and standardisation ............................................................ 45
7.2.4 PCR amplification............................................................................................. 46
7.3 Results..................................................................................................................... 46
7.4 Discussion ............................................................................................................... 47
8. Effect of fungal inoculation on weed competition ............................................................ 49
8.1 Introduction .............................................................................................................. 49
8.2 Material and methods............................................................................................... 49
8.3 Data analysis ........................................................................................................... 50
8.4 Results..................................................................................................................... 50
8.5 Discussion ............................................................................................................... 52
9. General discussion and conclusions ............................................................................... 54
10. Implications ................................................................................................................ 56
11. Recommendations...................................................................................................... 56
12. References ................................................................................................................. 57
vii
Executive Summary
Damasonium minus (R.Br.) Buch. (starfruit) is a problem weed in Australian rice fields where the
majority of rice is sown aerially on flooded fields. Rhynchosporium alismatis (Oudem) J.J. Davis
[Spermosporina alismatis (Oudem.) U. Braun] is an endemic fungus that causes disease in the weed. A
suitability study of this pathogen as a mycoherbicide for integrated management of starfruit was
carried out. These studies included growth and sporulation in artificial culture, optimisation of disease
development in the plant, microscopy of the infection process, fungal interaction with chemical
herbicides, variability of host (starfruit) populations and the effect of the disease on starfruit-rice
competition.
The influence of cultural and environmental conditions on sporulation, conidial germination, germtube elongation and conidial infectivity were investigated. Rhynchosporium alismatis sporulated on a
range of media but those based on lima bean were the most successful in producing large numbers of
viable and infective conidia. Sporulation, germination, and germ-tube elongation were greatest at 25C
and 30C. Germ-tube growth was inhibited if the conidia were submerged in water.
The effects of various factors (plant growth stage, fungal isolates, inoculum density, water level,
temperature and dew period) influencing disease development in starfruit were investigated under
controlled conditions. Infection by R. alismatis affected juvenile and adult starfruit plants differently.
Necrosis and chlorosis developed on aerial parts, mostly on leaves, of adult plants, while juvenile
plants were stunted.
Twelve fungal isolates tested caused different levels of disease that was apparent only on juvenile
plants where the minimum inoculum density necessary to cause significant level of disease was 105
conidia mL-1.
Water level during inoculation had no effect on disease development in adult plants. It influenced, but
did affect, disease on juvenile plants. Growth suppression did not occur if the plants were submerged
under water during inoculation.
It was found that a dew period was not critical for disease development. Lesion development on
starfruit leaves was enhanced at 25C and reduced at temperatures equal to, or higher than, 30C.
Microscopy of the infection sequence in starfruit leaves showed that appressorium formation was the
principal means of infection. Appressoria started to form after 4 h and infection hyphae were clearly
visible within 24 h of inoculation. The rate of appressorium formation was greater at 25C than at 30
C.
Bensulfuron-methyl (Londax) and MCPA (MCPA250) had the least effect on conidial germination
whereas propanil (Ronacil) had the greatest effect on germination. The interactive effect of
R. alismatis and sublethal doses of Londax and MCPA250 on starfruit was measured with a
synergistic interaction occurring between the fungus and Londax.
The use of simple sequence repeat primers (SSRs) for PCR analyses to study genetic variability of
different populations of the host, revealed little variation between starfruit plants from different
regions.
A glasshouse experiment examined competition between rice and starfruit, with and without fungal
inoculation. When plants were inoculated, the above ground biomass of rice did not vary as the weed
density increased; however, rice biomass declined with increasing weed density in the absence of the
fungus. These studies concluded that R. alismatis has characteristics to be a successful mycoherbicide.
It can be reproduced abundantly in artificial culture, a dew period is not essential for disease
development, it reduces competition with rice by suppressing the weed and it has synergistic effects
with at least one chemical herbicide (Londax). Therefore, the fungus has excellent prospects for
further
development
and
should
be
evaluated
in
field
experiments.
viii
1. Introduction
Damasonium minus (R. Br.) Buch (starfruit) is an Australian native plant species considered to be the
most important broad leaf aquatic weed in rice growing areas of Australia. The majority of rice in
Australia is aerially sown; this procedure particularly favours starfruit (McIntyre et al. 1991;
McDonald 1994). So far, the control of this weed has been almost exclusively reliant on the use of
only one herbicide, Londax (bensulfuron methyl) (Graham et al. 1996). This practice has the
potential to develop herbicide-resistant weed biotypes, especially in rice fields where the mechanisms
for development of resistance appear to be endemic (Hill, Smith & Boyer 1994). In fact, herbicide
resistance in starfruit has already been reported (Graham et al. 1996).
There is a growing interest in biological control of weeds using inundative applications with inoculum
of naturally occurring plant pathogens to cause disease epidemics. This approach is also known as the
bioherbicide strategy (Charudattan 1991; TeBeest 1996) The term mycoherbicide is used to indicate
that the pathogen is a fungus, and several fungal pathogens have been commercially developed as such
for weed control (Templeton & Heiny 1989).
The endemic fungus Rhynchosporium alismatis (Oudem.) J. J. Davis [Spermosporina alismatis
(Oudem.) U. Braun], was isolated from starfruit leaves and its pathogenicity to this plant and other
weed species in the Alismataceae was confirmed. Leaf necrosis was caused on leaves of D. minus,
Sagittaria guyanensis H.B.K., Alisma lanceolatum L (alisma). and A. plantago-aquatica L (water
plantain) (Cother & Gilbert 1994a,1994b). Studies by Cother and Gilbert (1994a, 1994b) indicated this
fungus may be suitable as a biocontrol agent for the management of starfruit in rice fields. Further
work by Fox (1995) showed the fungus sporulates well in artificial culture and the disease reduces
seed production in starfruit. Research is underway, at Charles Sturt University, to study the genetic
variability of the fungus and to extend the host range of this fungus to other Alismataceae weed.
The purpose of this project was to study the suitability of R. alismatis as a mycoherbicide for
integrated management of starfruit. To attain this aim the following research objectives were achieved:
1. to further investigate the effect of cultural conditions that influence inoculum production and
virulence of R. alismatis;
2. to determine factors that affect disease development in starfruit;
3. to study the microscopy of the infection process;
4. to examine fungal interaction with the most commonly used herbicides in rice;
5. to investigate the genetic variability in the host (starfruit) population;
6. to determine the effect of the disease on weed-crop competition.
2.
Literature review
2.1 Rice (Oryza sativa)

Seeds of the wild forms of rice were probably used as a source of food 10,000 to 15,000 years ago and
the cultivation of rice first began somewhere in South Asia or China 9,000 years ago. Cultivated rice
with Oryza sativa L. probably also originated in this area (Lewin & Heenan 1984). The spread of rice
cultivation to new areas brought the development of new types which enabled it to grow well in a wide
range of climates, environments and soil conditions. It is now grown as far north as Hungary and the
Czeck republic (50N) and as far south as Uruguay and New South Wales, Australia (35S), as a
rainfed hillside crop, in water up to 6 m deep, on flood plains and at altitudes up to 2400m (Brennan et
al. 1994; McDonald 1994).
Currently, more than one-third of the human population rely on rice for their daily sustenance, making
it the most important of the world's food crops. Worldwide, 530 million tonnes of paddy rice at an
average yield of 3.5 tonnes per hectare is harvested from 150 million hectares annually and provides
21 % of worlds food calorie supply. Almost 90% of the total crop is produced in Asia, where China
and India are the major producers; only a small proportion of the worlds rice is grown in temperate
regions (Zimdahl 1988; McDonald 1994).
The rice industry, worldwide, is facing the challenge of a dramatic rise in demand due to population
pressure where almost 100 million additional people must be fed each year. Thus, the global rice
production needs to be increased by 75% by 2020 to feed the world rice consumers. Except in Africa
and Latin America, there is little room to increase the area of land cultivated to rice, therefore, much
of the additional rice must come from existing land at higher average yields. The increase in demand,
however, is of high significance to temperate rice-growing areas that include the Australian industry,
where the capacity to produce rice will be fully extended if that demand is to be satisfied in the 21st
century (McDonald 1994).
2.2 The Australian rice industry

Although Australias contribution to the worlds total production is less than 0.2%, the Australian rice
industry is characterised by its high efficiency and productivity. Rice production is centred on the
temperate irrigation areas and districts of the Murrumbidgee and Murray Valleys in southern New
South Wales where some 2500 family farms produce up to 1.2 million tonnes of paddy rice annually
(Brennan et al. 1994; Ricegrowers' Co-operative Ltd. n.d.). The rice industry has grown rapidly over
the past three decades. Since the early 1960s, the area of rice has increased by an average of almost
5% per year from 22 000 ha to about 120 000 in the 1990s. During the same period the average paddy
yield has increased more than 1% per year rising from 6.3 to 8.6 t/ha, the highest in the world
(Brennan et al. 1994). The New South Wales industry has the advantage that it is free of major
diseases and pests that limit yield in other countries. It shares with other temperate regions, however,
some limiting factors including low temperatures during the production period, and weed competition
(McDonald 1994).
In recent times, economic pressure has changed the traditional rotation of one rice crop in 4-5 years of
pasture, to much closer rotation intervals with a higher proportion of rice crops (Beecher et al. 1994).
In addition, further increase in production is expected due to world demand that will exert even further
pressure on land and water. In addition costs of production should remain low enough to ensure that
temperate Australian rice is affordable for those needing it. The augmentation in cropping intensity
needs a better use of resources to establish and maintain a sustainable production system, where weed
management is considered as an essential component of such systems (McDonald 1994; Lovett &
Knights 1996).
2.3 Weed management in rice

The term weed is a concept of people, not of nature. A plant may be a weed to some and beneficial
to others. The simplest definition is that a weed is an unwanted plant referring to plants growing in
habitats modified or managed by humans, and plants that cause damage to some segments of the
human population (Norris 1992). Thus, a further definition of a weed would be that it is any plant
growing at levels that interfere with management objectives in a given area and at a given point in
time (Ross & Lembi 1985; Charudattan & DeLoach 1988; Norris 1992).
Weeds have always been, and will be, an integral part of agricultural systems (Lovett & Knights
1996). More than 1000 have been reported as weeds of rice (Baltazar & DeDatta 1992). Smith (1983)
reported that the most important plant families with the greatest number of species as weeds were
Alismataceae, Poaceae, Asteraceae, Fabaceae, Lytheraceae and Scriphulariceae. In rice crops, weeds
are the major constraint to high yield production and the species that cause problems vary with soil,
temperature, latitude, altitude, rice culture, seeding method, water management and weed control
technology (Smith 1983; Hill, De Datta & Real 1990). In Australia, changes in cropping systems have
caused a shift from grass weeds to broad leaf aquatic weeds (McDonald 1994). Four species
characterise the weed community in rice fields; Cyperus difformis L. (dirty dora), Elatine gratioloides
L., D. minus (starfruit) and Echinocloa crus-galli (L.) Beauv. (barnyard grass). These species appear
to be positively correlated with intensive rice cropping systems (McIntyre et al. 1991).
Weed management in rice is a combination of cultural and chemical tools (Baltazar & DeDatta 1992).
In Australia, chemical herbicides are used routinely for controlling major weeds in rice (McIntyre &
Barret 1985; Graham et al. 1996). This emphasises the fact that current management practices are
aimed at weed control by elimination (Watson & Wymore 1990; Cother 1994). This approach is
challenged, among others, by Glauninger and Holzner (1982), Watson and Waymore (1990), Cother
(1994), and Fischer and Ramirez (1993) who stated that a weed management or control strategy
should have as its basis on economic threshold and longer term productivity, keeping the weed
population below the level of economic competitiveness. This implies that no single approach can
solve the problem of weed populations in temperate rice. An integrated management approach that
combines chemical and biological approaches with other means of weed management is clearly
needed (Hill, Smith & Boyer 1994).
2.4 Losses due to weed competition

Weeds compete with crop plants for light, nutrients, water and space (Glauninger & Holzner 1982;
Kropff 1993). In rice crops worldwide, losses due to competitive effects of weeds are estimated at 10
to 15% of potential production (Smith 1983; Zoschke 1990; Baltazar & DeDatta 1992). Without weed
control, yield losses have been estimated to range from 16% to 86% or even 100% (Zoschke 1990;
Baltazar & DeDatta 1992; Kropff 1993). Losses are influenced by several factors which can be
summarised as: (i) competitive efficiency of weeds and rice, (ii) weed density, (iii) duration of
competition, (iv) planting method, (v) rice variety, (vi) fertility level, (vii) water management and
(viii) interaction among the preceding factors (Chisaka 1977; Smith 1983). Weeds may also interfere
with rice in the following ways (Chisaka 1977): (i) reduce quality, (ii) serving as hosts for other pests
and disease, (iii) reduce harvesting and processing efficiency, (iv) reduce efficiency of irrigation
systems by restricting the flow of water in reservoirs, canals and ditches, (v) cause consumption of
energy for their control, (vi) may be poisonous and injure animals and humans, and (vii) reduce the
value and productivity of the land. It is now understood that for rice, there is a critical weed free
period during which weeds must be controlled to prevent quantitative losses (Zimdahl 1988; Alstrom
1990).
Research reports over the past three decades have shown that losses caused by weeds are greater in
direct seeded rice than in transplanted rice (Baltazar & DeDatta 1992). This is because two or threeweek-old transplanted seedlings have a competitive advantage over newly emerging weeds. Therefore,
the dominant constraint of direct seeded rice is competition with weeds because weed control is more
difficult (Hill, De Datta & Real 1990).
2.5 Weed control

The cost-efficient control of weeds is an essential part of productive agriculture. The development of
modern chemical herbicides during the past 50 years has been a key factor in the increased
productivity of agriculture (Templeton & Heiny 1990). Weed control evaluations, however, are
generally determined solely on visual estimates of weed mortality with little or no reference to crop
yield or economic thresholds (Watson & Wymore 1990; Fischer et al. 1993).
Most of the literature agree in two concepts for more efficient and profitable weed management: (i) the
critical weed free period which is defined as a span of time between that time period after seeding or
emergence, when weed competition does not reduce crop yield, and the time period after which weed
competition will no longer reduce crop yield (Zimdahl 1988; Alstrom 1990). Research has shown (e.g.
Zimdahl 1988) that the critical weed free period in rice can last 5 weeks and may start 3 weeks after
sowing or emergence to 8 weeks after sowing or emergence, depending on several factors such as
environment, weed species and planting methods, (ii) the concept of economic threshold of weed
control that refers to weed density level in crop stands (Alstrom 1990), as a means to optimise
profitability, reduce environmental problems with pesticides and prevent or forestall the development
of herbicide resistant weeds (Hill, De Datta & Real 1990).
In rice, the reduction in labour forces and increase in direct sowing has lead to a greater reliance on
chemical herbicides, becoming a standard weed management practice for control of weeds (Barker &
Hayami 1983; Klerk et al. 1985). Rice in Australia is direct seeded in its totality and more than 80% of
it is aerially sown on flooded fields (McDonald 1994). This practice has controlled grass weeds such
as Echinocloa spp. but increased the problems associated with aquatic weeds such as C. difformis and
D. minus (McIntyre et al. 1991). In addition, the shift to the less competitive semi-dwarf varieties has
added greater problems to weed control in rice (Carey, Smith & Tablert 1994; Hill, Smith & Boyer
1994).
Since aquatic weeds assumed much greater importance, more efficient selective herbicides for their
control have been developed and these have become an integral component of rice cropping (Dilday,
Lin. & Yan 1994; Hill, Smith & Boyer 1994). The release of Londax (bensulfuron-methyl) in the late
1980's in Australia dramatically increased the effectiveness of aquatic weed control (Sanders 1994),
becoming the main, if not the only, herbicide used in rice. Two other herbicides, MCPA and 2,4-D, are
available but, due to problems such as drift, are considerably less desirable than bensulfuron (Sanders
1994, Hill et al. 1994). Londax is used routinely in controlling important weeds such as C. difformis,
Sagittaria montevidensis Cham. Et Schlecht. (arrowhead), D. minus, A. plantago-aquatica and A.
lanceolatum (Sanders 1994). However, the reliance upon Londax has not been matched with an
increase in the understanding of the biology of major aquatic weeds (Lattimore 1994; Sanders 1994).
The use of herbicides can be complemented with other cultural practices such as rice rotation, water
management and fertiliser management to enhance the crop and minimise weed growth. It seems
certain, however, that neither cultural practices nor herbicides alone can solve weed problems in direct
seeded rice. With fewer herbicides and a cultural system highly vulnerable to weed losses, an
integrated approach with better information on which to base weed control decisions will be needed to
solve problems in temperate rice (Klerk, Smith & TeBeest 1985; Hill, Smith & Boyer 1994; Sanders
1994).
2.6 Problems associated with chemical herbicides

Although the development of selective and broad spectrum herbicides has been a major contributory
factor to highly productive modern intensive agriculture, too heavy a reliance upon them has created
certain problems. According to Templeton & Heiny (1990) and Williams (1992), these problems have
arisen due to use, misuse, increasing costs, production contaminants, transport accidents and storage
and disposal of toxic wastes.
In addition, during the last decade there has been mounting public concern about the effect of
herbicides on both environment and public health. These include the effects of herbicides on surface
and ground water, spray drift and long term impact of herbicide residues in agricultural products. This
has led many governments to set specific targets to reduce the use of herbicides and the adoption of
more sustainable crop production and protection methods by encouraging integrated pest management
techniques (Evans, Rowland & McLean 1996; Lovett & Knights 1996).
The problems associated with chemical herbicides in rice fields, although basically similar to those of
other crops, are unique because rice fields are a disturbed wetland community (McIntyre, Ladiges &
Adams 1988), and the control of major weeds in the Australian rice crop is almost exclusively reliant
on one herbicide, Londax (Graham et al. 1996). One of the serious and global issues affecting many
agricultural areas around the world is the development of resistance to herbicides. This phenomenon
has increased dramatically since mid-1970s where 111 weed species were reported to have developed
resistance to various herbicides that originally were effective (Baltazar & DeDatta 1992; Holt &
Lebaron 1990). It seems that the trend towards the build-up of herbicide resistance, within or among
species, increases with the repeated use of herbicides with similar modes of action or chemically
similar herbicides exerting selection pressure (Baltazar & DeDatta 1992; Holt & Lebaron 1990;
Powles & Howat 1990). However, Powles and Howat (1990) raise the alarming issue of cross-resistant
weed biotypes, defined as a biotype that has developed resistance after selection from one herbicide
and then exhibits resistance to herbicides that differ chemically and have a different mode of action.
They reported several biotypes of two weeds in Australia and Europe presenting cross resistance to
different herbicides.
The appearance of cross-resistance in weeds means that practical control of herbicide-resistant weeds
may not be achieved simply by changing herbicides and this phenomenon may become evident in
other weed species (Powles & Howat 1990). This likelihood raises an alarm for the rice growing
industries since it appears that the mechanisms for development of resistance are endemic to many, if
not most, rice fields. In addition, herbicide resistance in rice fields has an aggravating fact since the
choice of herbicides in rice crops is extremely limited. In only five years after the introduction of
Londax, four weeds were found to be resistant to Londax in California (Hill et al. 1994). In
Australia, resistance to Londax has been reported in major weeds of rice such as D. minus, C.
difformis and S. montevidensis (Sanders 1994, Graham et al. 1996).
Herbicide drift is another problem associated with the use of chemical herbicides. Temperate rice may
be grown amid a diverse array of broadleaf crops intolerant to Londax and other rice herbicides (Hill
et al. 1994). Phytotoxicity and herbicide injury to rice have also been commonly reported. The risk of
herbicide injury in rice is greater in water-seeded than dry-seeded rice, Bensulfuron (Londax) may
inhibit root growth, therefore inhibiting growth in young rice plants. This phenomenon can decrease
the number of established seedlings, leading to reduction of yield (Yim & Bayer 1996).
Moreover, there is evidence that chemical herbicides have a significant impact on both composition
and dynamics of microbial, faunal and floral communities of rice fields (Wardle 1992; Roger et al.
1994). Therefore, they can reduce soil fertility through effects on microorganisms responsible for
maintaining the fertility. Roger et al. (1994) reported that herbicides affected more microorganisms
and their activities than did insecticides and fungicides. Furthermore, the use of herbicides in rice
fields create conflicts due to multiple uses of water such as drinking, fishing, irrigation and cleaning
(Charudattan, DeValerio & Prange 1990). Fish killed in agricultural drains of the Sacramento Valley,
California, in 1970s and off-taste in municipal drinking water of the city of Sacramento in the 1980s
were attributed to herbicides used in rice fields (Hill et al. 1994).
2.7 Integrated weed management systems

The current challenge of agricultural production systems is to produce economically rewarding crop
yields while maximising local regional and global environmental sustainability. Weed management
plays a vital role in this process to maintain the balance between environmental, economic, social and
political concerns (Swanton & Murphy 1996; Lovett & Knights 1996).
A broad approach to systems management, referred to as Integrated Weed Management Systems
(IWMS), has been developed (Templeton & Heiny 1990; Smith 1991; Lovett & Knights 1996). IWMS
is usually considered as part of Integrated Pest Management Systems (IPMS), that is an approach in
which principles, practices, methods, materials and strategies are chosen to control pests while
minimising undesirable results (Shaw 1982; Smith 1991). Elmore (1996) and Swanton and Murphy
(1996) however, argue that it should be considered as part of a total management system called
integrated crop management, emphasising a total crop management approach and not just pests.
IWMS is a systems approach (Swanton & Murphy 1996), encompassing different methods of weed
management together with effective education and extension of the management components that
takes into account the whole range of issues from agricultural production, to economic losses, risks to
human health and the environment and energy requirement (Shaw 1982; Blair & Parochetti 1982;
Lovett & Knights 1996). The objectives of IWMS are the reduction of losses caused by weeds, costs
of control, energy and labor requirements, assure adequate supply of food, improve environmental
quality and maximise producer profits (Penner 1982; Shaw 1982; Fischer et al. 1993). Thus, it is a
directed agroecosystem approach for the management and control of weeds at threshold levels that
prevent economic damage in the current and future years (Shaw 1982, Swanton & Murphy 1996),
reconciling the differences between short term economic gain of the landholder and long term
environmental stability (Cother 1994).
IWMS combines the use of several methods (Chisaka 1977; McWhorter & Shaw 1982; Smith 1991;
Lovett & Knights 1996) which can be summarised as:
i) Ecological methods that include multiple-pest-resistant, high yielding, well-adapted crop cultivars
that resist weed competition, fertiliser management to give the crop competitive advantage, careful
crop rotation, optimum crop plant population and the use of crop cultivars that form a canopy for
shading early season weed growth, are viable parts of the system as well as the use of allelopathy in
crop plants to interfere with weeds. Such methods also include the use of judicious irrigation practices.
ii) Physical methods include appropriate cultivation, field sanitation and harvesting methods that do
not spread weed seeds, appropriate seedbed tillage and seeding methods that enhance crop growth
while minimising weed growth. Minimum tillage, direct drilling and zero tillage to reduce disturbance
of the soil systems which can be enhanced by chemical methods will also fall into this category of
control method.
iii) Chemical methods of weed management include the use of herbicides and genetically engineered
herbicide-resistant crops. However, it is important that herbicide-resistant crops are not promoted as a
panacea, but as a component of IWMS.
iv) Biological methods usually include the use of organisms such as pathogens, insects, plants and
herbivores. The use of bioherbicides (see below) may present the best choice of biological methods in
cultivated crops.
In addition, all these management strategies will be integrated with other pest management systems
(e.g. insect pest management) that are incorporated into agroecosystems on farm areas or regions
(McWhorter & Shaw 1982; Elmore 1996; Swanton & Murphy 1996).
Although IWMS integrates all those preventive, cultural, chemical and biological practices, there is a
general agreement among weed scientists that chemical herbicides will continue to be key components
of IWMS. In addition, many of the elements needed for an effective IWMS are currently limited or
unavailable (McWhorter & Shaw 1982; Hill 1982; Lovett & Knights 1996). In order to meet both
short and long term goals, technological advancement is needed and this should include the use of
bioherbicides (Bewick 1996). For instance, more than 140 herbicides are available for control of
weeds; in contrast, only three bioherbicides are registered for use in crop sites for rice, soybean., citrus
and wheat (Charudattan, DeValerio & Prange. 1990; Smith 1991; Makowski & Mortensen 1992).
Research and development of new management practices with presently available chemical herbicides
would be much more cost effective than equivalent research inputs into developing new herbicides
(McWhorter & Shaw 1982), and bioherbicides offer greatest potential for alternative weed
management (Bewick 1996; Kremer & Kennedy 1996). The objective of biological control is not the
eradication of the weed but to keep the weed below an economic loss level, since eradication is seldom
feasible (Huffaker 1957; Aldrich 1984; Burge 1988).
It is important to highlight that the biological control methods, such as bioherbicides are not to replace
chemical herbicides but to supplement their judicious use (Templeton 1983). The integration of both,
biological and chemical herbicides into ecologically based weed management systems is an essential
process if agriculture is to become sustainable (TeBeest 1990; Watson 1992; Bewick 1996).
2.8 Biological weed control

2.8.1
Definitions and terminology
Biological weed control is the use of a natural process to decrease damage to useful plants by weeds
(Woods & Way 1988). It is based on the premise that biotic factors significantly affect the distribution
and abundance of plant species (Watson 1991).
Some definitions of biological weed control (e.g. Jutsum 1988) cover a broad spectrum of approaches
including the use of obligate parasites and pathogens, facultative parasite and pathogens, competitors,
toxin-producing pathogens, toxin produced by pathogens, non-toxic behaviour-modifying chemicals
and even the use of selective agrochemicals. This view differs from the most common definitions of
biological control:
Huffaker (1957): Biological control is the use of natural enemies to reduce the densities of weed to
levels largely noninjurious to mans interest.
Debach (1964): Biological control is the action of predators, parasites or pathogens in maintaining
another organisms population density at lower average than would occur in their absence.
Ross and Lembi (1985): Biological weed control is the use of natural enemies to reduce weed
populations to economically acceptable levels.
Watson and Wymore (1990): Biological weed control is the deliberate use of natural enemies to
suppress the growth or reduce the population of a weed species.
Debach and Rosen (1991): Biological control is the use of natural enemies to reduce the damage
caused by noxious organisms to tolerable levels.
Bruzzese (1993): Biological control is the use of natural enemies of a pest to control its population to a
level where it is no longer considered a problem.
Shepherd (1993)): Biological control is a process by which the natural enemies of a plant are
introduced to control that plant.
Crump, Cother and Ash (1999): Biological control is the use of living organisms to suppress a pest
population, making it less abundant and thus less damaging than it would otherwise be.
The list of definitions is longer but most of them coincide in the use of living organisms,
synonymous to natural enemy or biotic agents, to keep the weeds below the level of economic
competitiveness.
Other biologically based methods are sometimes referred to as biological control. These may include
the use of sterilised males to control insect pests or the control by breeding and selection of resistant
crops. According to (Woods & Way 1988) and Debach and Rosen (1991) these methods are based
entirely on different principles and should not be included in biological control terminology. This
argument is logical if we consider biological control as the use a natural process that involves the
utilisation of what is called natural enemies or biotic agents. Under such consideration falls the use of
herbivores, as well as the use of weak or non-competitive plant species as biocontrol agents (Aldrich
1984). Auld, Menz, & Tisdell 1987) however, discriminate against vertebrates and plants and refer
to their use as an ecological method.
Nevertheless, taking into account this consideration of crop, weed and biotic agent, the definition,
therefore, may include vertebrates, invertebrates, microorganisms and plants that are not the weeds
and the useful plants to be protected (Woods & Way 1988; Debach 1964). In addition, the terms
living organisms and natural enemies may be referred to as biotic agents or biocontrol agents
(Aldrich 1984; Watson 1991).
2.8.2
History and general principles
The two major disciplines involved in the historical development of biological control of weeds are
entomology and plant pathology (Charudattan & DeLoach 1988; Cate 1990). The management of
pathogens and insects is the basis of biological control and has served as an important component of
weed management in agriculture (Charudattan & DeLoach 1988; Kremer & Kennedy 1996). The term
biological control, however, was first used by Smith (1919) and referred to the use of natural
enemies to control insect pests. Nevertheless, the concept of biological control of weeds is not new. In
1795 the cochineal insect, Dactylopius ceylonicus was introduced onto cactus in India in the mistaken
impression that it was useful dye-producing species. The insect moved from its intended cactus host to
another exotic species, Opuntia spp., which had become a weed and dramatically reduced its density.
This insect was subsequently put to good use through redistribution of the insect to other cactus
infested parts of India. The first major program of biological weed control was in 1902 and was aimed
to control Lantana camara (L.) in Hawaii by insects collected in Mexico (Waage 1992; Shepherd
1993).
In recent years, the stature of biological control as a viable practice in modern agriculture has
increased greatly. Biological control of weeds is seen as an economical, effective and environmentally
sound method of weed control. There are several reasons for this accelerated interest in biological
control such as the public awareness of the potential ecological hazards posed by the use of chemical
pesticides, demand for more environmentally acceptable methods, and pressure to decrease greatly the
release of chemicals into the environment (Burge 1988; Watson & Wymore 1990; Medd 1992;
Templeton & Heiny 1990; Combellback 1992). Biological control is an alternative as well as a method
with which chemicals can be integrated, so decreasing the loading on the environment with
undesirable pesticides. Its great virtue is that it makes use of natural mechanisms by which damage to
plants is kept at minimum levels. It is based on the fact that all pests are affected by one or more biotic
agents (Huffaker 1957; Templeton 1983; Aldrich 1984; Woods & Way 1988). In natural systems, the
populations of plants and biotic agents are in an equilibrium phase which may shift a little one way or
the other according to factors such as the environmental conditions (Cullen & Hasan 1988); successful
biological control depends on the manipulation of these biotic agents by various means to facilitate the
reduction of pest population below the level of economic threshold (Burge 1988). Biological control
is, therefore, a scientific endeavour that deals fundamentally with interaction between organisms (i.e.
weeds and biotic agents) which can range from parasitism, predation or antagonism (Cate 1990).
The strategies to achieve the desired objective may vary according to the nature of the problem
(Charudattan 1990a). Firstly, a weed may not be affected by any biotic agent in the system because it
has been introduced to an area where it has no natural enemy. Another scenario is when the natural
balance between the weed, introduced or native to the area, and biotic agents has been broken due to
human action and/or natural phenomena. Moreover, it is also possible that the existing natural balance
still gives a higher plant population than required by the management objectives (Burge 1988; Cullen
& Hasan 1988; Ehler 1992; Cullen 1992).
Although there is some difference between entomology and plant pathology in terminologies, there are
common criteria on the strategies used to establish or re-establish the desired balance between weeds
and biotic agents, in the broadest term it may be achieved by:
i) classical approach that involves the introduction of the biocontrol agent for release;
ii) habitat management that is for conservation and enhancement of existing biocontrol agents;
iii) periodic release of biocontrol agents that have been cultured artificially or under controlled
conditions (Charudattan 1985; Cate 1990; Debach & Rosen 1991; Templeton 1990).
The most common approach to using plant pathogens for weed management is the introduction and
periodic release of biocontrol agents (Charudattan 1985; Charudattan & DeLoach 1988; Charudattan
1990a).
2.8.3
Biological control of weeds with pathogens
Although the biological control of weeds with pathogens is a relatively new concept compared to the
use of insects, they have been successfully used in weed management, challenging the domination of
insect agents in control programs (Aldrich 1984; Watson 1991; Waage 1992). The use of pathogens
for weed management could offer specificity in control, environmental safety, avoidance of herbicideresistant weed biotypes and they may also be less costly to develop than chemical herbicides
(Charudattan 1990a).
However, disease development in plants is often limited by various restraints (Watson & Wymore
1990). Strategies for biocontrol of weeds using plant pathogens include (i) classical approach (as
mentioned earlier) and (ii) periodic release that can be divided in (a) augmentative approach and (b)
inundative or bioherbicide approach involving a massive dose of inoculum applied in the same manner
as a chemical herbicide (Altman, Neate & Rovira 1990; Charudattan 1991). These strategies will be
discussed in detail later.
Although pathogens are the most extensively studied aspect of microbial control of weeds, it is
important to remember that they are only one facet of the tripartite biological interaction of host,
pathogen and environment that is essential for disease development. Much of the theoretical
information concerning population dynamics of plant pathogens for weed control is obtained from
epidemiology of crop pathogens. Epidemics occur when host and environmental conditions satisfy the
requirements of a pathogen for an extended period of time (Templeton 1983; Huber & Gillespie 1992;
TeBeest, Yang & Cisar 1992). For conventional epidemiology, the emphasis is on prevention of
epidemics in homogeneous host populations. According to TeBeest, Yang and Cisar (1992) the
emphasis for weed biocontrol is on the induction of diseases or enhancement of an epidemic in
heterogeneous populations by determining and manipulating epidemic constraints. The study of each
of these factors (i.e pathogen, environment and host) is therefore important for the success of this type
of biocontrol (see also Bos & Parlevliet 1995).
Pathogen: Major constraints include pathogenicity, dissemination or dispersal, reproduction, survival
and specificity. It should be noted however, that each constraint is of differing importance according
to the strategy used in biological weed control. For example, high dispersal and survival rate are
critical factors in the classical approach. On the contrary, low ability for dispersal and survival are
required in the inundative approach (Charudattan 1989; Watson 1991; TeBeest, Yang & Cisar 1992).
Among the different pathogens it is possible to use, fungi have been the focus of interest. They have
been used in attempts to control weeds in various parts of the world since the 1950s (Auld 1990).
Fungi have multiple advantages over other pathogens such as viruses or bacteria. They are the most
commonly encountered pathogens of plants, many are destructive and they offer more specificity to
the host than bacteria or viruses. In addition, unlike viruses or bacteria which usually enter the host
through openings or by vectors, fungi can actively penetrate the host. (Powell & Faull 1989; Altman,
Neate & Rovira 1990;.Charudattan 1990a) .
Environment: Initial infection by the pathogen as well as the speed of disease development, depends
on environmental conditions (Auld & Morin 1995). Perhaps the most important constraint on disease
development in plants is moisture (Templeton & Heiny 1989). Free moisture and relative humidity
play key roles during many epidemiological episodes. Temperature is another restricting factor,
although it has not been considered as important as moisture (TeBeest, Yang & Cisar 1992). Light, the
UV radiation component of sunlight and nutrient status of the soil may also influence the interaction
between pathogen and host. These environmental factors influence the physiology of both plant and
pathogen (TeBeest 1993; Auld & Morin 1995; Bos & Parlevliet 1995).
In aquatic environments, especially for submerged plant species, biological control using pathogens
may be constrained by the extremely high level of inoculum that is needed to cause disease epidemics
(Joye 1990).
The importance of environmental constraints varies with the biocontrol approach. For instance,
temperature seems to be more important in the classical approach than in the bioherbicide approach,
whereas nutrient status of the soil is more critical in the bioherbicide approach (TeBeest 1993; Auld &
Morin 1995).
Host: Although variation in susceptibility to a pathogen or to a particular strain is normal (Cullen &
Hasan 1988; Ahmed, Mundt & Coakley 1995), according to Charudattan (1985) the development of
resistance of weeds to pathogens is not a serious concern, compared to chemical herbicides. This is so
because genetic variation in weeds necessary to achieve resistance to a pathogen would be ostensibly
more substantial than what is needed to develop resistance to a chemical herbicide. Chemicals have
such specific sites of attack on cellular biology of the pest that a single mutation in the pest genome
has been sufficient to remove the sensitive site (Burge 1988). However, genetic variation within and
between weed populations may be an important limiting factor to biological control, since in a weed
population there may be resistant biotypes that will increase in population (Auld & Morin 1995).
Moreover, host resistance can change as the plant develops. Many studies have shown that although
some pathogens, especially foliar pathogens, can kill seedlings, the same pathogens may only reduce
reproductivity as plants mature (Charudattan 1989; TeBeest, Yang & Cisar 1992).
2.8.4
Biocontrol strategies using pathogenic fungi
Traditional entomological terminologies to classify strategies of biological weed control have been
changed slightly aiming to a more precise and adequate description of approaches involved when
using pathogens as biocontrol agents (Charudattan 1985; Charudattan 1991; Morin 1993).
2.8.4.1 Approach
The classical approach has its basis in entomology (Cate 1990). A more pathology-based terminology
for this approach is called the inoculative strategy. Basically the inoculative strategy using pathogens
is similar to that utilising insects and involves the importation of the biocontrol agent to an area where
the weed, normally an introduced species, exists but has no natural enemy. Therefore, the pathogenic
fungi, usually obligate parasites, for inoculative control are sought from the original geographic range
of the weed (Mortensen 1986; Adams 1988; TeBeest 1990). The reason for the feasibility of an
inoculative strategy to control introduced species is that the weed which has been spatially separated
from its native pathogen for some time, tends to lose its resistance to the pathogen. Therefore, when
reintroduced to its long lost enemy the plant tends to be vulnerable; (Charudattan 1990a, b). The term
inoculative is used similar to the introduction of a small sample of a substance into a large body to
start a massive response. The inoculative pathogen is merely released over small weed infestations
relative to the total infestation. Hence, a small dose of inoculum is used to eventually suppress the
weed population over a large area of land (Charudattan & DeLoach 1988; McRae 1988). However, the
process is very slow, since it depends on the gradual increase in disease which may take several years,
but, because of its relative low cost, the inoculative strategy is best suited to manage weeds that are
distributed over vast areas that yield low marginal or economic returns (Altman, Neate & Rovira 1990;
Charudattan 1990b).
The pathogenic fungi used in this strategy must have high capacity for self-dissemination, which must
be sufficient to allow rapid spread through the target plant population, without human intervention.
Therefore, the most suitable areas for inoculative approach are undisturbed areas such as pastures,
rangeland and waterways where there is no interference with the spreading of the disease (Mortensen
1986; Adams 1988; Charudattan 1985; Charudattan & DeLoach 1988). Another important
characteristic of the pathogen is that it must be able to self perpetuate and be capable of maintaining a
self-sustaining epidemic; this implies that the inoculum of the fungus must survive from year to year.
Thus, the most suitable fungi in an inoculative approach are those with wind-dispersed propagules
such as rusts which cause endemics after initial release (Mortensen 1986; TeBeest, Yang & Cisar.
1992).
The inoculative approach is an self-regulating mechanism where the level of disease increases with an
increase in weed population and revert to lower levels with a reduction in weed population
10
(Charudattan 1985; Watson 1992). Since the biocontrol agent is left without further human
intervention and once it is released it can not be stopped, the pathogen must be specific to the weed to
be controlled (Watson 1991; Waage 1992).
The most dramatic demonstration of the inoculative strategy has been the control of skeletonweed
(Chondrilla juncea L.) by a rust fungus (Puccinia chondrillinia Bubak. & Syd) in Australia. After its
introduction from the Mediterranean region, skeletonweed became a serious weed in cereal crops and
rangelands of Australia. Following considerable research on virulence and host specificity the rust,
from the Mediterranean region, was found to be the most virulent and specific to the most common
form of skeletonweed (narrowleaf) and was subsequently introduced into the continent. After
inoculative release the rust spread rapidly and reduced the populations of narrow-leaf form of
skeletonweed, while the other non-susceptible forms of skeletonweed generally increased. Thus,
further research to find suitable biocontrol agent was carried out (Watson 1991; Charudattan &
DeLoach 1988).
2.8.4.2 Augmentative approach

This approach falls into the category of periodic release and can be regarded as midway between the
classical and the inundative approach (Charudattan 1985).
In the augmentative approach pathogens may be native or naturalised in their respective region
causing endemic disease, and similar to the classical approach, they can self-disseminate causing
certain level of self-sustaining epidemics after inoculation. Some human intervention, however, is
needed for inoculum distribution. The amount of inoculum required is larger than for the inoculative
approach and less than the inundative strategy. The area that needs to be inoculated is also larger than
in the classical approach. Using this strategy large amounts of inoculum are stockpiled (e.g. inoculum
may be harvested from infected plants under controlled conditions) applied over several hectares.
However, there are limitations to the amount of inoculum that can be gathered and applied and it is
necessary to periodically apply a surplus of inoculum (Charudattan 1988; Charudattan 1991).
Therefore, it differs from the inundative approach in that the inoculum is not mass produced in
artificial culture nor applied as an inundative dose (Charudattan 1991). This strategy is slow acting and
depends on a gradual build up of epidemic, therefore, its suitability for agricultural systems is limited
(Charudattan 1985).
An example of this approach is the control of yellow nutsedge (Cyperus esculentus L.) by the rust
fungus Puccininia canaliculata (Schw.) Lagerh. (Charudattan 1991). In northern Australia another rust
fungus Maravila cryptostegiae (Cummins) for controlling rubbervine (Cryptostegia grandiflora R.Br.)
is being assessed (Tomely & Hardwick 1996). The inoculum of this fungus can be bulked up in vivo
and applied in suspension over several hectares..
2.8.4.3 Inundative approach

The inundative strategy is also known as the bioherbicide approach and is best suited to control
agricultural weeds in IWMS (Templeton 1982). It involves the repeated application of a massive dose
of inoculum of a non-self-sustaining pathogen over a specific area, such as a paddock, where the weed
control is desirable (Baker & Henis 1990). The term mycoherbicide is used instead of bioherbicide to
indicate the biocontrol agent is a fungus. Crump, Cother and Ash (1999), however, propose, the term
'bioherbistat' or ''mycoherbistat' for an inundative biocontrol agent that does not kill the target plant but
suppresses the weed by reducing its competitive ability.
Many opportunities exist in temperate areas to increase disease pressure on weeds to the extent that
they are either killed or debilitated enough to be considered controlled. Thus, from an ecological
perspective annual weeds in annual crops or in waterways in temperate regions are ideal targets for
mycoherbicides (Templeton & Heiny 1989). This concept was introduced by Daniel et al. (1973) who
showed that an endemic fungus that can sporulate abundantly in artificial culture may be used as a
biological herbicide in the same manner as chemical herbicides. These authors based their work on an
11
endemic anthracnose disease of northern jointvetch [Aeschinomene virginica (L.) B.S.P.] caused by
the fungus Colletotrichum gloesporioides (Penz.) Sacc. f sp. aeschynomene. This mycoherbicide,
named Collego, was registered in 1982 for commercial use to control northern jointvetch in rice and
soybean as a liquid formulation in the USA (Templeton 1986).
The first commercially available mycoherbicide was DeVine registered in 1981. The product consists
of a liquid concentrate of chlamydospores of Phytophthora palmivora (Butl.) which is used to control
strangler vine [Morrenia ordorata (H. & A). Lindl] in Floridas citrus groves. It took, however, 10
years for the third mycoherbicide to be commercially registered. The fungal pathogen Colletotrichum
gloeosporioides (Penz.) Sacc. f sp. malvae , discovered in 1982, was registered in Canada under the
tradename Biomal to control of Malva pusilla Sm. (round-leaved mallow) (Makowski & Mortensen
1992). There are several other fungi that have the potential as mycoherbicides with proven ability for
weed control but which have not been developed further that may be termed orphans (Templeton
1992a). Despite the fact the mycoherbicide concept has received recognition around world, and there
are several orphaned mycoherbicides, the development of only three mycoherbicides since 1982 is
not big progress (TeBeest 1992; Klein 1992 ).
The three main steps, discovery, development and deployment of a mycoherbicide, mean that the
inundative approach is more time and resource consuming than the other strategies. However, there
are commercial incentives since a mycoherbicide can be commercially produced with a subsequent
return on investment (Kenney 1986; Templeton 1982; Templeton 1990).
Kenney (1986) argues, from an industrialists perspective that the most important constraint on the
development of bioherbicides is return on investment and every step in the discovery, development
and deployment should be financially justified. Templeton (1990b) however, argues that human and
environmental hazards should also be taken into account and since the private sector is not willing to
invest, public sector investment on research is required (Pennell 1994).
Characteristics of mycoherbicides
Since the purpose of mycoherbicides is a quick economic control of agricultural weeds, they must
have a number of characteristics. A mycoherbicide should be highly virulent and destructive; this does
not necessarily mean killing the plant as suppression or debilitation may achieve the same desired
level of control (Templeton & Heiny 1990; Watson & Wymore 1990). Charudattan (1989) states the
guidelines for assessment of the efficacy of a mycoherbicide in terms of (i) the speed of weed control,
(ii) amount or the level of weed control and (iii) ease with which a mycoherbicide can be used that
refers not only to application tools but also the range of environmental conditions in which it can be
used.
The pathogen must easily produce abundant durable and genetically stable inoculum in artificial
culture, to the extent that it may be produced on a commercial scale (Altman, Neat & Rovira 1990;
Trujillo 1992). The mycoherbicide can be incorporated into ongoing IWMS together with other weed
management technologies such as chemical herbicide for a more efficient weed management
(Templeton 1982; Charudattan 1985). Interaction between herbicides and pathogens have
demonstrated that disease can be increased on plants when sublethal rates of herbicides are used. Very
low doses of herbicides can act as metabolic inhibitors of host defence and when synergistically
combined with a pathogen would achieve weed control as part of an IWMS (Lovett & Knights 1996).
Host specificity of a mycoherbicide is important. Pathogens with unrestricted host range are not
recommended for biological control. From an economic point of view, however, large commercial
companies may not be interested in extremely host specific mycoherbicides (Cother, E.J. pers. comm.
1999).
According to Charudattan (1990a) a mycoherbicide is not expected to survive beyond the season in
which it is applied. In fact, lack of survival is required for safety. argues that short-term survival would
minimise inoculum build up in treated areas and the potential for long-term genetic interactions in the
field. Poor capacity for self-dissemination is another desirable characteristic of mycoherbicides
12
(Trujillo 1992). This is important since weed control is desirable only in a specific area, without
affecting the plants beyond that area.
In addition, large commercial companies may not be interested in mycoherbicides due to the smaller
margin of benefits compared to chemical herbicides. As an example the mycoherbicide DeVine can
survive long after the time in which it is applied, controlling the weed over five consecutive growing
seasons (Altman, Neate & Rovira 1990). This great advantage has made DeVine less attractive for
large commercial companies. Despite this DeVine can be considered as a very successful
mycoherbicide from a practical and scientific viewpoint and is still available to growers.
2.9 The host target and its pathogen

2.9.1 The weed- Damasonium minus (starfruit)
Damasonium minus (starfruit) is an Australian native broad-leaved aquatic monocot plant belonging to
the family Alismataceae. It is the most significant broad leaf weed in aerially sown rice of New South
Wales, Australia (McIntyre, Ladiges & Adams 1988; McIntyre et al. 1991).
It is an annual or biannual plant, bright to olive green, fleshy, hairless with oval to lance shaped leaves
with a long petiole supporting the leaf on the water surface. In dense infestations and shallow water
the leaf blade may held upright. Flowers are small and white to pink on an erect stem. Star-shaped
fruits contain small black seeds 1-2 mm long with one seed per fruit (Cox 1984). The plant is
widespread, but not particularly abundant in natural and semi-natural wetlands. Plant numbers have
built up considerably in rice fields and D. minus was recognised as an important weed in the 1960s.
McIntyre, Ladiges and Adams (1988), and McIntyre et al. (1991) reported that D. minus occurred in
more than 70% of sites in surveys of rice fields and it can be considered ubiquitous in rice crops in
NSW. The herbicide Londax (bensulfuron) is employed routinely for its control. Little is known about
the biology of starfruit and research in the School of Agriculture at Charles Sturt University, Riverina
has been underway to study the ecology and biology of the weed.
2.9.2 The fungus (Rhynchosporium alismatis)

2.9.2.1 Taxonomy
The taxon, a Hyphomycete, was originally described as Septoria alismatis by Oudemans (1875) and it
has since been reclassified several times as shown in Table 2.1.
According to Punithalingam (1988) the fungus is correctly known as Rhynchosporium alismatis with
the following description:
Mycelium subcuticular (as well as ramifying the leaf tissue and becoming inter- or intra-cellular),
composed of branched, septate hyphae forming stromata which are composed of short, broad hyphal
cells from which conidia are formed from short conidiogenous cells. Conidiophores absent.
Conidiogenous cells hyaline, acrogenous, percurrently proliferating. Conidia holoblastic, hyaline,
straight or slightly curved towards the ends or slightly beaked medianly 1-septate 16-19 x 2.5-3 m.
Braun (1995) created the genus Spermosporina designating this fungus as Spermosporina alismatis and
describes the fungus as follows:
Mycelium immersed, hyphae hyaline, septate, branched, forming intra-epidermal hyphal aggregation
(stromata), rarely substomatal, variable, almost lacking to well developed sometimes confluent.
Conidiogenous cells solitary or in small groups, arising from and integrated in hyphal
aggregations, erumpent through the cuticle, ampulliform, hyaline, smooth with narrowed apex,
rounded to truncate. Conidia solitary, ellipsoid-ovoid, subcylindric, rarely fusiform (8-) 10-20(24) x 2-4.5 m, 0-2-septate, rarely somewhat constricted at the septa, hyaline, smooth, apex
mostly rounded, base rounded or attenuated.
13
Table 2.1. Different taxonomical classification of Rhynchosporium alismatis

Classification and authority
Source
Ascochyta alismatis Ell. & Eva.
(Punithalingam 1988)
Ramularia alismatis Fautrey.
Rhynchosporium alismatis (Oudem) J.J. Davis
Didymaria alismatis (Oudem.) J.J. Davis
Ovularia alismatis Pass.
(Braun 1995)
Didymaria aquatica Starback
(Braun 1995)
Cylindrosporium baudysianum Sacc.
(Braun 1995)
Didymaria cinerea Starback
(Braun 1995)
Braun (1995) differentiates the genus Rhynchosporium from Spermosporina as the former being
graminicolous. Nevertheless, this fungus is still referred to as Rhynchosporium alismatis by the
herbarium of the New South Wales Agriculture where all known isolates of the species in Australia
are held. Therefore, throughout this work, the fungus will be referred to as R. alismatis.
2.9.2.2 Distribution and host range

The first record of the fungus was in Netherlands on leaves of A. plantago-aquatica in 1874
(Punithalingam 1988). It also occurs in Asia, Africa, Europe and North America on A. plantagoaquatica and Sagittaria heterophylla Bert. ex Steud. The first record of the fungus in Australia was on
A. plantago-aquatica in metropolitan Sydney and later it has also been observed on A. lanceolatum,
and D. minus (Cother, Gilbert & Pollock 1994). It also occurs on A. loeselii Gorsk (Central Asia), A.
gramineum Lej. (Europe), A. latifolia (Asia, Europe and N. America), and S. subcordatum Raf. (N.
America) (Braun 1995). According to Braun (1995), the fungus occurring on Sagittaria species are
classified as Spermosporina sagittariae (Bres.) U. Braun.
The incidence of disease caused by R. alismatis in rice growing areas of NSW seems much higher in
A. lanceolatum and A. plantago-aquatica than in D. minus, which may suggest that the fungus has
recently expanded its host range to the latter. It is possible that the fungus was introduced to Australia,
possibly with A. lanceolatum or several other Sagittaria species regarded as ornamentals (Cother,
Gilbert & Pollock 1994).
Using the inundative approach, R. alismatis (a wet-spored fungus) has the potential as a biocontrol
agent to suppress D. minus and other hosts in the same family. This potential has been confirmed by
Cother and Gilbert (1994a, 1994b), who observed the effect of R. alismatis on D. minus and the other
species in terms of the reduction in the plant biomass after inoculating young plants with spore
suspensions. They suggest that leaf infection by R. alismatis early in the growing season could have a
physiological effect on plant growth similar to that caused by leaf removal or by chemical herbicides.
Cother, Gilbert and Pollock (1994) described the lesions caused by the fungus as dark brown, necrotic
spots (1-3 mm diameter) on mature leaves usually with a pale green to yellow halo. On the main vein
lesions are lens shaped and may coalesce to form an elongated lesion. Symptoms may also occur on
petioles and stalks. Further work carried out by Fox (1995) concluded that the fungal inoculation of
starfruit reduces significantly total seed production of the plant. In these reports the inoculum was
spore suspensions produced abundantly under artificial conditions in solid or liquid shake cultures.
Other hosts of the fungus are A. canaliculatum A. Br. & Bouche., Echinodorus rostratus (Nutt.)
Engelm., S. brevisrostra Mack & Bush, S. guyanensis H.B.K. and S. pygmaea Miq. These species are
considered weeds of rice in different parts of the globe (Cother & Gilbert 1994b).
Host range studies of R. alismatis on 28 species of aquatic plants in the Alismataceae and related
families as well as 39 cultivars of 25 species of agriculturally significant species in the rice growing
areas of Australia have been carried out (Cother 1999). These studies have indicated that although
fungal inoculation caused apparent symptoms on 11 aquatic species (including A. lanceolatum and S.
guyanensis), it was re-isolated only from 5 of these species. Perhaps one the most prominent aspect of
14
this studies were disease symptoms and re-isolation from some cultivars of cucumber and rockmelon,
tomato and soybean. However, the presence of lesions did not appear to have any apparent effect on
the plant development. It must be emphasised the host range studies were carried out under extremely
severe experimental conditions that tend to predispose plants to infection resulting in artificial
expansion of host range which may not occur under normal conditions. In addition, there are no
records of the fungus as a pathogen of any of these economically important species (Cother 1999).
15
3.
Environmental and Cultural Factors

Affecting Fungal Growth, Sporulation
and Conidial Infectivity
3.1 Introduction
One of the most important attributes of a mycoherbicide is the production of abundant, viable and
infective inoculum in artificial culture (Templeton 1982). Auld and Morin (1995) and Jackson and
Schisler (1992) consider the lack of commercial success of biocontrol agents is largely due to the
difficulty in producing the agents in artificial cultures. Production of propagules in liquid culture is an
important characteristic of a mycoherbicide since liquid fermentation is the most economical and
simplest way to produce large quantities of inoculum (Auld 1993). The production medium appears to
be a significant factor affecting the attributes of the inoculum of a biocontrol agent (Jackson &
Schisler 1992). Some environmental factors during inoculum production are said to effect disease
severity. Shabana, Charudattan and Elwakil (1995) reported some effect of light regime during the
inoculum production of Alternaria eichhorniae Nag Raj & Ponnappa in disease development on
waterhyacinth [Eichhornia crassipes (Mart.) Solms].
It has been shown that R. alismatis can produce large amounts of inoculum readily in liquid cultures in
a short period of time (Jahromi, Ash & Cother 1998). Further work by Cother & van de Ven 1999
found no significant trend between conidial production and the pH of the media. However, cultural
conditions for R. alismatis need to be studied further. The objectives of these experiments were to
study the effects of several factors affecting in vitro sporulation, conidial germination and germ-tube
elongation of R. alismatis and subsequent influence on lesion development. This work was focussed
on agar culture since it is the most convenient method for small-scale inoculum production for
research purposes.
3.2 Materials and methods

3.2.1
Fungal isolation and storage
Isolate RH139 (DAR73145) of R. alismatis was isolated by excising the margin of lesions from
starfruit leaves. Margin tissue was surface sterilised in 2% chlorine solution for 1-2 minutes and
placed on either potato dextrose agar (2% Amyl Media) or lima bean decoction agar. An antibiotic
mixture containing ampicillin, rifampicin and pentachloronitrobenzene was added to both media just
before they were dispensed in plates. Excised leaf pieces were placed on the agar and incubated at
25C in the dark for several days. Fungal tissue was collected from the colony edges by scraping with
a sterile microspatula and was used to inoculate culture media for conidial production. Inoculum was
either freeze-dried or conserved in soil following the method described by Smith and Onions (1994).
3.2.2
Sporulation experiments
3.2.2.1 Effect of media

The extent of sporulation of R. alismatis was determined on a) Difcolima bean agar (DLBA pH 5.6);
b) potato dextrose agar (Amyl Media pH 5.6); c) yeast extract-dextrose agar (YEA pH 5.9); d) malt
extract agar (MEA pH 6); e) V8 juice agar (V8 pH 6.1); f) gamma irradiated sterile carnation leaves
(approximately 20 pieces per petri dish) sprinkled on water agar (CLA pH 6.8); g) blended lima bean
agar (LBA pH 6); h) malt-peptone-dextrose agar (MPDA pH 6.2); i) lima bean decoction agar (LBDA
pH 6) and j) malt and yeast extract agar (MYEA pH 5.8).
Molten media (20 mL) were poured into 200 mL rectangular bottles, autoclaved and allowed to
solidify with the bottles in a horizontal position. Bottles were inoculated with 2 mL of a suspension of
1.5 x 106 conidia mL-1 washed from four-day-old PDA cultures using sterile distilled water. Conidial
16
suspensions were poured into the bottles and then out to inoculate the entire agar surface making a
"lawn" of culture. Four replicates per media were incubated for four days at 25C in the dark, after
which conidia were harvested with 5 mL of distilled water by gently agitating the bottles for 30
seconds. A drop of lactophenol cotton blue was added to the conidial suspensions to kill and preserve
them (Preliminary tests had indicated that the optimum harvest time ranges between 3 and 5 days after
inoculation). The number of conidia per millilitre of suspension was determined with a
haemocytometer.
3.2.2.2 Effect of light cycle
PDA was prepared and inoculated as described earlier with a suspension of 2 x 106 conidia mL-1.
Bottles were exposed to 24 hours continuous fluorescent light, 24 hours dark or 12 hours light/12
hours darkness at 25C. Four replicates per light cycle were used. After four days incubation conidia
were harvested and counted as previously described.
3.2.2.3 Effect of temperature
DLBA was prepared and inoculated as previously described with 1 x 106 conidia mL-1. Four replicates
were incubated at 15C, 20C, 25C, 30C or 35C and after four days incubation in the dark, conidia
were harvested and counted as described above.
3.2.3
Conidial germination

Conidia were produced on each of the ten media used in the sporulation test described earlier and
conidial germination was determined in two separate experiments.
In the first experiment conidia were harvested with 10 mL of sterile distilled water from four-day-old
cultures growing on each of the media. To avoid losses of nutrient due to dilution or washing, the
conidial concentrations were not standardised and ranged from 1x106 to 8x106 conidia mL-1.
Germination medium was 2% water agar (WA) in plastic Petri dishes inoculated with 2 mL conidial
suspension from each medium. Suspensions were spread evenly on the surface of agar and the excess
was poured off.
The second experiment differed in that the conidial suspensions were centrifuged and washed twice
with sterile distilled water to remove any nutrient carried over from the media. Conidial suspensions
were adjusted to approximately 1.5 x 106 conidia mL-1 prior to inoculation of the WA plates.
Each experiment was replicated four times and plates were incubated for six hours at 25C in the dark,
after which lactophenol cotton blue was added to the plates to stain conidia and prevent further
growth. Germination was determined by counting the total number of conidia and the number of
germinated conidia in ten fields of view at 200x magnification. The total number of conidia counted
was at least 200 per replicate and they were considered germinated if the length of germ-tube was
equal to the width of the conidium.
WA plates were inoculated as described in 3.2.3.1 with a suspension containing 1 x 106 conidia mL-1
prepared from four-day-old DLBA cultures. Plates were incubated at 10C, 15C, 20C, 25C, 30C
or 35C. Preliminary tests had shown that germination at temperatures lower than 10C was very slow
(e.g 2% germination at 5C after 12 hours). Four petri dishes from each temperature were removed
from the incubators 2, 4, 6, 8 or 12 hours after inoculation and germination were determined as
described previously. In addition, germ-tube length was measured on 25 randomly selected conidia per
replicate.
17
3.2.3.3 Effect of temperature during conidial production

Conidia were harvested from four-day-old DLBA cultures produced at 20C, 25C or 30C and
suspensions were adjusted to 5 x 105 conidia mL-1. Four inoculated WA plates per treatment were
incubated at 25C for six hours and germination was determined as described earlier.
3.2.3.4 Effect of water depth
This experiment was considered necessary because it was observed that the conidia of R. alismatis
may sink in suspension and in an aquatic environment the sinking of spores may affect their
germination. Glass microscope slides were marked every 10 mm with water-proof markers and stuck,
end to end, onto autoclave tape forming a rigid band. They were wrapped in aluminium foil and
autoclaved. Water agar was poured onto glass slides and later inoculated with 4 mL of a suspension
(2.5 x 106 conidia mL-1) harvested from DLBA cultures. Slides were left in a laminar flow cabinet
until excess water evaporated, allowing conidia to adhere to the agar surface. Autoclave tapes with
inoculated slides were immersed vertically into sterile distilled water in 250 mL measuring cylinders.
The deepest end of the bands were at 10 cm below the surface, the upper parts were above the surface.
Four replicates were used. Cylinders were incubated for 15 hours at room temperature after which
strips of agar were removed every 10 mm, placed on microscope slides for observation and stained
with lactophenol cotton blue to determine conidial germination and germ-tube elongation.
3.2.4
Leaf disc bioassays for lesion development
Water agar containing 1 ppm benzylaminopurine (BAP) was prepared and dispensed in Petri dishes.
Starfruit floating leaves were harvested from glasshouse-grown plants, washed and surface sterilised
with 1% chlorine solution for 1 minute and then rinsed and dried with sterile filter paper. The floating
leaves selected for the bioassays were of the same size and harvested from plants of the same age.
Leaf discs (10 mm diameter) were cut with a cork borer and placed randomly on the surface of the
water agar and pressed lightly to ensure maximum contact (BAP was used to avoid early senescence).
Leaf discs were inoculated with 5 L of a spore suspension (see below) and Petri dishes were placed
in 12 hours light/dark cycle at room temperature. Non-inoculated treatments (water only) were always
used. A band of thin plastic film was used to seal the petri dishes. Percentage of area affected by
necrosis or chlorosis was visually determined daily after three days after inoculation. As an additional
aid for visual evaluation, patterns prepared with millimetric paper were used. These patterns were
prepared by drawing 10 mm circles, simulating the leaf discs on millimetric paper. The respective
areas simulating the diseased areas were calculated and drawn in the middle of each of the 10 mm
circles. The areas were then scanned into a computer and the areas were calculated again using image
processing software (Imagepro). As another additional aid to evaluate the lesions, a scoring system
was used based on visual estimation of the disease severity (Table 3.1).
Table 3.1. Scores of lesion development on starfruit floating leaves.
Score
Symptoms
0
no symptoms;
1
mark left by the inoculation droplet;
2
sparse necrotic flecking at inoculation site;
3
dense necrotic flecking;
4
dense or sparse necrotic flecking with chlorosis;
5
whole inoculation site necrotic;
6
necrosis and/or chlorosis extending beyond the inoculation site;
7
necrosis and chlorosis equal to, or less than, half of the disc;
8
necrosis and chlorosis extending more than half of the disc;
9
necrosis and chlorosis extending to entire leaf disc.
18

All media used in the sporulation experiment were tested together with two additional media, starfruit
leaf extract agar and starfruit leaf agar. Media were inoculated with a conidial suspension and
incubated at 25C in the dark for five days. Six leaf discs were each inoculated with 5 L conidial
suspensions (8 x 10 5 conidia mL -1) prepared from cultures grown on each of the media.
Conidial suspensions (2 x 106 conidia mL-1) were prepared from four-day-old DLBA cultures
incubated at 20C, 25C or 30C in the dark. Seven leaf discs per suspension were prepared,
inoculated and assessed as above.
3.2.5
Data analysis
Data were transformed as appropriate using arcsine transformation for percentage data or square root
transformation for number of conidia. Data analysis was performed using ANOVA and treatment
means were compared by least significant difference (LSD) procedure at 5% significance level. For
the leaf disc bioassay only the percentage area of the disc affected was analysed. Scores of the
diseased area on the discs were averaged and used as an additional aid for evaluation. No statistical
analysis was performed on the scores as qualitative data.
(Conidial production mL -1 x 10 6)
3.5
3
2.5
2
1.5
1
0.5
0
15
20
25
Temperature (C)
30
35
Fig 1. The effect of temperature on conidial production of R. alismatis. Error bars = standard errors.
3.4 Results
3.4.1
Sporulation experiments

R. alismatis sporulated on all media tested. However, the sporulation differed significantly (P<0.05)
between media. LBA and MPDA produced the maximum number of conidia whereas the lowest
number was produced on MEA followed by CLA (Table 3.2).
19
3.4.1.2 Effect of light

There was no significant (P>0.05) effect of the three light regimes on conidial production.
No sporulation was observed at 35C and conidial production at 15C was negligible. Maximum
sporulation was observed at 25C and 30C and the number of conidia was not significantly (P>0.05)
different at these temperatures. Sporulation at 20C was significantly (P<0.05) lower than at 25C and
30C. Figure 3.3 shows the results of the effect of temperature on sporulation.
3.4.2
Conidial germination experiments

There were significant (P<0.05) differences in germination of washed and unwashed conidia produced
on different media. When the conidia were not washed the highest germination rates were observed for
MYA, MPDA and DLBA with no significant (P>0.05) difference between these media (Table 3.2).
Washing the conidia altered the order of germination rates where conidia produced on MPDA had the
least germination and conidia produced on MYA germinated significantly (P<0.05) less than those
produced on DLBA. Germination of conidia produced on DLBA was not significantly (P>0.05)
different from those produced on LBDA, V8, CLA and PDA (Table 3.2).
Table 3.2. Effect of solid culture media on sporulation, conidial germination and infectivity of R.
alismatis.
Media
Sporulation
(x10 6 mL-1)
Difco lima bean agar

Blended lima bean agar
Lima bean decoction agar
Starfruit leaf agar
Starfruit leaf extract agar
Carnation leaves agar
V8 juice agar
Potato dextrose agar
Malt and yeast extract agar
Yeast extract-dextrose agar
Malt-peptone-dextrose agar
Malt extract agar
21.0 bcA
28.6 a
15.4 de
3.7 f
18.8 cd
19.1 cd
13.0 e
16.4 cde
26.5 ab
1.3 g
% conidial germination
Unwashed
94.0 ab
60.3 de
89.8 b
74.3 c
65.2 cde
73.2 cd
98.6 a
38.5 f
96.3 ab
55.5 e
Washed
95.8 a
87.0 bc
95.6 a
95.5 a
94.4 ab
89.7 abc
80.0 c
52.6 d
49.3 d
60.2 d
% leaf disc
area affected
(score)
62 a (7.8)
54 ab (7.7)
53 ab (7.7)
53 ab (7.7)
52 ab (7.7)
48 ab (7)
43 bc (7)
42 bc (7)
42 bc (7)
40 bc (6.6)
40 c (6.6)
38 c (6.6)
Mean values within columns followed by the same letter are not significantly different.
20

Germination began after 2 hours at temperatures between 20C and 35C with higher rates occurring
at 25C and 30C. No significant (P>0.05) difference in germination was observed between 25C and
30C at any time. Germination at these two temperatures was not significantly (P>0.05) different from
35C after 4 hours. The lowest germination after 4, 6, 8 and 12 hours occurred at 10C. Maximum
germ-tube elongation after 2 hours occurred at 25C, 30C and 35C. Germ-tube elongation was
greatest after 6 hours at 25C and 30C. At this time there was no significant (P>0.05) difference
between 20C and 35C.
After the six hours incubation time, there was no significant (P>0.05) difference in germination of
conidia that had been produced at 20C, 25C or 30C.
3.4.2.4 Effect of water depth
There was no significant (P<0.05) difference in percentage germination of conidia above the surface
and at different depths. Length of germ-tube was significantly higher (P <0.001) for conidia
germinating above the surface. The average germ-tube were 66.9 m compared to 14.9m for above
and under the surface respectively). There was no significant difference between germination rates of
conidia at different depths.
3.4.3
Leaf disc bioassay for lesion development

There was a significant (P<0.05) difference after seven days in the size of leaf lesion caused by
conidia produced on different media. The more severe symptoms were observed on leaf discs
inoculated with conidia produced on DLBA, and conidia grown on MEA and MPDA caused the
mildest symptoms (Table 2.2).
There were no significant (P>0.05) differences in lesion severity caused by conidia that produced at
20C, 25C or 30C.
3.5 Discussion
The most important environmental factor affecting germination and infectivity of conidia is the culture
media. Rhynchosporium alismatis sporulates readily in artificial culture. This is an important
characteristic in a mycoherbicide (Trujillo 1992). Jackson et al. (1996), and Shabana, Charudattan and
Elwakil (1995) highlight the importance of the effect of production medium and cultural conditions on
quantity, viability and efficacy of propagules of several fungi investigated as biocontrol agents.
The results in this study indicate that sporulation of R. alismatis is considerably influenced by culture
media and temperature, whereas fluorescent light has no significant effects. In some cases those media
producing large numbers of conidia (e.g. YEA or MPDA) did not produce the optimum conidial
germination and virulence. The results indicated that high rates of sporulation should not be
considered as the main criteria to select any particular media for inoculum production. Considering
virulence as the most important attribute for inoculum production, media based on lima bean, in
particular DLBA, were the most successful. DLBA also produced large numbers of conidia which,
unlike those produced on MPDA, are not adversely affected by losses of nutrient due to the process of
washing or dilution. This may be an important issue for the downstream processing of the
mycoherbicide.
21
Rhynchosporium alismatis sporulates well in liquid and solid media ((Jahromi, Ash & Cother 1998;
Cother & van de Ven 1999). Emphasis has been given in this project to solid media, as it was the most
convenient method for small scale production of conidia to assess the factors influencing the potential
of this pathogen as a biocontrol agent. Liquid media produces both mycelial fragments and conidia,
some of which germinates during the incubation period, and hence the inoculum density for leaf disc
bioassay and glasshouse experiments is more difficult to standardise.
In the current work the pH values of the media were not overly extreme, therefore, it is not possible to
separate the pH effects from the media components. However, Cother and van de Ven (1999) found no
trend between the pH and conidial production.
The results of the effect of temperature on sporulation differed to some extent from those found by
Fox (1995) who, under different conditions and in shake cultures, found that conidial production at
25C was significantly higher than at 30C. It can be concluded, however, that the optimum
sporulation rate occurs between 25C and 30C which are also the optimum temperatures for
germination and growth. Furthermore, the viability and virulence of conidia are not affected by the
temperature at which they are produced.
Water depth does not appear to affect the germination of conidia. However, germ tube length was
shorter under a still-water surface, suggesting that growth of submerged germ-tubes is either slower or
delayed. However, the fungus sporulates abundantly in submerged shake culture (Jahromi et al. 1998),
suggesting, therefore, that slow or delayed growth may be due to lack of oxygenation necessary for
germ-tube growth. This is important since conidia of R. alismatis are large and heavy and it has been
observed that a portion of them can eventually sink. Reduced fungal growth under the surface of still
water calls for measures that keep the conidia on the surface of water. The optimum level of
oxygenation for fungal growth needs further research.
Temperature has profound effects on conidial germination and growth which are slow and delayed
below 20C. High temperature (35C), although inducing earlier germination, inhibits further growth
and sporulation. In fact, germ-tube elongation drops sharply within hours of being exposed to 35C
and no conidia were produced after four days at this temperature. In addition, temperatures equal to or
higher than 30C, appear to reduce lesion development (Jahromi, Ash & Cother 1998).
In summary, conidial production, viability and virulence are affected strongly by the production
media. Light does not appear to influence sporulation whereas temperature affects both conidial
production and fungal growth. The level of oxygenation under submerged conditions can affect fungal
growth.
Therefore, R. alismatis complies with the criterion of abundant production of viable and virulent
inoculum in artificial culture. Future work will focus on epidemiological aspects of the disease in
starfruit.
22
4.
Factors affecting disease development

in starfruit
4.1 Introduction
In the development of a mycoherbicide it is important to study factors affecting infection and disease
development in plants. These would include conidial virulence, spore density and optimum
environmental conditions (Smith 1986; McRae & Auld 1988). Several isolates of R. alismatis
collected from rice growing farms of Murrumbidgee Irrigation and Murray Valley Irrigation Areas are
available and the likely difference between the virulence of these isolates needs to be tested. The
inoculum concentration of other fungal pathogens (potential and actual mycoherbicides) has been
shown to have an effect in the efficacy of the disease in suppressing the weed in question (McRae
1988; Makowski 1993; Klein & Auld 1995). The most important environmental factors affecting the
disease development of many potential mycoherbicides are temperature and dew period (Makowski
1993; Luo & TeBeest 1999).
Some nutrient additives have been reported to increase the severity of disease caused by potential
mycoherbicides. Sugars have been shown to increase infectivity of some fungal pathogens after
addition to the inoculum suspension and applied onto host plants (Boyette et al. 1991). Furthermore,
Shabana, Charudattan and Elwakil (1995) have shown the positive effect of a nutrient-humectant
(Metamucil) on disease development of A. eichhorniae on waterhyacinth. Current research on
biological control of saffron thistle at Charles Sturt University has also concluded that this product
enhances greatly severity of the disease caused by Phomopsis spp after 48 h dew period (Crump, N.
1997 pers. comm.).
Makowski (1993) reported that the growth stages of velvetleaf [Abuliton theophrasti Medik. ABUTH]
and round-leaved mallow [Malva pusilla (Smith)] are differently affected by disease caused by C.
gloeosporioides f. sp. malvae. Starfruit has different growth stages which are likely to be affected
differently by disease.
Furthermore, the water level during inoculation is another factor that may be important. Plants in the
Alismataceae family are submerged in the initial stages of growth as the water level is kept between 3
and 10 cm above the soil (Stant 1963). Aerial parts gradually emerge as plants grow. It is likely
therefore, that water level may influence the efficacy of the inoculation.
The objective of this research was to study the optimum conditions for disease development caused by
inoculation of R. alismatis in starfruit.
4.2 Leaf disc bioassays

4.2.1
Materials and methods
Leaf disc preparation: Leaf disc bioassays were prepared and percentage diseased areas were
determined and scored as in section 3.2.4
Inoculum preparation and inoculation:. The inoculum was obtained by harvesting conidia with 2
mL of sterile distilled water from four-day-old cultures on Difco lima bean agar slopes in McCartney
bottles. The cultures were always incubated at 25C in the dark. Leaves were inoculated by placing
5L of a conidial suspension on the centre of the discs.
Data collection and analysis: Percentage disease area on the discs were transformed if required using
arcsine transformation and analysed using ANOVA and if significant difference was found LSD at 5%
was used to compare different treatments. However, scores of the diseased areas were averaged and
ANOVA was not performed as being qualitative data.
23
4.2.1.1 Isolate virulence

Table 4.1 shows all isolates used in this experiment. The isolates had been collected from diseased
starfruit plants by researchers at New South Wales Agriculture.
Conidial suspensions of each isolate were standardised to approximately 2 x 106 conidia mL-1 and used
to inoculate leaf discs. In all experiments there was a control treatment inoculated with sterile
distilled water. Treatments were replicated five times.
Table 4.1. Origin of isolates of R. alismatis used in this study. DAR is the accession number in the
plant pathology culture collection herbarium of NSW Agriculture, Orange.
Isolate
RH 46
RH 47
RH 100
RH 108
RH 136
RH 137
RH 138
RH 139
RH 143
RH 144
RH 145
RH 146
DAR No
67511
73145
73152
73153
73154
73155
Location
Yanco (NSW)
Yanco (NSW)
Yanco (NSW)
Yallakoo (NSW)
Hanson (Deniliquin, NSW)
Hanson (Victoria)
Hanson (Victoria)
Hanson (Victoria)
Barmah (Victoria)
Barmah (Victoria)
Barmah (Victoria)
Barmah (Victoria)
Date of isolation
Jan 1991
Jan 1991
May 1991
Feb 1993
Jan 1995
Jan 1995
Jan 1995
Jan 1995
Mar 1997
Mar 1997
Mar 1997
Mar 1997
4.2.1.2 Additives
Solutions (5% w/v) of sucrose, maltose, sorbitol and glucose were autoclaved and used separately to
suspend conidia (2.5 x 106 conidia mL-1) of isolate RH139. The suspensions were used to inoculate the
leaf discs, prepared and evaluated as above; in addition two control treatments of conidial suspensions
in distilled water and distilled water only were used for comparison. Each treatment was replicated six
times.
4.2.1.3 Conidial density
Conidia of isolate RH139 were harvested from agar slopes and quantified. Serial dilutions were made
to obtain 9 conidial densities; 104, 105, 5 x 105, 106, 2.5 x 106, 5 x 106, 7.5 x 106, 107, 2.5 x 107 conidia
mL-1. Leaf discs were replicated six times for each treatment including a control inoculated with water
only.
4.2.1.4 Effect of post inoculation temperature
Leaf discs were inoculated with a conidial suspension (5 x 106 conidia mL-1) prepared as described
earlier and used to inoculate leaf discs. Discs were incubated for 24 h at 5C, 10C 15C, 20C, 25C,
30C or 35C in the dark. The plates were placed at room temperature (20-26C) after 24 h and
evaluated as above. The experiment was replicated seven times.
4.2.2
Results
4.2.2.3 Isolate virulence

The severity of lesion development caused by different isolates was significantly different (P<0.05).
Isolate RH145 produced the highest significant level of diseased area on the leaf discs and the highest
score for lesion development after five days. All leaf discs were completely necrotic two weeks after
inoculation with each isolate.
24
4.2.2.4 Additives
No significant difference (P>0.05) was found between the four sugars and the conidial suspension in
distilled water. The average scores of the leaf area affected after 5 days were sucrose (7), maltose
(6.5), sorbitol (7), D-glucose (6.5) and distilled water (6.7).
Lesion development was significantly (P<0.01) affected by spore density after five days. The lowest
lesion development was observed at 104 conidia mL-1 and the lowest average score was also rated at
this density. There was no significant difference for spore densities between 2 x 106 and 2.5 x107
conidia mL-1. All leaf discs inoculated with each spore density were completely diseased two weeks
after inoculation.
4.2.2.6 Effect of post inoculation temperature

Significant (P<0.001) differences were observed four days after inoculation. The greatest lesion
development occurred on leaf discs incubated at 25C. The lowest level of lesion development always
occurred at 35C. Leaf discs were completely necrotic, with the exception of the 35C treatment, two
weeks after inoculation.
4.3 Glasshouse experiments

4.3.1
General
Seed germination: Plants used in these experiments were grown from seeds harvested at the Yanco
Agricultural Institute field sites (Yanco, New South Wales, Australia) during 1995 by NSW
Agriculture. Tests on starfruit seed germinability showed that the seeds do not germinate readily and
need Ethrel (Ethephon 480gr L-1) in concentrations ranging from 50 to 200 ppm to stimulate
germination. The germination tests indicated that within the first ten days a minimum of 40 to a
maximum of 65% (average 55%) of the seeds germinate. Therefore, seeds were always soaked and
germinated in Ethephon solution in Petri dishes before transplanting in the glasshouse.
Planting method: Germinated seeds were transplanted into soil in flat containers placed in plastic
boxes and then covered slowly with water. The aquatic nature of the plant made the logistics of
growing plants somewhat more complicated than with terrestrial plants. Thus, the method of potting
was changed and improved with time. During this research three types of potting were used for all
experiments.
Method 1: Plants were transplanted from flat trays into 10 cm diameter pots which were placed in
plastic boxes filled with water. The main inconvenience of this method was that it needed relatively
large glasshouse space, and handling and transporting the plants was more difficult.
Method 2: Plants were transplanted from the flats to 850 mL (round) transparent plastic food
containers filled with 350 mL of soil. Containers were then filled with water. This method required
less space.
Method 3: The same as method 2 except that the plants were transplanted in 1500 mL square food
containers. The advantage of this method was that more plants could be planted in each container.
The soil used in all experiments was Red Earth and came from the Charles Sturt University farm at
Wagga Wagga campus. Plants were always grown in glasshouse with cooling and heating facility,
with temperature between 17-28C, under natural light conditions.
Growth stages of starfruit: Definition of the growth stages of the host plant is important in the
epidemiological studies. Some authors Cother and Gilbert (1994a); Ditommaso and Watson (1995)
have used days after sowing or days after planting as the criteria for the plant age. These chronological
expressions do not necessarily coincide with a defined phenological age of the plants. In the case of
starfruit it was observed that growth is greatly influenced by the environmental conditions such as
25
temperature and water depth. For instance, in shallow waters or if the young plants are not completely
submerged, they reach maturity earlier than in deeper water. Empirical observations were performed
to define the criteria for the expression of plant age. These observations were carried out as follows:
Starfruit plants were grown using potting method 2 described earlier in this section. Twenty
four pots with one plant per pot were divided in eight plastic boxes with two water levels, 5
and 10 cm above the surface of the pots. Boxes were placed in the glasshouse with
temperature ranging from 17 to 28C. Observations on the plants were noted every day.
In this report I will refer to the initial phases of the growth as juvenile stage. In both water
levels, the juvenile plants were completely submerged. At this age plants start developing
narrow leaves forming a rosette; up to 9 or more of these leaves may grow before the second
growth stage starts. These narrow leaves will be referred to as juvenile leaves. In this latter
stage the plants develop floating leaves which are oval shaped with long petioles. This growth
phase will be referred to as floating leaf stage. The floating leaves are rolled during emergence
and open as they expand on the surface. These leaves initially are 3-5 mm wide and 10-20 mm
long, however, the size of the subsequent leaves increases as the plant matures, in contrast the
length of the petioles decreases with time and floating leaves become gradually erect standing
leaves. Based on visual evaluations no difference between the two water levels were observed.
Inoculum preparation and inoculation: The inoculum was always obtained from four-day-old
cultures of R. alismatis in flat square bottles on DLBA as described in the section 3.2.2.1. Inoculation
was performed using an air-brush at 0.75 atmosphere.
Isolates: During these experiments two isolates RH139 and RH145 were used as specified for each
experiment. Initially RH139 was chosen and later, when more isolates were available, RH145 was
used as it proved to be more virulent in the leaf disc bioassay. The experiments were carried out at the
School of Agriculture (CSU, Wagga Wagga campus) in glasshouse bays with cooling and heating
facilities.
Disease assessment: The effect of the disease on juvenile plants (growth stage 1) and the floating leaf
stage (growth stage 2) of starfruit was determined. The variables measured in all these experiments
were (unless otherwise specified):
a) the total number of leaves;
b) the number of diseased leaves;
c) the disease index, as the ratio between (a) and (b);
d) green leaf area of the floating leaves;
e) dry above-ground biomass (green dry biomass), obtained by drying the samples in a plant
dehydrator at 65C for 48 h;
f) visual estimation of the percentage of diseased area of leaves (Figure 4.5). The areas in Figure 4.5
was measured as follows: a) area of two different size leaves were measured using an Area Meter
(Delta-T Devices Ltd), b) the shape of the leaves were copied on 1:1 scale on overhead
transparencies, c) the total area of the leaves was measured a second time by placing the
transparencies over millimetric paper, d) using the millimetric paper respective percentages of
diseased areas were calculated and drawn. The areas were measured several times to achieve
uniformity; e) the drawn areas were scanned and calculated again on a computer screen as
described in section 3.2.4.
26
Data analysis: Data transformation was performed, if required, using square root or arcsine
transformation for discrete number or proportions respectively. Data analysis was performed using
ANOVA, and if they were significantly different, LSD at 5% significance was used to determine the
difference between individual treatments.
Table 4.2 Different growth stages of starfruit as determined by observations of glasshouse grown
plants.
Juvenile stage
No. of juvenile leaves
1.1
up to 4
1.2
5
1.3
6
1.4
7
1.5
more than 7
Floating leaf stage No. of floating Leaves
2.1
1
2.2
2
2.3
3
2.4
4
2.5
5
2.6
6
2.7
7
2.8
8
2.9
more than 8 floating leaves
Maturity
Floral stem emerging
4.3.2 Effect of plant growth stage at constant water level on susceptibility to

disease
Starfruit plants were grown using method 1 described in section 4.3.1. Four pots each containing a
plant from each growth stage 1.2, 2.2, 2.4 and 2.7 (Table 4.2) were placed in plastic boxes. The water
level in the boxes was kept at 10 cm above the surface of the pots, so that only the blades of the
floating leaves were on the water surface. A conidial suspension (5x10 6 mL-1) was prepared and
applied as described earlier and all boxes were covered with a transparent plastic sheet immediately
after inoculation to simulate a dew period for 24 h. Five replicates set in a split-plot design were used,
with inoculation or no-inoculation (sprayed with distilled water only), as the subplots. Plants were
harvested three weeks after inoculation.
4.3.3
Results
Inoculated leaves developed necrosis and chlorosis. The reduction in green leaf area and above-ground
biomass as the result of fungal inoculation increased with increase in the plant growth stage (Table
4.3). There was no significant difference in the total number of leaves between the control and the
inoculated plants of the same age (P>0.05). ANOVA showed that for inoculated plants there was no
significant difference between the number of dead or diseased leaves and the number of floating
leaves per plant at the time of inoculation. Disease severity index increased significantly as the number
of floating leaves present at inoculation increased (P<0.0001). Disease on the leaves that emerged
27
after inoculation was either not observed or it was observed randomly in small necrotic areas. This
may be due to the non-wetting surface of the floating leaves, so that droplets containing suspended
conidia may have little adhesion to the leaf. Results for green leaf area, dry green biomass and disease
index are shown in Table 4.3.
Table 4.3. Effect of inoculation on different growth stages of starfruit at constant water level. Data
marked with common letters within columns of green leaf area, biomass and disease index are not
significantly different.
Green leaf area
mm2
No. of floating
leaves at
inoculation
0
2
4
7
LSD
4.3.4
Above ground biomass

g
Control
Inoculated
Control
Inoculated
15.65e
48.06c
65.62b
106.30a
10.34e
32.14d
42.48cd
76.04b
0.141e
0.284d
0.387c
0.666a
0.143e
0.211de
0.251d
0.518b
10.8
0.077
Disease index
0.08d
0.35c
0.46b
0.60a
0.1
Factors affecting disease development on the floating leaf stage

Plants were grown to growth stage 2.5 using method 1 described in section 4.3.1 Each pot contained
one plant and two pots were placed in each tub. Plants were spray-inoculated with 106, 2.5 x 106, 5 x
106 or 107 conidia mL-1 of isolate RH139 and the control plants were sprayed with water only. Tubs
were immediately covered with transparent plastic for 24 h to simulate a dew period. Each treatment
was replicated four times in a randomised block design. Plants were harvested after four weeks and, in
addition to the variables mentioned in section 4.3.1, the total biomass (as the sum of green and dead
biomass) was also determined.
4.3.4.2 Effect of dew period
Starfruit plants were grown using method 2 described in section 4.3.1 to growth stage 2.4 before they
were spray-inoculated with a conidial suspension (5x106 conidia mL-1) of RH 139 or with water only.
Containers, each containing two plants, were covered with lids immediately after inoculation and
placed in incubators at 15C or 25C. They were taken out of the incubators every 0, 8, 12, 16, 20 or
24 h and randomly placed in a glasshouse chamber. The experiment was replicated four times and was
harvested after four weeks. The relative humidity in the glasshouse during this experiment was
between 60 and 85%.
4.3.4.3 Effect of a nutrient-humectant additive
Preliminary tests gave some indication that Metamucil may be useful for the increasing the severity
of disease caused by R. alismatis.
Plants were grown using method 1 (section 4.3.1) to growth stage 2.5. Two pots were placed in each
tub. A spore suspension of RH139 (5x106 conidia mL-1) was prepared as described earlier and plants
were sprayed with a) spore suspension containing 2% (w/v) Metamucil, b) spore suspension only, c)
2% Metamucil only or d) water only. A dew period was simulated by covering the tubs as described
above. Five replicates of each treatment were placed in a randomised block design in the glasshouse.
Plants were harvested after four weeks.
4.3.4.4 Effect of water level in disease development
Plants were grown to growth stage 2.7 using method 1 described earlier (section 4.3.1) and inoculated
with 5 x 106 conidia mL-1 (RH139) or with water only. Water level of the tubs during inoculation was
28
kept constant (10 cm above the surface of the pots) or was lowered to the soil level. All plants were
subjected to 24 h dew period simulated as described above, after which the low water level was raised
to 10 cm. Plants were harvested four weeks after inoculation.
4.3.4.5 Effect of isolates
Based on the results of the leaf disc bioassay, six isolates RH46, RH139, RH143, RH143, RH145 and
RH146 were tested in the glasshouse. Plants were grown to growth stage 2.6, using method 1 (section
4.3.1), with two pots per tubs. Each tub was inoculated with a conidial suspension (5 x 106 conidia
mL-1) of one of the isolates or distilled water for the control. A 24 h dew period was simulated by
covering the tubs with clear plastic sheets. Four replicates for each treatment were harvested after 4
weeks.
4.3.5
Results

All conidial densities had similar effects on the plants. There was no significant (P>0.05) difference in
any of the variables between the conidial densities.
4.3.5.2 Dew period
No significant (P>0.05) effects of the dew period length and the dew period temperature were found
on the plants in any of the variables measured.
4.3.5.3 Effect of nutrient-humectant
The addition of Metamucil had no significant effects in disease development at the floating leaf stage
(P>0.05).
4.3.5.4 Effect of water level
All variables measured showed that the water level during inoculation of plants in floating leaf stage
had no significant (P>0.05) effects.
4.3.5.5 Effect of isolate
All isolates significantly (P<0.05) reduced the green leaf area and the above ground biomass in
relation to the control (no inoculation). However, the difference between the isolates was not
significant.
4.3.6 Effect of plant growth stage after lowering the water level to the soil
level on susceptibility to disease
The plants were grown using method 1 described earlier (section 4.3.1) to the growth
stages 1.2 and 2.5. Two pots with a plant from each growing stage were placed in each plastic box.
The experiment was set in a split-plot design with five replicates with inoculation or no-inoculation
(distilled water only) as sub plots. A conidial suspension 5 x 106 mL-1 was prepared and applied as
described earlier. Before inoculation, the water level was lowered to the surface of the pots to expose
the whole plant to the inoculum.
Boxes were covered for 24 h with a plastic sheet, after which it was removed and the water level was
raised to 10 cm above the surface of the pots. Plants were harvested four weeks after inoculation
4.3.7
Results
No necrosis or chlorosis were apparent in the plants inoculated at the juvenile stage. However, the
inoculated juvenile plants were stunted and the total number of floating leaves, above ground biomass
and green leaf area were significantly (P<0.001) reduced (Table 4.4). The ratio of growth suppression,
29
in terms of the biomass and leaf area reduction [(untreated-treated)/untreated] were determined and
they were significantly (P<0.05) higher when the plants were inoculated at the juvenile stage
Table 4.4. Effect of inoculation on juvenile and floating leaf stages of starfruit after lowering the
water level. Data marked with common letters within the columns of green leaf area dry green
biomass and number of leaves are not significantly different.
Growth stages
Juvenile stage
Floating leaf
stage
Green leaf area

mm2
Control
Inoculated
51.82c
17.12d
83.98a
60.80b
Above ground biomass

g
Control
Inoculated
0.323c
0.141d
0.791a
0.509b
19.43
0.116
LSD
4.3.8
No. of Leaves
Control
7.2b
12.8a
Inoculated
2.8c
11a
2.52
Factors affecting disease development at the juvenile stage

It was assumed that the leaf disc bioassays would be more extrapolative to the floating leaf stage than
to the juvenile stage, since they were performed on the discs from floating leaves. Therefore, due to
the differences between the two growth stages, a wider of spore densities were tested on juvenile
plants.
Plants were grown to growth stage 1.3 using method 2 (section 4.5.1). Containers each with two
plants, were drained and sprayed with 104,105, 106, 5 x 106, 107 conidia mL-1 of isolate RH145 or with
water only. Containers were placed immediately in a dew chamber where high humidity was provided
by an ultrasonic mist generator, at 25C in the dark. After 24 h, plants were taken to a glasshouse and
reflooded. Each treatment was replicated six times in a randomised block design and plants were
harvested after four weeks.
4.3.8.2 Effect of a nutrient-humectant additive
Plants were grown as in the above experiment; isolate RH145 was used. Treatments consisted of a) 5 x
106 conidia mL-1 only, b) conidial suspension with 2% (w/v) Metamucil, c) 2% Metamucil or d) water.
Plants were subject to 24 h dew period as above. Five replicates of each treatment were placed in a
randomised block design in the glasshouse and harvested after 4 weeks.
4.3.8.3 Isolates
All isolates used in the leaf disc bioassay were tested on juvenile plants as their reaction may be
different from that of floating leaves.
In this experiment, plants were grown to growth stage 1.3 using method 2 (section 4.3.1) drained and
sprayed with 5 x 106 conidia mL-1 of one of the isolates or water only. Immediately after inoculation, a
24 h dew period was applied in the dew chamber at 25C in the dark, and the plants were reflooded
after placement in the glasshouse. Seven replicates of each treatment were placed in a randomised
block design and harvested four weeks after inoculation.
4.3.8.4 Dew period
Plants were grown using method 3 (section 4.3.1) Each container had five plants, two from stage 1.2,
two from stage 1.3 and one plant from stage 1.4. Plants were sprayed with 5 x 106 conidia mL-1 or with
water only and immediately placed in the dew chamber for 0, 12 or 24 h. The plants not subject to a
dew period (0 h) were placed in the same room but outside the dew chamber to experience the same
temperature. The relative humidity was between (55 and 70%). All containers were reflooded 24 or 48
30
h after inoculation and were placed to the glasshouse after 48 h. Six replicates for each level of
inoculation, the dew period and time to re-flooding were placed in a randomised block design. Plants
were harvested after four weeks and the above ground biomass was measured.
4.3.8.5 Maximum age of juvenile plants
The purpose of this experiment was to study the effect of inoculation at the intermediate stage between
the juvenile and floating leaf stages. The juvenile plants producing their first floating leaf were grown
using method 2 (section 4.3.1) with two plants per container. They had 8-10 juvenile leaves and the
emerging floating leaf was small and under the surface of the water. The containers were drained
before inoculation with 5 x 106 conidia mL-1 (RH145) or with water. A 24 h dew period was applied
before re-flooding the containers. Seven replicates were placed randomly in a glasshouse and were
harvested after 5 weeks.
4.3.8.6 Growth suppression over time
Juvenile starfruit plants were grown, inoculated and a dew period was applied as in the conidial
density experiment (section 4.3.8.1). Six replicates were placed in a randomised block design in the
glasshouse and plants were harvested 21, 35 or 56 days after inoculation. In addition to the variables
measured earlier, the average height of the floral stem for each replicate was determined. The ratio of
growth reduction over time [(untreated-treated)/untreated)] in terms of the reduction ratio in green leaf
area, above ground biomass and the height of the floral stem was determined and compared across the
sample times for each of the variables.
4.3.9
Results

Significant reductions in green leaf area, the total number of leaves (P<0.01) and above ground
biomass (P<0.001) were recorded in plants inoculated with spore densities equal to, or higher than, 105
conidia mL-1. Growth suppression of plants inoculated with conidia at rates between 105 and 107
conidia mL-1 were not significantly different.
4.3.9.2 Effect of a nutrient-humectant

No significant effects (P>0.05) of addition of Metamucil in the growth suppression of the juvenile
plants were found.
4.3.9.3 Isolate
There were significant (P<0.0001) differences in green leaf area, above ground biomass and the total
number of leaves between the isolates tested. The greatest growth suppression occurred on plants
inoculated with isolates RH145 and RH146.
4.3.9.4 Dew period
Different dew periods did not have an effect on biomass as all inoculated plants had significantly
(P<0.01) less biomass than the uninoculated plants. Time to re-flooding had no significant (P>0.05)
effect on plant biomass.
4.3.9.5 Maximum age of juvenile plants

There was a significant (P<0.001) difference in growth between the inoculated and the control plants
for green leaf area, above ground biomass (P<0.0001) and the total number of floating leaves
(P<0.05).
31
4.3.9.6 Growth suppression over time

Fungal inoculation reduced starfruit growth in terms of leaf area and biomass at all sampling times.
Plants still had no floral stem 21 days after inoculation; thereafter the height of the stem was
significantly (P<0.0001) reduced in inoculated plants. The ratio of biomass and leaf area reduction
was not significantly (P>0.05) different between 21, 35 and 56 days after inoculation. However, the
reduction in height was significantly (P<0.05) different between 35 and 56 days. There was a decrease
in the leaf area from 35 to 56 days in the control due to the natural senescence of leaves and dry
biomass.
4.4 Isolation from inoculated juvenile plants

Juvenile plants were grown using method 1 to growth stage 1.4 (section 4.3.1). Ten plants were
inoculated as in the earlier experiments (5 x 106 conidia mL-1) followed by 24 h dew period before
they were reflooded. Re-isolation was attempted at two times. Firstly, before the plants had started to
grow floating leaves a week after inoculation.
Juvenile leaves of five plants were cut and surface sterilised in 1% chlorine solution for one minute
only and were placed on LBDA containing RAP (Appendix 1.) and incubated at 25C. The fungus was
successfully re-isolated after one week. The second isolation attempt was carried out when the plants
had four floating leaves. Only one leaf had visible disease symptoms. Floating leaves of the remaining
plants were cut into four to six square pieces (approximately 5mm x 5mm), surface sterilised, placed
on LBDA and incubated as above. The fungus was not seen growing on agar from the leaf pieces up to
three weeks later; isolation was successful only from pieces with disease symptoms.
4.5 Discussion
Infection by R. alismatis has different effects on juvenile and adult starfruit plants. While in the adult
plants it produces necrosis and chlorosis, in the juveniles it causes a significant reduction in growth by
stunting the plants without causing high levels of necrosis and chlorosis.
The proportional reduction in biomass and green leaf area in juvenile plants, as the result of
inoculation, is higher than in the adult plants. The stunting effect of other fungal pathogens on plants
has been previously reported (Daniel et al. 1973). Water level during inoculation influences the
effectiveness of the fungal inoculation in the juvenile plants. If the juveniles are submerged when the
inoculum is applied to the water surface, it does not affect the plants' growth, because not enough
inoculum reaches the plants. However, the water level at the time of inoculation does not appear to
affect disease development in adult plants. The inoculation of juvenile plants may be more desirable
from an agronomic point of view, since the growth suppression of the weed early in season may give
rice plants a competitive edge over starfruit.
Inoculation at the later growth stage of juvenile plants when the first floating leaf is emerging, can still
significantly stunt the plants. At this age, plants are still completely submerged and need to be exposed
before inoculation. Therefore, it can be assumed that as the plants mature the inoculation will produce
more necrosis and chlorosis, and less growth suppression.
Another Rhynchosporium species, R. secalis (Oud.) Davis, causes scald disease in barley producing
similar symptoms of necrosis and chlorosis. Ayesu-Offei and Clare (1971) found that when barley
seedlings were submerged in sterile culture filtrates of R. secalis, they developed similar symptoms to
leaves infected by the fungus. Mazars et al. (1989) showed that phytotoxic compounds are produced
by R. secalis. One of them, a high molecular weight compound, a glycoprotein, is capable of
producing necrotic and chlorotic symptoms found on barley leaf scald. Therefore, there is reason to
believe that R. alismatis may also produce a toxin that causes both stunting of juvenile plants, and
necrosis and chlorosis of adults.
The fungus was isolated successfully only from submerged leaves after inoculation but not from
floating leaves emerging after inoculation (only one floating leaf had the disease symptoms),
suggesting that the fungus is not systemic. Fox (1995) found that R. alismatis may be seed-borne at
32
low levels. However, seed borne inoculum may arise from infection of the inflorescences rather than
movement of the fungus through the plant. Infection of floral stems and fruits of starfruit can be
commonly observed in the field.
There are some differences in the speed of disease development between the isolates tested, but all of
them produced necrosis and chlorosis and eventually killed the floating leaves that came in contact
with the inoculum. This fact and the lesser possibility of disease development in floating leaves
emerging after inoculation, under glasshouse conditions, explains the lack of significant differences
between the isolates tested on adult plants after four weeks. In juvenile plants, however, the difference
between the isolates is more important. These results agree with the leaf disc bioassay indicating that
isolate RH47 is the least pathogenic, whereas isolate RH145 is the most pathogenic on floating leaves
causing more significant stunting on juvenile plants. It is, therefore, recommended for future
experiments.
The conidial density necessary to cause disease in adult and juvenile plants does not constitute a
limiting factor since high levels of sporulation in culture were achieved Jahromi, Ash and Cother
(1998). The stunting effect on juvenile plants can be achieved by applying 105 conidia mL-1.
Nevertheless, the experimental trends suggest that densities of at least 106 and up to 5x106conidia mL-1
may be recommended for future glasshouse and field trials.
The leaf disc bioassay, simulating the effect of dew period temperature on infection, showed that
temperature has a significant effect on development of disease symptoms. Taking into account the
effects of temperature on fungal growth and sporulation, it can be concluded that temperatures below
25C delay lesion development while 25C enhances it. As the temperature increases to 30C, the
severity of the symptoms is reduced, dropping sharply at 35C. Cother and van de Ven (1999) also
indicated that lesion development in leaf disc bioassays is slower at 30C than at 25C if temperatures
persist. It appears that 30C promotes only vegetative growth and sporulation in artificial culture while
infectivity of conidia is reduced.
Glasshouse dew period experiments indicated that under those experimental conditions there was no
discernible effect of additional moisture in disease development. This is because the relative humidity
above the water surface, in the case of adult plants, and the humidity provided by the saturated soil in
the case of the juvenile plants keep moisture on the leaf surface long enough for the conidia to
germinate and penetrate the plants. Daniel et al. (1973) reported that anthracnose disease caused by C.
gloesporioides f. sp. aeschynomene developed in northern jointvetch, without additional moisture in
the glasshouse. In rice fields the fungus controlled the weed in rice when sprayed at dusk. TeBeest,
Templeton and Smith (1978) indicated that moisture in the rice fields is enough for the anthracnose
disease to develop in the weed.
In summary, R. alismatis does not kill starfruit. The disease in adult starfruit plants causes necrosis
and chlorosis on aerial parts. In the juvenile plants it reduces significantly plant growth and this may
assist rice to out-compete the weed. Growth suppression can be achieved with spore densities as low
as 105 mL-1 applied in spray suspension. To produce the stunting effect on juvenile plants it is
important to lower the water level to expose the plants to fungal inoculum during spray inoculation.
There are some small, but significant differences in the virulence of the isolates that are expressed
especially in the stunting effect of juvenile plants. If high temperatures (>30C) persist, disease may
be inhibited. Therefore, and in future field trials, inoculation should be done at times when ambient
temperatures are lower. The most important advantage of this weed situation is the lack of requirement
for a dew period. It appears, therefore, that no formulation to trap moisture for conidial germination
and infection is necessary. However, this needs to be verified in future field experiments where
climatic conditions are more variable than in the glasshouse.
33
5.
Microscopy
5.1 Introduction
The initial stages of the infection process of starfruit leaves by R. alismatis have not been previously
studied. Moore-Landecker (1990) states that the next step after the contact of fungal inoculum and the
host is indirect or direct penetration. In the former, a fungus enters the plant by means of wounds or
natural openings. Wounds may be caused by insects or humans although stomata are considered to be
the most important natural opening for fungal entrance. Direct penetration through an intact epidermis
occurs usually after the formation of an appressorium. Some fungi can use one or more of these means
of entrance (Moon-Landecker 1990). Howard (1997) states that fungi that employ either direct or
indirect penetration means are common. Fungi using different means of penetration may do so because
of the environmental effect.
Staples and Hoch (1997) differentiate other infection structures according to whether they are formed
on the tip of germ tubes (appressoria) or arise from mycelia (hyphopodia or infection cushions). Yet
the infection structures arising from hyphae or germ tubes may appear identical; therefore they should
be differentiated by their function Howard (1997). Similarly, Emmett and Parbery (1975)
distinguished these structures by their function rather than by whether they are formed on germ-tubes
or mycelia. All structures that adhere and cause penetration into the host are grouped as infection
structures that are considered synonymous with appressoria (Emmett & Parbery 1975); the term used
in this report.
Rhynchosporium alismatis does not appear to require wounds to penetrate starfruit tissue since no
wounds were needed for infection in the previous studies (Chapters 3 and 4). However, the process of
infection is temperature sensitive with high temperatures (above 25C) slowing down the process.
Some authors Roderick and Thomas (1997) attribute the existence of differential growth and germtube orientation in some fungi to a response to different textures of leaf surfaces. Furthermore, in vitro
studies of the infection structures are important as an experimental approach to understand the hostpathogen interaction (Howard 1997).
The objectives of this study were to (i) identify pattern of conidial germination, (ii) define the time
sequence and means of fungal penetration of the fungus and how this process is affected by
temperature and topography of the leaf surface, and (iii) verify if conidia are formed on the leaves
after infection. This knowledge is fundamental in understanding of the epidemiology of the disease
and to predict the behaviour of the pathogen in the field.

5.2.1 Effect of temperature on growth and appressorium formation on
cellophane paper
Cellophane paper was cut in 2 cm2 pieces, boiled and autoclaved in beakers containing distilled water.
When cool, they were placed on the surface of water agar in 9 cm diameter petri dishes. Five pieces
were placed in each dish.
A conidial suspension (4 x 105 conidia mL-1) was prepared from four-day-old cultures of isolate
RH145 on slopes of Difco lima bean agar in McCartney bottles. The cellophane pieces in 32 Petri
dishes (with 5 pieces in each dish) were inoculated with 50 L of conidial suspension and 16 plates
were placed in incubators set at 25C or 30C. Agar plates and the sterile water used to dilute the stock
conidial suspension had been placed in the incubators the night before. Two plates were removed from
each incubator at 2, 4, 6, 8, 10, 12, 16 and 24 h after inoculation, from which five pieces per
temperature were removed and placed on a glass slide. A drop of lactophenol cotton blue was added
and a cover slip was placed on top of the paper and sealed. Eight fields of view were observed under a
light microscope at x200 magnification for each piece of cellophane paper. The total number of
conidia, number of germinated conidia, number of conidia geminating from two conidial cells, number
34
and length of the germ-tubes and the number of appressoria were recorded. The conidia were
considered germinated if the length of the germ tube was equal to the width of the conidium.
5.2.2 Conidial germination, growth and appressorium formation on starfruit

floating leaves
Leaf-discs were prepared in Petri dishes as previously described (Section 2.2.4) and 10 discs were
place in each Petri dish. Conidia from four-day-old cultures of isolate RH145 grown on DLBA were
washed as described earlier. A conidial suspension (4 x 105 conidia mL-1) was sprayed onto the leaf
discs using an air-brush and placed in an incubator at 25C. Plates was taken out at 2, 4, 6, 8, 10, 12,
14, 16, 24 and 48 h intervals. Six leaf discs were removed, fixed in 1:2 solution of acetic acid:ethanol
until cleared and rinsed for a minimum of 30 min. in distilled water. Lactophenol cotton blue was then
added onto the discs, left for one minute and rinsed for a few seconds in distilled water before the
discs were mounted on glass microscope slides with glycerol. A number of staining methods had been
tested previously including acid lacto-fuchsin (Dhingara & Sanders 1985), and fluorescent staining
(Rohringer, Kim & Samborski 1979). In addition to the counts listed earlier, the number of stomatal
penetrations was recorded.
5.2.3 Effect of temperature on growth and appressorium formation on floating

leaves
Petri dishes each containing four leaf discs were prepared and inoculated as above and incubated at
25C or 30C. Two plates per temperature were removed at 2, 4, 8, 12, 16, 24 and 48 h after
inoculation. Five leaf discs per temperature were removed cleared and fixed and mounted on glass
slides for observation as described previously (section 5.2.2).
5.2.4
Conidial germination and appressorium formation on juvenile leaves
Juvenile leaves were cut from plants and surface sterilised for 45 seconds in 1% chlorine solution and
were placed on the surface of BAP agar, spray inoculated with 4 x 105 conidial mL-1 and incubated at
25C in the dark. Juvenile leaves were observed as described above at 2, 4, 8, 12 and 24 h after
inoculation.
5.2.5
Scanning electron microscopy
The leaf discs and the inoculum were prepared as described above. They were inoculated in the centre
with 5 L of a conidial suspension under aseptic conditions. This method was prepared to avoid any
possible contamination visible through SEM. Leaves were inoculated 14, 24 and 48h before
observation. Water agar and LBA plates were also inoculated for observation for comparison. Samples
were prepared and viewed as described by Craig and Beaton (1996) except that samples were frozen in
liquid nitrogen before insertion in the cold stage. Samples were viewed at 15 kV in a JOEL 6400
SEM via a Bio-Rad E7400 cryotrans system.
35
5.3 Data analysis

The data were transformed where appropriate using arcsine or square root transformation for
percentages and discrete numbers respectively and analysed using ANOVA.
5.4 Results
5.4.1
Effect of temperature on growth and appressorium formation on

cellophane paper
Germination approached 90% at the two temperatures after 8 h. Observations were recorded up to 16 h
after inoculation. After this time abundant germ tube growth was observed at all temperatures. The
conidia (12-17 m) were usually septate although the septum was not always clearly visible. No
significant differences between temperatures in germination and germ tube length were observed at
any time. Approximately 30% had germinated after 2 h and only 2% of the conidia had more than one
germ tube. By 10 h after inoculation more than 80% had more than one germ tube. Appressorium
formation started 6 h after inoculation at both temperatures and was significantly (P<0.05) higher at
25C after 10 h. Appressoria were either round or oval shaped (2.8 to 4 m). The infection structures
had formed either at the end of short (1-3 times the length of the conidium) or on long and branched
germ-tubes. No significant difference in germ tube length was observed at any time.
After 12 h, 1.4% and 4% of conidia at 25C and 30C respectively had produced swellings at the end
of germ tubes that were too large to be considered appressoria and resembled conidia. These structures
resembled conidial formation at the end of the germ-tube
5.4.2
Conidial germination and appressorium formation on floating leaves
Similar rates and patterns of conidial germination and germ tube length were observed on the leaf
tissue. Conidial germination was observed on different parts of the leaves and the germlings grew in
an apparently random way in all directions. Forty percent of the conidia had germinated after 2 h and
over 90% had germinated after 8 h.
The appressoria were observed on most leaf surfaces. Three types of appressoria were observed: (i)
appressoria forming at the end of long or branched germ-tubes, (ii) appressoria forming at the end of a
short germ tube, (iii) sessile appressoria.
Very few penetrations through the stomata were observed, less than 10% of the germ-tubes had
penetrated through stomata between 16 and 24 h after inoculation. Although the stomatal penetrations
appeared to increase as the germ-tube length and the hyphal ramification increased, this increase
lacked a clear pattern and fluctuated with time. In addition, germ tubes often were observed passing
over stomata without penetrating. Sixteen hours after inoculation it was possible to observe the
subcuticular hyphae erupting out of the leaf tissue through the epidermis. Aerial conidial formation
was observed after 48 h on mycelium erupting through the epidermis or through stomata. No
subcuticular conidial formation was observed. No non-appressorial swelling at the end of germ-tube,
as described in section 5.4.1, was observed.
5.4.3
Effect of temperature on growth and appressorium formation on

floating leaves
There was no significant difference in germination, germ tube length and penetration through stomata
at 25C or 30C on starfruit leaves. The number of conidia producing appressorium was significantly
higher at 25C than 30C after 8 h after the inoculation. No direct eruption of subcuticular hyphae nor
hyphae coming out of stomata were observed at 30C.
36
5.4.4
Conidial germination and appressorium formation on juvenile leaves
The rates and trends of germination, growth and appressorium formation were similar to the floating
leaves. No stomatal penetrations were detected in these observations.
The juvenile starfruit leaves are submerged in water and have far fewer stomata than floating leaves.
Stomata are almost nonexistent on the bottom half of juvenile leaves but become more numerous on
the upper part towards the apex of the leaves. Appressorium formation rates were 0, 13, 38, 42 and
51% at 2, 4, 8, 12 and 24 h after inoculation respectively.
5.4.5
Scanning electron microscopy
The SEM revealed a lack of ornamentation on the conidial surface. Germinated conidia on water agar
that were thought to be stressed through drying, had formed swellings at the end of germ-tubes as
described in section 5.4.1. On the leaf tissues there was evidence of physical pressure around the
appressorium due to the action of penetration. Signs of possible adhesive materials similar to that
described by Yates and Cason (1996) were visible around appressoria. Most of the germinated conidia
which had formed appressoria or had penetrated through stomata had began to collapse as a result of
the transfer of nutrients to the infection sites. It also revealed that, although rare, some penetration
through intact epidermis may occur without the formation of appressoria. Conidial formation on the
leaves was observed 48 h after inoculation.
5.5 Discussion
This study reveals that germ-tubes formed appressoria, which in turn produced infection hyphae in the
sub-epidermal cells within the first hours after inoculation. Appressorium formation on the juvenile
starfruit plants is important since inoculation of the juvenile plants is preferred to cause disease early
in the plants' growth. The observations suggest that penetration through the stomata appears to be a
random event and does not appear to be caused by any stimulus. Hyphae and germ tubes frequently
passed over the stomatal cavity but did not penetrate.
It can be concluded that the appressoria are the main means of penetration by R. alismatis and any
artificial measure or stimulus that encourages appressorium formation would be useful.
There are some similarities between the process of penetration by R. alismatis and that described by
Ayesu-Offei and Clare (1970) for the infection of barley by R. secalis (i. e. both produce sessile
appressoria). However this potential mycoherbicide is faster in its development. Appressorium
formation starts just 4 h after inoculation, subcuticular hyphae is visible 16 h after inoculation, lesions
on leaf tissue are visible just three days after inoculation (Chapter 2) and conidial formation on leaves
is visible within the first 48 h after inoculation. In contrast R secalis produces visible subcuticular
hyphae in barley leaves after 4 days and conidial formation after 10 days.
The deformation of the leaf as a result of physical pressure and the collapse of the germinated
conidium is due to the transfer of the cell contents through to the appressorium and the infection site (i
e. the penetration peg) to the infection hyphae (Roderick & Thomas 1997). It can be assumed,
therefore, that direct penetration starts soon after appressorium formation; however, further study to
observe the penetration peg may be may be useful.
The lack of subcuticular conidial formation and the existence of solitary and erumpent conidiogenous
cells differ from that described by Ayesu-Offei and Clare (1970) for R. secalis but it matches the
description by Braun (1993) who classified R. alismatis as Spermosporina alismatis (U. Braun).
Previous work (Chapter 4) simulating the effect of dew period experiment on subsequent lesion
development showed that high temperature (30C) reduced the speed of disease development. This can
be related to appressorium formation being significantly higher at 25C than at 30C. Staples and
Hoch (1997) described several factors that affect appressorium formation such as the physical and the
chemical cues but they did not specify any temperature effect other than heat shock. However,
Roderick and Tomas (1997) reported significant effects of temperature during the first hours after
inoculation on appressorium formation by three rust species on ryegrass.
37
Appressorium formation on cellophane paper started after 6 h, and on leaf discs 4 h after inoculation.
In addition, the rates on paper were lower than on discs. Although the results of the two experiments
(on the cellophane paper and the leaf discs) may not be directly compared (they were performed at
different times), different patterns and rates of appressorium formation suggest their formation could
be a fungal response to the leaf surface. Fungal responses may be due to thigmotropism, chemical
stimulus, surface hardness or any combination of these factors.
According to Staples and Hoch (1997), thigmotropic responses occur when the fungus detects
different features on the host surface such as the topography or surface thickness; such responses have
been reported in several rust fungi. Thigmotropism may be expressed in directional changes or cell
differentiation. In the case of R. alismatis there is no indication of any directional growth of the germ
tube.
Chemical cues in the form of exudate or surface wax from starfruit leaves may also be a cause for
appressorium induction. Chemical exudates from pepper fruit (Capsicum frutescens L.) stimulated
germ tube differentiation of appressoria in Colletotrichum piperatum conidia (Staples & Hoch 1997).
Podila et al. (1993) found that surface wax of avocado fruit (Persea americana Miller.) stimulates
appressorium formation of C. gloeosporioides.
In addition, we have never found appressoria forming from germinated conidia of R. alismatis on
liquid or agar media suggesting that surface hardness is another factor necessary for appressorium
formation. The surface hardness is considered one of the most important factors for appressorium
formation (Staples and Hoch 1997). In addition, Staples and Hoch (1997) stated that the appressorium
formation is regulated by different mechanisms where there is a requirement for "well-defined"
environmental conditions for most species. They highlighted that these environmental conditions are
required for the expression of the genotype controlling the appressorium formation. The fact that not
all conidia produced appressoria can be explained by the likelihood that different conidia in a
population may require different conditions for appressorium formation (Emmett & Parbery 1975).
Conidial germination does not appear to be regulated by topography nor substrate composition as they
germinate readily on water agar. However, germination and germ tube growth is influenced by
temperature (Chapter 3). Yates and Cason (1996) reported responses very similar to R. alismatis for
conidial germination and appressorium formation of the fungal pathogen Cladosporium caryigenum
(Ellis & Langl.) Gottwald affecting pecan [Carya illinoinensis (Wangenh.) K, Koch].
The lack of any type of attachment structures such as ornamentation on the conidial surface means that
the adhesive ability of the conidia is low. This could be a reason for the low levels of infection on the
floating leaves that emerge to the surface after inoculation (Chapter 4), since the conidia do not adhere
to the newly emerged leaves. Future formulation work may include the study of ways of making the
conidia stick to the emerging leaf surface.
In summary, conidia of R. alismatis germinate, produce infection structures and infection hyphae on
starfruit leaves within the first 16 h after inoculation. Appressorium formation is the main means of
penetration, especially on the juvenile leaves, and is temperature dependent. The optimum temperature
for appressorium formation (25C) coincides with the optimum dew period temperature previously
reported (Chapter 2), and is likely to be influenced by conditions on the starfruit leaf surface.
38
6. Fungal interaction with chemical

herbicides
6.1 Introduction
In the development of a mycoherbicide it is important to study its integration with chemical herbicides
(Templeton 1982). It is accepted that chemical herbicides do have effects on disease development in
plants (Altman & Campbell 1977). Levesque and Rahe (1992) reported increased disease severity as
the result of interaction between trifluralin and Fusarium oxysporum Schlecht. and fenoxapop and
Pythium arrhenomanes Dreschsler, in soybean and rice respectively. However, herbicides may also
decrease disease severity in plants (Prasad 1994).
The interaction between chemical herbicides and bioherbicides can be antagonistic, synergistic or
additive. Anatagonism occurs when applied together, they have no or less effect than when applied
separately; in contrast when there are synergistic and additive effects, their combined effect is larger
than when applied separately. These two interactions are highly desirable since they can increase
disease severity caused by a mycoherbicide (Prasad 1994). In particular, the use of synergistic
interaction between sublethal rates of herbicides and mycoherbicides has gained interest as part of an
integrated weed management system (Lovett & Knights 1996). It is thought that plant defence
mechanisms can be diminished by herbicides, weakening the plant and increasing the efficacy of the
pathogen in infecting the host (Levesque & Rahe 1992). Charudattan (1986) reported that treatments
of waterhyacinth with Cercospora rodmanii Conway, and sublethal rates of 2,4-D resulted in
improved weed control compared to using C. rodmanii or the herbicide alone. Wymore, Watson and
Gotlieb (1987) showed synergistic interaction between Colletotrichum coccodes (Wallr.) Hughes, and
low rates of thidiazuron for controlling velvetleaf.
In looking at the interaction between chemical herbicides and mycoherbicides it is necessary to
determine any adverse effect the chemical may have on the fungus in order to determine whether or
not they can be mixed. These studies are normally performed in vitro (Charudattan 1986; Prasad
1994).
The objectives of this study are to determine the effect of the most commonly used herbicides to
control starfruit on conidial germination of R. alismatis, and to find possible synergistic or additive
effects between the most compatible herbicides and the fungus to increase disease severity in starfruit.
6.2 Effect of herbicides on conidial germination

6.2.1
Materials and methods
The herbicides used to study the effects on conidial germination of R. alismatis are shown in Table
6.1. Londax and MCPA 250 are the most common herbicides used to control starfruit; Ronacil is
used in lesser extent. Although Round-Up is not specifically recommended for this weed it was
selected for this experiment as it is the world's most important herbicide (Franz, Mao & Sikorski
1997). This herbicide is also used in rice for the control of other weeds.
39
Table 6.1. Selected herbicides for the in vitro studies of their effect on conidial germination of R.
alismatis.
Active Constituent
Herbicide
Herbicide dose ha-1
85 g
Bensulfuron-methyl 600 g Kg-1
Londax
MCPA 250
1.5 - 2.8L*
MCPA sodium salt 250 g L-1
11 L
Propanil 360 g L-1
Ronacil
1L
Glyphosate 450 g L-1
Round-Up 450
Source: NSW Agriculture and the RIRDC Rice Research & Development Committee (1996), Rice
Notes, a supplement to Rice Growing in New South Wales.
*Different rates of MCPA 250 are recommended according the growth stages of weed and crop.
Serial dilutions of a stock solution of each herbicide, equivalent to application of 200 L of water per
hectare, were made to obtain 1, 0.5, 0.25, 0.125, 0.0625, 0.03125, 0.015625 times of the recommended
dose per hectare. MCPA 250 was tested at rates equivalent to 1.5 and 2.8 L ha-1.
Conidia of the isolate RH145 were harvested with 10 mL of sterile distilled water from four-day-old
cultures on Difco lima bean agar in flat media bottles as described in section 3.2.2.1. A conidial
suspension (2 x 107 conidia mL-1) was centrifuged for 5 min. at 2500 rpm and excess water was poured
off. Conidia were resuspended in 3 mL sterile distilled water to obtain a stock suspension (2 x 108
conidia mL-1), 5 L of this suspension was added to each millilitre of the herbicide dilutions, or water
only, to obtain suspensions of 1 x 106 conidia mL-1. Herbicide dilutions with conidia were left for 30
minutes before they were used to inoculate water agar in 9 cm petri dishes. Each dish was inoculated
with 1mL of the suspension. Three dishes for each treatment were incubated for 12 h at 25C in the
dark, after which the percentage germination of conidia was determined. Conidia were considered
germinated if the length of the germ tube was at least equal to the width of the conidium.
Data across all herbicide dilutions were analysed using ANOVA followed by LSD at 5% significance.
6.2.2
Results
MCPA 250 had no effect on conidial germination at any dosage. The herbicide with maximum
deleterious effect was Ronacil, no conidial germination was observed at any concentration. The effect
of Londax and Round-Up on conidial germination was reduced as the herbicide concentration
decreased, although Round-Up had a larger inhibitory effect on germination.
6.3 Glasshouse experiments

6.3.1
Fungal interaction with MCPA
Preliminary tests on adult plants: To determine the effects of sublethal rates of MCPA 250 on
starfruit plants, preliminary tests were carried out. A pot sprayer was used to apply the equivalent of
100 L of water per hectare. Several dilutions based on the recommended product dose of 2.8 L ha-1
were made and herbicide effects on plants were visually estimated. The water level was not lowered
before the spray, as this is the case when adult plants are inoculated with conidial suspensions
(Chapter 4).
It was found that only the full recommended dose eventually killed the plants. Under glasshouse
conditions, doses below 0.2 of the recommended rate did not have a visual effect on the plants.
Preliminary tests on juvenile plants: In the case of the juvenile plants it was necessary to drain the
containers before herbicide application. Herbicide rates which produced no visual effects on the plants
were lower than required for adult plants. It was found that these need to be equal to or less than
0.0625 times of the recommended dose of 1.5 L ha-1. Dead plants were infrequently observed at these
low rates; such phenomena have also been reported before in other herbicides (Wymore, Watson &
Gotlieb 1987).
40
6.3.1.1 Fungal interaction with MCPA at floating leaf stage

Starfruit plants were grown to growth stage 2.7 (section 4.3.1) in 10 cm diameter pots placed in 45 x
30 x 23 cm white plastic boxes (method 1 described in section 4.3.1). A conidial suspension (5 x 106
conidia mL-1) of R. alismatis (RH 145) was harvested from four day-old-cultures on Difco lima bean
agar as described earlier. Herbicide rates were 0.2 and 0.15 of the recommended product dose (2.8 L
ha-1), referred to as M1 and M2, respectively.
For ease, the conidial suspension and the herbicide solution were sprayed separately with an airbrush
and an average of 1.5 mL of each spray was applied per box. Each replicate (box) contained two pots
with one plant each. Treatments consisted of:
a)
b)
c)
d)
e)
f)
g)
h)
i)
j)
control sprayed with water only;

fungal inoculation (F),
fungal inoculation and M1 applied at the same time (FM1);
fungal inoculation and M2 at the same time (FM2);
fungal inoculation followed by M1 after four days (F/M1);
fungal inoculation followed by M2 after four days (F/M2);
M1 followed by fungal inoculation after four days (M1/F);
M2 followed by inoculation after four days (M2/F);
M1 only;
M2 only.
All boxes were covered with clear plastic for 24 h to simulate a dew period. Boxes containing the
plants were placed in a randomised block design with five replicates in the glasshouse and plants were
harvested after five weeks. All variables described in section 4.3.1 were measured.
6.3.1.2 Fungal interaction with MCPA at the juvenile stage
Juvenile plants were grown to growth stage 1.3 in 850 mL food containers using method 2 of potting
(section 4.3.1) with two plants per container. Herbicide rates were 0.0625 (M1), 0.03125 (M2) and
0.015625 (M3) of the recommended dose (1.5 L ha-1). Herbicide was applied using a pot sprayer as in
the preliminary tests. Fungal inoculum (5 x 106 conidia mL-1) was prepared on Difco lima bean agar
as in the previous section and applied with an air brush. Juvenile plants were drained before a spray
application. According to the product label; it is recommended that plants should not be reflooded
until 24 to 48 h after the herbicide application. The treatments consisted of :
a) Control treatment sprayed with water only followed by 24 h dew period and reflooded 24 h later;
b) fungal inoculation only (F);
c) M1 only;
d) M2 only;
e) M3 only;
f) fungal inoculation and M1 at the same time followed by 24 h dew period and reflooded 24 h later
(FM1);
g) fungal inoculation and M2 at the same time followed by 24 h dew period and reflooded 24 h later
(FM2);
h) fungal inoculation and M3 at the same time followed by 24 h dew period and reflooded 24 h later
(FM3);
i) M1 before fungal inoculation after 24 h, followed by a 24 h dew period and reflooded (M1/F);
j) M2 before fungal inoculation after 24 h, followed by a 24 h dew period and reflooded (M2/F);
k) M3 before fungal inoculation after 24 h, followed by a 24 h dew period and reflooded (M3/F);
l) fungal inoculation then 24 h dew before M1, and reflooded 24 h later (F/M1);
m) fungal inoculation then 24 h dew before M2, and reflooded 24 h later (F/M2);
n) fungal inoculation then 24 h dew before M3, and reflooded 24 h later (F/M3).
Six replicates were placed in a randomised block design and plants were harvested five weeks later.
Aerial biomass was measured by drying the samples for 48 h at 65C.
41
6.3.2
Fungal interaction with Londax
The manufacturer recommends that Londax should be used for treatment of juvenile starfruit plants
and it should be applied in still water. Applications need not be precisely uniform. In addition, after its
application no draining or new water should be added to the treated area for at least five days after
application. Therefore, the fungus/herbicide interaction was studied only at the juvenile stage.
Preliminary tests were performed applying different rates of the herbicide by injecting the solution
under the water surface. It was observed that sublethal rates of Londax (i.e 0.03125 of the
recommended dose) had some effect on starfruit, causing some deformation and/or stunting of the
plants. However, it was observed that most plants would eventually recover after approximately 8
weeks
Fungal inoculum and starfruit plants were prepared as in section 6.3.1.2. Herbicide was injected
directly underneath the water surface with a pipette (at a rate of water equivalent to 100 L ha-1) and a
conidial suspension (5 x 106 conidia mL-1) was applied with an airbrush. A 24 h dew period, in a dew
chamber at 25C in the dark, followed all fungal inoculations.
Fungal interaction was determined with two herbicide rates 0.03125 (L1) and 0.015625 (L2).
Treatments consisted of:
a) control without drainage;
b) fungal inoculum only (F);
c) L1 only;
d) L2 only;
e) L1 followed by fungal inoculation after six days (L1/F);
f) L2 followed by fungal inoculation after six days (L2/F);
g) fungal inoculation followed by L1 after six days (F/L1);
h) fungal inoculation followed by L2 after six days (F/L2) and
i) control was drained, sprayed with water and followed by a 24 h dew period.
Seven replicates of each treatment were placed in a randomised block design and were harvested after
six weeks. Green leaf area and aerial biomass was determined as in section 4.3.1.
Data analysis Data for all glasshouse experiments were transformed using square root transformation
and analysed using ANOVA; if significant, LSD was determined at 5% significance.
6.4 Results
6.4.1
Fungal Interaction with MCPA
6.4.1.1 Fungal interaction with MCPA at the floating leaf stage

There was a significant difference (P<0.001) between treatments. There was no significant difference
between the control and treatments where herbicide alone was applied. Biomass and leaf area were
reduced for fungal inoculation only and there was no significant difference between all fungusherbicide combinations.
6.4.1.2 Fungal interaction with MCPA at juvenile stage
3.2.2.2 Effect of light cycle
There was a significant difference between treatments (P<0.001). The dry biomass of the control
plants was not different from the M2 and M3. All herbicide-fungus interactions reduced the biomass
relative to the control, the fungus and M1 combinations appear to reduce the biomass more than other
combinations as shown in Figure 6.2.
6.4.2
Fungal interaction with Londax
All treatments reduced significantly (P<0.001) the biomass and leaf area relative to the control. The
resulting growth suppression when the herbicide was applied prior to the fungal application was, in
most cases, higher compared to fungal inoculation or herbicide only, and in comparison to when the
fungus was applied prior to the herbicide.
42
6.5 Discussion
Templeton (1982) emphasised the importance of not losing the biological activity of mycoherbicides
in a tank mix situation. The in vitro study of the herbicides' effect on conidial germination was an
attempt to simulate the tank mix of R. alismatis with the herbicides commonly used in rice. The two
most important herbicides used for the control of starfruit (Londax and MCPA 250) had the lowest
impact on conidial germination. The effect of the herbicides on conidial germination may be due to the
adjuvants or surfactants used in the products' formulation rather than the active constituent. This
possibility needs to be determined in future experiments.
Charudattan (1986) found that when C. rodmanii was applied a week before 2,4-D, the disease
severity was the greatest. However, there was no conclusion on whether or not the interaction was
synergistic or additive. Wymore, Watson and Gotlieb (1987) however, suggested the interaction
between Thidiazuron and C. coccodes may be synergistic.
The interaction of MCPA 250 and R. alismatis on juvenile plants showed the effect was greatest at the
highest dose (0.0625 of the recommended). There is no evidence to suggest any synergistic effect,
however, there may be some additive effect between the higher sublethal rates of the MCPA 250 and
the fungus. Further studies on this interaction are needed.
The sublethal rates of Londax reduced growth of the starfruit similar to plants only inoculated with
the fungal inoculum. The highest Londax dose (L1 = 0.03125) may be "too high" to draw any
conclusion on the type of interaction with the fungus. The effect of L1 alone was significantly higher
than that of R. alismatis alone. However, if Londax application precedes the fungus the growth
suppression is greatly increased. Particularly at the lower sublethal dose (L2 = 0.015625). This growth
suppression is greater than when the fungus or herbicide is applied alone and is observed when L1
precedes the fungal inoculation. However, if the fungus is applied before Londax no increase in the
growth suppression occurs, indicating there is a synergistic effect between sublethal rates of Londax
and R. alismatis.
The basic theory that low rates of herbicides may act as metabolic inhibitors of plant defence
mechanisms, hence increasing the disease caused by fungal pathogens (Altman, Neate & Rovira
1990), appears to particularly apply in this case. Application of sublethal rates of Londax a few days
before the fungal inoculation, appears to predispose starfruit plants to disease caused by R. alismatis. It
is possible that this synergy may be achieved with further lowering the dose of the herbicide.
Furthermore, the actual mechanisms responsible for the synergistic effect between the fungus and
Londax need to be studied and understood to improve the combined effect of the two.
In summary, the formulation agent in the herbicides that has an inhibitory effect on the fungus needs
to be determined. There is no evidence of synergism between MCPA and R. alismatis, however further
studies on the interaction between them is necessary. There is clear evidence of synergistic interaction
between Londax and the fungus, which needs to be studied further.
43
7. Genetic variability between starfruit

populations
7.1 Introduction
Biological control using plant pathogens should be viewed in the context of the disease triangle
(Templeton 1983). One of the principal factors in the epidemiology of biological weed control is the
genetic diversity of the target species over a geographic area (TeBeest, Yang & Cisar 1992). In a
genetically diverse weed population, there may be biotypes resistant to the pathogen. Although the
host variability is a very important factor in the classical strategy, it has not been reported in the
inundative approach and may be a constraint to the success of a bioherbicide (Auld & Morin 1995).
Nissen et al. (1995) highlights the need to determine the genetic diversity of weed species in
biological control using DNA-based marker techniques. Most of such techniques are based on the
methodology known as polymerase chain reaction (PCR). PCR has had a significant impact on
molecular biology. It allows for amplification of preselected segments of DNA from a very small
amount of an organism's total DNA, given that the sequence information for the target DNA is known
(Nissen et al. 1995) .
PCR is an enzymatic method, the amplification is achieved with two synthetic oligodeoxynucleotide
single-stranded-primers acting on both ends of the template DNA, a polymerase enzyme, called Taq
polymerase and the four deoxyribonucleotide triphosphates. In this process, performed in
programmable heating blocks, first the template DNA is denatured into a single strand, in the next
phase primers anneal to homologous sequences of the DNA and in the final stage the Taq polymerase
copies the DNA. These three cycles are repeated several times and as a result the segments of the
DNA bounded by the two primers are copied over and over. These amplified sequences are then
separated and visualised using gel electrophoresis techniques (Nissen et al. 1995) .
There are several PCR-based techniques. One such method involves the use of simple sequence repeat
primers (SSRs), also called microsatellites. SSRs are repetitive DNA sequences, consisting of
tandemly repeating mono-, di-, tri- or penta-nucleotide units that are arranged throughout the genomes
of a significant portion of higher eukaryotic species (Powell, Machray & Provan 1996). Powell,
Machray and Provan (1996) describe SSRs as an easy and effective method to detect polymorphism
between closely related plant individuals, and it can be used to detect variation in the gene pools of
plants.
Charters et al. (1996) successfully used SSR primers for PCR analysis of several cultivars of Brassica
napus. They found that this method showed greater variability between the cultivars than other
methods such as restriction fragment length polymorphism (RFLP) and the PCR-based randomly
amplified polymorphic DNA (RAPD). Another advantage of microsatellites is the repeatability of the
results which is due to the longer length of the SSR primers, 17-20 bases, compared to RAPD primers
(10-12 bases). Charters et al. (1996) consider the speed in the analysis (PCR, product resolution and
band detection) as a further advantage of SSRs.
The objective of this study was to determine possible genetic variability between different starfruit
populations throughout the major rice growing areas of southern Australia using SSRs.

7.2.1
Plant materials
During the 1997-98 growing season, starfruit seeds were harvested from plants at 34 sites from rice
farms located in the Murrumbidgee (MIA), Coleambally (CIA) and Murray Valley (MVIA) irrigation
areas (Table 7.1), in the North, centre and the South of the rice growing region of NSW respectively.
Seed lots harvested from 18 of these sites were selected, soaked in 200 ppm Ethephon solution
(Section 4.3.1) and transplanted into 10 cm pots placed in 45 x 30 x 23 plastic boxes which were then
44
gradually filled with water. At least five pots with three germinating seeds from each seed lot were
grown in a glasshouse chamber. Plants were later thinned to one plant per pot. The MVIA is by far the
largest irrigation area, therefore a greater number of sites from this region were sampled. In addition to
the starfruit seeds, water plantain (A. plantago-aquatica) which appears to be closely related to the
starfruit, were also grown for DNA extraction.
Floating leaves from the healthiest plants were harvested, briefly surface-sterilised with 1% chlorine
solution and rinsed. Leaves were freeze-dried immediately after harvest.
7.2.2
DNA extraction
DNA of the freeze dried samples were extracted using Qiagen DNeasy Plant Mini Kit following
manufacturers instructions. In total, 48 DNA samples (47 starfruit and one water plantain) were
extracted. Table 7.1 shows the number of DNA extracted from plants grown from the selected seed
lots. Samples were labelled with an identification number (ID) drawn from their seed lot number
(Table 7.1). The DNA stock solutions were preserved in -20C freezer.
7.2.3
DNA quantification and standardisation
DNA grade agarose (Progen), 1% w/v, was rehydrated in Tris-acetate (TAE) with 3 L 100mL-1 of
ethidium bromide stock solution (Appendix 2). Gels were loaded with DNA samples (5ul of the stock
solutions) and 800ng/L of DNA; each with 2 L of loading dye (Appendix 2). Electrophoresis was
conducted at 100 V for 90 min. Gels were viewed on an UV light box using a digital camera (Isis)
and analysed using an image processing software (ImagePro) based on the brightness of the bands.
According to the quantification, different amounts from each DNA stock solution were diluted in trisEDTA buffer (TE buffer) to obtain a final dilution of 10 ng L-1 (see appendix 2 for the recipes of the
chemical solutions).
Table 7.1. Sites from which starfruit seeds were harvested and grown in a glasshouse for DNA
extraction and analysis. The latitude and longitude of each site was recorded with a GPS.
Seed
lot
No.
1*
4
8
10
15
17
19
20
21
22
23
25
26
27
30
31
33
34
Location
Latitude
Longitude
No. of DNA
samples
ID code of DNA
samples
MVIA
MVIA
MVIA
MVIA
MVIA
MVIA
CIA
CIA
CIA
MIA
MIA
MIA
MIA
MIA
MVIA
MVIA
MVIA
MVIA
35 29' 11.7'' S
35 20' 58.3'' S
35 23' 1.8" S
35 19' 08.7"S
35 47' 45.9" S
34 45' 12.9" S
34 51' 50.6" S
34 51 42.5" S
34 54 17.2" S
34 35 40.3" S
34 33' 38.6" S
34 31' 18.9" S
34 30' 40.4" S
34 37' 16.3" S
35 40' 56.4" S
35 42' 24.25"S
35 41' 53.3"S
35 34' 54.2"S
144 36' 37.2'' E

144 13' 32.7'' E
144 13' 57.4" E
144 12' 11.7" E
144 36' 45.9" E
146 00' 20.75" E
145 50' 4.0" E
145 51' 52.3" E
145 55' 06.7" E
146 09' 33.5" E
146 15' 53.9" E
146 18' 02.5" E
146 15' 47.6" E
146 24' 54.4" E
145 40' 19.3" E
145 47' 13.1"E
145 38' 46'E
145 11' 56.1"E
3
3
3
2
2
1(1), 1(2), 1(3)

4(1), 4(2), 4(3)
8(1), 8(2), 8(3)
10(1), 10(2)
15(1), 15(2)
17(1), 17(2)
19(1), 19(2), 19(3)
20(1), 20(2), 20(3)
21(1), 21(2), 21(3)
22(1), 22(2)
23(1), 23(2), 23(3)
25(1), 25(2)
26(1), 26(2), 26(3)
27(1), 27(2), 27(3)
30(1), 30(2), 30(3)
31(1), 31(2), 31(3)
33(1), 33(2), 33(3)
34(1), 34(2), 34(3)
45
3
3
2
3
2
3
3
3
3
3
3
7.2.4
PCR amplification
Due to the large number of samples (48) preliminary tests were carried out for primer selection. Initial
screening of 52 primers (UBC SSR primers set 100/9) was performed on six DNA samples [1(1),
17(1), 19(1), 23(1), 26(2), 30(1)]. The primers were selected on the basis of their sequence and
melting temperatures (Tm). According to Powell, Machray & Provan (1996) poly (A-G) and poly (AT) are more frequent in plants than other sequences. The Tm of the selected primers were 50C, 52C,
53C or 55C.
PCR reaction mixtures (25 L) contained the following components:
a) one unit (0.2 L) of Taq DNA (Promega 5 L-1);
b) 2 L of MgCl2 (Promega 25 mM);
c) 2.5 L PCR Buffer (Promega);
d) 2.5 L of dNTPs (2 mM each);
e) 2.5 L of DNA (10 ng L-1);
f) 3 L of a single SSR primer (0.3 M) and
g) 12.3 L of sterile distilled water.
PCR amplification was performed using a thermal sequencer (FTS-960 Thermal Sequencer Corbet
Research) programmed with 37 cycles as follows:
1) One cycle: initial denaturing of 2 min at 94C;
2) 35 cycles: denaturing for 1 min at 94C,
annealing for 2 min at Tm-5C
extension for 3 mints at 72C
3) One cycle: final extension for 5 min at 72C.
Electrophoresis of PCR products was performed as described earlier at 100 V for 3 h. after which gels
were photographed on an UV light box. Of 52 primers tested, ten showed clear and well differentiated
bands. These primers were tested in all 48 samples.
The banding profiles were scored and compiled as data matrix and a cluster analysis was performed to
examine the distinction of the samples based on their banding profiles, using NTSYS pc (Exeter
software). Only bright polymorphic bands that were well identifiable and clearly different between
samples were scored. Bands were scored as present (1) or absent (0).
7.3 Results
All ten primers produced different banding profiles between samples of starfruit and water-plantain.
Only five primers (Table 7.2) produced polymorphic bands between starfruit DNA samples. The
dendrogram generated as the result of banding profiles produced by these primers, showed little
variability between the DNA samples extracted from plants originating from the same seed lot, with
some exception such as plants fro seed lot number 15, 19 and 30. There was not a clear pattern
between the relationship of samples and their geographic location. The major difference between the
starfruit samples was at 0.275 meaning that they share 72.5% similarity. Water-plantain was separated
from all starfruit samples at 0.76.
46
Table 7.2. SSR primers that showed polymorphism between the populations of starfruit.
Primer No.
Sequence
Annealing
No. of
Tm (C)
temperature
polymorphic
bands
(Tm-5C)
811
(GA)8C
52
47
1
818
(CA)8G
52
47
1
55
50
1
835
(AG)8YC*
53
48
2
852
(TC)8RA**
853
(TC)8RT**
53
48
3
*Y = (C,T)
**R = (A,G)
7.4 Discussion
There is a general agreement that the success of biological control may be limited by high levels of
genetic diversity occurring in target weed species (Nissen et al. 1995; Auld & Morin 1995; TeBeest,
Yang & Cisar 1992).
Prior to the analysis, it was expected that samples from each irrigation area would form clearly
identifiable clusters in the dendrogram; however, the pattern of variability did not always relate to the
geographic location of the samples. This may be because the Australian rice crop is reduced to a
relatively small area where farmers buy seeds from the same source, may use the same contractors for
farm operations, share machinery or other management practices that help the starfruit seeds to travel
from one area to another. In addition, it may also be due to small variability between the plants from
these locations.
Previous work to study the biology of starfruit plants did not show phenotypic differences between
starfruit plants from six locations of MIA, CIA, MVIA (Graham R. J. 1998 pers. comm.). Charters et
al. (1996) found that oilseed rape plants (B. napus) of the same cultivars that do not show phenotypic
variation, contained substantial levels of genetic variability indicated by SSRs. They suggest that such
intra-cultivar variability may take place during the breeding process (i. e. intercrossing)
The results of this experiment showed that in most cases there was little or no variation between plants
grown from the same seed lot. The very small variability between plants originating from seeds on the
same site suggests that intercrossing between plants is not common.
SSRs are powerful tools to detect variation between closely related individuals and they reveal more
polymorphism compared to other assay methods (Powell, Machray & Provan 1996). Nevertheless,
from the ten primers tested on all starfruit samples, only eight polymorphic bands were found from
five primers. Current screening of the genetic variability of different populations of A. lanceolatum,
another potential target of R. alismatis, using SSRs has shown three times more polymorphism that
has been found in the case of starfruit (Raman R. 1999 pers. comm.)
Therefore, there is strong evidence to suggest that the genetic variability between starfruit plants from
different geographical locations is low and should not pose a problem to the success of R. alismatis as
a mycoherbicide. It has been observed however, that starfruit populations from different areas produce
different responses to chemical herbicides. Indeed, a number of herbicide-resistant populations of
starfruit have been reported (Graham et al. 1996). This variable response to chemical herbicides could
be due to the different management approaches rather than genetic variability of starfruit plants. In
addition, it is not expected that plants develop resistance to a disease with the same speed as they do to
chemical herbicides. According to Burge (1988) genetic variation in weeds necessary to cause
resistance to a pathogen needs to be far more extensive than what is required to develop resistance to a
chemical herbicide, due to the specificity of the chemicals in their site of action in the plant.
It is also considered necessary to study the genetic variability of a biocontrol agent (Kutcher &
Mortensen 1999). The variability between different isolates of R. alismatis is currently under study at
47
Charles Sturt University (NSW, Australia) in a project that aims to extend the host range of the fungus
to other Alismataceae weeds.
In summary, SSRs showed little polymorphism between 48 starfruit plants grown from seeds from 18
locations across the rice growing areas of southern Australia. It appears the level of out-crossing
between starfruit plants is small. The variation between them should not cause concern with regard to
the problems it may pose to the use of the biocontrol agent.
48
8. Effect of fungal inoculation on weed

competition
8.1 Introduction
The efficacy of a potential mycoherbicide needs to be established and this is normally done after the
study of inoculum production and epidemiological studies (Morin, Watson & Reeleder 1989).
Charudattan (1989) stated that efficacy can be measured in terms of weed control, the level of disease
stress or increase in crop yield resulting from reduced weed competition. Several weed control
parameters have been established (Charudattan 1989). These may be on the basis of weed kill or
reduction in weed growth and/or reproduction, resulting in reductions in biomass and/or leaf area. This
in turn affects the competitiveness of weeds, favouring the plants of interest (i. e. crops).
Rhynchosporium alismatis does not kill starfruit but it reduces the growth when applied at the juvenile
stage; therefore, it could negatively affect starfruit in competition with rice. The effects of non-lethal
mycoherbicide agents that cause reduction in growth or stunting in weeds have been evaluated in
competition studies. Ditommaso, Watson and Hallet (1996) found that inoculation of velvetleaf by C.
coccodes reduces the competition between the weed and soybean by reducing growth variables of the
weed such as biomass and leaf area.
Furthermore, R. alismatis, if applied to the adult starfruit plants, produces necrosis and chlorosis on
aerial parts, causing leaf mortality. Lee & Bazzaz (1980) investigated the effect of defoliation of
velvetleaf finding that it reduces the competition of the plants in a high-density situation. In addition,
Guntli et al. (1999) found that hedge bindweed (Calystegia sepium R. Br.) may be suppressed as the
result of necrosis caused by Stagonospora convolvuli and by competition with a competitive ground
cover plant such as red clover (Trifolium pratense L.). Therefore, it is possible that inoculation of adult
starfruit plants may also influence the competition of this weed with rice. The objective of this study
was to determine whether inoculation of starfruit plants at the juvenile and adult stages could reduce
the competitiveness of the weed and, therefore, improve rice production as measured by rice biomass
in comparison with a non-inoculated situation.
8.2 Material and methods

The criteria to sow rice and starfruit plants was based on the sequence of direct seeding procedures
that are performed on most Australian rice farms. The experiment, in a glasshouse, was carried out in
20 cm pots placed in 120 x 120 x30 cm tubs which were filled with water to 5 cm above the soil.
Table 8.1 shows the time sequence of the operations.
Starfruit seeds were soaked in 200 ppm ethephon solution (section 4.3.1) and after germination they
were transplanted into 30 x 35 x 5 cm flat trays. Vigorous plants were transplanted to the 20 cm
diameter pots. There were 0, 1, 3 or 5 starfruit plants per pot. Rice seeds (cv. Amaroo) were soaked in
water for 48 h before they were sown in the pots (Table 8.1). Ten seeds were placed on the soil surface
of each pot. Rice seedlings were later thinned to the five most vigorous plants in each pot.
A conidial suspension (5 x 106 conidia mL-1) was harvested from four-day-old cultures on Difco lima
bean agar, as describe in section 3.2.2.1, and applied with an airbrush to run off. Inoculation was
carried out twice; at the first inoculation the inoculum or water was applied (Table 8.1) and followed
by a 24 h dew period (25C in a dew chamber in the dark). This inoculation was performed when the
starfruit plants were between growth stages 1.3 and 1.4 (section 4.3.1) and rice plants were between
10-20 cm tall. The second application of inoculum suspension or water (Table 8.1) was performed
only on three replicates selected at random, without the application of a dew period. In this
inoculation, starfruit plants were between stages 2.7 and 3 (section 4.3.1) and rice plants had tillers.
The experiment was in a split-plot design with no inoculation as sub plot, with six replicates. Each tub
contained 12 pots placed, three for each weed density and in total there were 24 tubs, two in each
49
replicate. Pots were placed in a latin square design along the edges of each tub. Tubs were selected
randomly for one or two inoculations. The four corner pots of each tub were harvested in the first
sample time and the remaining pots for each density were harvested three weeks later. At the first
sample time leaf area and starfruit and rice biomass were measured. In the second sample time only
biomass was determined for both starfruit and rice.
Table 8.1. Calendar of operations for the glasshouse competition experiment.
Operation
Day
soak starfruit in ethephon solution

transplant of starfruit in flat trays
1
7
soak rice (var. Amaroo)
11
saw rice in pots
14
Transplant starfruit onto pots
18
st
1 inoculation
25
nd
2 inoculation
73
st
94
nd
117
1 harvest
2 harvest
8.3 Data analysis

The analysis was performed separately for the two harvest times. To model each trait at a given
sampling time a linear mixed model (LMM) was fitted to take into account all variables.
Letting Y denote the variate of interest, corresponding to the trait expressed in units per plant:
Y = mean + NI + WD + NI.WD + NR + Rep + Tub + error
where NI is a factor corresponding to the number of inoculations, WD corresponds to the number of
weeds, NI.WD denotes the effect of weed density for 0, 1 and 2 inoculations, NR is a covariate for
number of rice plants to allow adjustments for pots with more or less rice plants, whilst Rep, Tub and
error are uncorrelated random components associated with the design. This is because the aim was to
measure the difference between treatments and not between replicates. The covariate NR is included
as it was observed that some pots had one rice plant short or extra than originally intended. The above
models were then simplified by removing non significant terms with the exception of the Tub term (a
design feature) which was only excluded from the model if its variance was almost zero.
8.4 Results
First sampling stage: Some necrosis and chlorosis were observed in the non-inoculated starfruit
plants. Although the severity of the disease on leaves based on visual estimation was higher on plants
inoculated twice, there was no significant difference (P>0.05) in the results for plants inoculated once
or twice. The biomass and leaf area for inoculated plants did not vary with weed density, while they
varied for non-inoculated plants. Table 8.2 shows the result for starfruit biomass and leaf area for the
first sample time.
50
Table 8.2. The effect of inoculation and density on starfruit biomass and leaf area at the first sample
time. Numbers with common letters within columns of biomass and leaf area are not significantly
different.
No of weeds/pots
1
3
5
Predicted values for biomass/plant

(g)
No inoculation
Inoculation
0.569 a
0.255 c
0.418 b
0.255 c
0.267 c
0.255 c
Predicted values for leaf area/plant

(mm2)
No inoculation
Inoculation
7666 a
2432 c
5416 b
2432 c
3166 c
2432 c
Rice biomass per plant did not vary significantly (P>0.05) between inoculated and uninoculated
plants, but did decline significantly with the density of the weed. The estimated mean biomass per rice
plant (Y) is given by:
Y=1.255 - 0.036 x Number of starfruit plants
Predicted values are given in Table 8.3.
Table 8.3. Predicted rice biomass at different starfruit densities at the first sampling time. Numbers
with common letters are not significantly different.
No. of weeds/pot
Predicted rice biomass (g)/plant
0
1.29 a
1
1.22 b
3
1.15 c
5
1.08 d
Second sampling stage: There was no significant difference (P>0.05) between one and two
inoculation in starfruit biomass. The rate of decline in biomass per plant with the increase in weed
density was the same for inoculated and non-inoculated plants (Table 8.4).
Table 8.4. The effect of inoculation and density on starfruit biomass per plant at the second sample
time. Numbers with common letters are not significantly different.
No. of weeds/pot
1
3
5
Predicted values for biomass (g)/plant

No inoculation
Inoculation
0.765 (a)
0.486 (c)
0.623 (b)
0.343 (d)
0.480 (c)
0.200 (d)
The analysis indicated that rice biomass was significantly (P<0.01) higher for inoculated than for non
inoculated pots. There was no significant difference (P>0.05) between the results for 1 or 2
inoculations. The rice biomass per plant did not change at the different weed densities when plants are
inoculated. However, when plants were not inoculated the rice biomass per plant declined with
increasing weed density. The final model for the expected value of Y (rice biomass per plant) is given
by:
Y = 1.97 (with inoculation)
and
Y = 1.97 - 0.101 x Number of starfruit plants
Predicted rice biomass values are given in Table 8.5 and the least significant differences for each weed
density and inoculation levels are shown in Table 8.6.
51
Table 8.5. Predicted values of rice biomass per plant at different starfruit densities, with and without
inoculation for the second sampling stage. Numbers with common letters are not significantly
different.
No of weeds/pot
0
1
3
5
Predicted values for rice biomass/plant (g)

No inoculation
Inoculation
1.97 a
1.97 a
1.87 b
1.97 a
1.67 c
1.97 a
1.47 d
1.97 a
Table 8.6. The least significance differences for comparisons of rice biomass in presence of different
densities of starfruit plants, with (+) and without (-) inoculation for the second sample time.
Inoculation
+
0
1
3
5
0-5
Inoculation
No of weeds pot-1
0
0.048
0.141
0.237
0
1
0.095
0.189
0.048
3
0.095
0.141
5
0.238
0-5
+
8.5 Discussion
The results confirmed that there is inter-specific, and possibly intra-specific, competition in a ricestarfruit system affecting the plants' growth and biomass production. Competition increases with
increasing plant density. Inoculation of starfruit plants may reduce the growth of starfruit, in terms of
biomass, and hence can decrease the level of competition between the weed and rice.
In the first sample time despite the fact that the weed growth had been reduced, the reduction in
competition was not apparent. As time progressed towards the second harvest, the rice continued to
grow but the starfruit was suppressed, hence the difference was bigger and more measurable. The rice
biomass in mixture with diseased weed was higher than that in presence of healthy weed, and not
significantly different from rice in monoculture.
Inoculation caused the highest growth suppression in starfruit at the highest density. At 5 plants per
pot, which is equivalent to approximately 200 plants m-2, it is expected that the plants experience more
inter- and/or intra-specific competition. This result is consistent with other findings that fungal
pathogens have greater negative effect on the host in higher plant density situations (Ditommaso &
Watson 1995; Massion & Lindow 1986). Ditommaso and Watson (1995) found that C. coccodes
affects the competitive ability of velvetleaf in the presence of soybean in more crowded environments.
Further work by Ditommaso, Watson and Hallet (1996) indicated that the effect of the reduction in
competition is largely due to stunting of the weed, which is the case of starfruit when it is inoculated at
the juvenile stage. Massion & Lindow (1986) also found that infection of johnsongrass [Sorghum
halepense (L.) Pers.] by Sphacelotheca holci Jack. resulted in greater reduction of above ground
biomass and fewer and shorter rhizomes at higher density of johnsongrass and/or alfalfa (Medicago
sativa L.).
Paul and Ayres (1987)) found that lettuce (Lactuca scariola L. had a competitive advantage in mixture
with groundsel (Senecio vulgaris L.). This advantage was further exaggerated if groundsel was
infected by a rust fungus (Puccinia langenophorae Cooke). They reported that the effect of the rust on
the weed was expressed in reduction in the dry yield of groundsel, which also appeared to be the case
of starfruit-rice system examined here. Mullerscharer and Rieger (1998) also reported similar results
on the effect of the same rust in reduction of the competitive ability of groundsel in celeriac.
Whether rice has a natural competitive advantage over starfruit, as reported by Paul and Ayres (1987)
for lettuce-groundsel, is beyond the scope of this work, but it is possibly a function of the rice variety.
52
The rice cultivar Amaroo used in this competition experiment is a slow growing rice cultivar (Lewin,
L. 1999 pers. comm.) and it is likely that a faster growing variety would be a more successful
competitor and the effect of inoculation in reducing weed competition would be greater.
Auld and Morin (1995) highlighted the variability in the host population as one constraint in
mycoherbicide development. Potter (1987)showed that the reduction in dry matter yield of a
susceptible biotypes of ryegrass (Lolium perenne L.) to P. coronata Corda f. sp. Lolii Eriksswas
greater in the presence of a resistant type only if plants were infected with the rust fungus. Given the
low genetic variability between starfruit populations (Chapter 7), the different responses of the weed
populations to the disease may not be a matter of concern. Therefore, the competitiveness of the rice
cultivar can influence the effectiveness of inoculation.
A second inoculation of starfruit in the glasshouse experiment, however, did not further reduce the
weeds competition with the rice, nor did it decrease starfruit dry weight compared to one inoculation
only, despite the fact that more severe necrosis was observed in plants inoculated twice. Guntli et al.
(1999) also found that more severe necrosis on hedge bindweed as the result of inoculation with S.
convolvuli does not necessarily translate into biomass reduction. Morin, Watson and Reeleder (1989)
found that the pathogen Phomopsis convolvulus Ormeno, had greater effect on field bindweed
(Convolvulus arvensis L.) in reducing the dry weight when inoculated twice than once; however, they
found that two inoculations are less effective than expected. Similar results were reported by
Ditommaso and Watson (1995) who found that three inoculations of A. theophrasti with C. coccodes
was not more effective than two. Further work by Ditommaso and Watson (1997) recommended a
second inoculation of C. coccodes for a more efficient control of velvetleaf.
It is possible that the second inoculation of starfruit was done too late to have any effect on
competition, or the plants were too developed and the level of defoliation, as the result of the second
inoculation, was not enough. Lee and Bazzaz (1980) found that relative weight of velvetleaf decreased
as the level of defoliation increased. However, they found this trend only in a high-density situation.
Therefore, another possible explanation for the lack effectiveness of the second inoculation of starfruit
may be that the absolute plant density was not high enough.
This research confirms that R. alismatis does not produce mortality of starfruit plants. It should be
emphasised that any consideration of the efficacy of a biocontrol agent should be based on crop yield
rather than injury or mortality of the weed (Paul & Ayres 1987). The difference in rice growth at the
first sample time was not apparent, however in the second sample the rice biomass significantly
higher. Glasshouse and field experiments must be carried out in the future to further study the effect of
the disease on weed competition. Nevertheless, it can be concluded that R. alismatis can suppress the
growth and there is strong evidence that it can reduce the competitive ability of the weed. However,
the virulence of R. alismatis may need improvement. As it was shown previously (Chapter 6) the
growth suppressing effect of the fungus can be greatly increased by sublethal doses of herbicide. The
efficacy of the combined effect of herbicide and R. alismatis needs to be evaluated in future
competition experiments.
53
9. General discussion and conclusions

The purpose of this research was to determine the interaction between R. alismatis and starfruit, with
the purpose of future development of the fungus as a mycoherbicide for the integrated management of
starfruit in rice.
Charudattan (1989); Altman, Neate and Rovira (1990), and Trujillo (1992) reviewed the process to
evaluate of fungal pathogens as mycoherbicides. In general, the pathogen-host system should be
assessed in terms of:
a) pathogen ability inoculum production in artificial culture;
b) the effectiveness in weed control in terms of the degree, speed and ease in weed suppression.
c) incorporation into IWMS
It has been shown (Chapter 3), that R. alismatis produces massive amounts of infective conidia in
artificial culture. Both the amount and infectivity of the inoculum are influenced by cultural
conditions. The most important factors were found to be the nutrient media and temperatures. Lima
bean-based media prove to be the most suitable. The temperature needs to be between 20C and 30C.
The fungus therefore fulfils the first criterion of a successful mycoherbicide.
The second criterion was evaluated in the laboratory and glasshouse (Chapters 4 and 5). Microscopic
studies (Chapter 5) of the early stages of the infection process revealed that appressoria are the
primary means of penetration They start to form soon (4 h) after inoculation. However, this process is
also temperature dependent.
There are a number of factors affecting the level of disease in plants (Chapter 4). Of these, the plant
age and water level at the time of inoculation appeared to be the most significant. It was shown that
the inoculation of juvenile plants caused greater reduction in above-ground biomass and leaf area than
it had on the adult plants. In fact, the disease is expressed in two different ways in juvenile and adult
(floating leaf stage) plants. In the former, the fungal inoculation stunts the young plants, provided the
water level is lowered to the soil level at inoculation, in order to expose the plants to the inoculum. In
adult plants R. alismatis causes necrosis, generally surrounded by chlorosis, on aerial parts of starfruit.
Strobel, Sugawara and Clardy (1987) state the most important group of phytotoxin-producing fungi
are deuteromycetes, the group to which R. alismatis belongs. They also stated that fungal pathogens
with wider host ranges are more likely to produce phytotoxins than those highly host-specific.
Rhynchosporium alismatis causes disease in a number of plant species in the Alismataceae (Cother &
Gilbert 1994b; Cother 1999). In addition, a related species (R. secalis) is known to produce toxins
(Ayesu-Offei & Clare 1971; Mazars et al. 1989). There are, therefore, indications that R. alismatis
may also produce phytotoxin/s responsible for causing the necrosis and chlorosis on adults and
stunting in juvenile plants.
The growth reduction caused by the fungus on juvenile plants is of greater significance than its effect
on floating-leaf plants. It is thought that growth suppression early in the development of plants would
help the crop to out-compete the weed (Charudattan 1989). This hypothesis was tested in a glasshouse
competition experiment where the rice plant growth was measured at different levels of weed
densities, with and without fungal inoculation (Chapter 8). Indeed, inoculation reduced the weed
competition to levels that did not significantly affect rice biomass production. Although inoculation of
adult plants did not appear to have further effects on weed competition, it may still assist in reducing
weed competition by defoliation; however, defoliation needs to be more severe to be effective. It has
been shown that massive defoliation (e.g. 75%) reduction in the foliage can reduce the competition
between plants (Lee & Bazzaz 1980). One of the problems in the development of disease at the
floating leaf stage of starfruit was that it developed to a lesser extent on leaves emerged after
inoculation, under the "still" glasshouse conditions.
Yang and TeBeest (1992) showed that spores of C. gloeosporioides are dispersed by rain to northern
jointvetch plants causing disease under rice field conditions. Glasshouse observations gave indications
that R. alismatis is disseminated to newly emerged floating leaves by water splash. In addition, under
54
field conditions the disease can be observed frequently on most aerial parts in mature plants.
Therefore, inoculation of floating-leaf plants is likely to cause more severe defoliation if followed by
rain. This needs to be evaluated in future field experiments
Other factors have been found to have significant impacts on disease development including fungal
isolate and conidial density. Although some isolates (e.g. RH145) proved to be more virulent, all
isolates used in this research produce significant levels of disease. The minimum concentration of
conidia to cause stunting in juvenile plants is 105 conidia mL-1 under the glasshouse experimental
conditions. However, it could be argued that under field conditions the conidial density may need to be
higher.
Most fungal pathogens that are assessed as mycoherbicides have been shown to require a dew period
to infect the host and to cause greater levels of disease Boyette et al. 1996; TeBeest, Templeton &
Smith 1978; Templeton & Heiny 1989). Boyette et al. (1996) state that formulation of mycoherbicides
is the mixing of the active constituent and inert materials in order to alter the characteristics of the
mycoherbicide to a desirable form. This includes diluting or improving stability and biological
activity. Much effort has been placed on formulation of most mycoherbicides to overcome the
problem raised by the dew requirements (Womack, Eccleston & Burge 1996; Boyette et al. 1996;
Daigle et al. 1990). The great advantage of R. alismatis-starfruit system is that a dew period is not
vital for disease development. Therefore the approach to formulation in this case is different from
other potential mycoherbicides where overcoming the dew period requirement is the main concern.
Future formulation work may focus on other aspects such enhancement of infection by increasing the
virulence.
Although the fungus does not cause plant mortality, it has good efficacy in suppressing the weed to
below the level of economic competitiveness. In spite of this, one important aspect that future work
needs to focus on is the improvement of the fungal virulence through formulation. This may be
achieved by combining the fungus with chemical herbicides.
Smith (1991) included the minimum use of chemicals and the inclusion of biocontrol agents as key
strategies in integrated pest management together with other approaches such as modification of
cultural practices to favour crops. One approach towards this integration is the use of low doses of
chemical herbicides together with a biological agent (Altman, Neate & Rovira 1990). In this work the
interaction of sublethal doses of the two most important herbicides in rice, Londax and MCPA 250,
were studied. Only Londax showed to have synergistic effects with the pathogen when the herbicide
is applied before fungal inoculation. This, clearly, should form a key research area in the formulation
of R. alismatis to improve its efficacy.
Screening of the genetic variability of different starfruit populations showed little variation between
them. This, undoubtedly, is a further advantage of the fungus-plant interaction. Therefore, despite the
fact that plants for all experiments in this work were grown from the same seed source (Yanco, NSW),
the results of the interaction studies may be extrapolated to other weed populations.
55
10. Implications
The most important findings of this research are:
Nutrient media influence inoculum (conidia) viability and infectivity of R. alismatis;
a dew period is not critical to cause disease;
the reduction in the growth of juvenile starfruit plants, as the result of inoculation, can give the
rice crop a competitive advantage over the weed;
there is a synergistic interaction between the fungus and Londax which can enhance the efficacy
of the fungus.
The main question as to whether or not R. alismatis is suitable as the fungal component of a
mycoherbicide has been answered. Further research is, however, needed to take this pathogen to the
final stages of development as a mycoherbicide. Included in this is the need to optimise the mass
production of conidial production. In addition, a more in-depth study of the fungal interaction with
chemical products should be undertaken and it is important to evaluate this pathogen under rice field
conditions.
11. Recommendations
This project is the second in a series successfully progressing along the chain from discovery to
commercial implementation. A continuation of this project (UCS 26A) at Charles Sturt University,
Wagga Wagga, is examining aspects of field application and formulation. By studying each aspect of
the biology of this host-pathogen interaction, this research will deliver real benefits for the rice grower
immediately on release to the industry.
Continued support from the rice industry for this additional approach to control of herbicide-resistant
weeds will ensure its ultimate success.
56
12. References
Adams, E. B. (1988). Fungi in classical biocontrol of weeds. In Fungi in biological control systems,
ed. M. N. Burge. Manchester University Press: Manchester, pp. 111-122.
Ahmed, H. U., Mundt, C. C. & Coakley, S. M. (1995). Host-pathogen relationship of geographically
diverse isolates of Septoria tritici and wheat cultivars. Plant Pathology, 44, pp. 838-847.
Aldrich, R. J. (1984). Principles in weed management. Breto Publishers: North Scituate,
Massachussetts
Alstrom, S. (1990). Fundamentals of weed management in hot climate peasant agriculture: A multidisciplinary study of principles, potentials and practices primarily based on Indian and
international weed science literature. Department of crop production Science, Swedish
University of Agricultural Sciences: Uppsala.
Altman, J. & Campbell, C. L. (1977). Effect of herbicides on plant disease. Annual Review
Phytopathology 15, pp. 361-385.
Altman, J., Neate S. & Rovira A. D. (1990). Herbicide-pathogen interactions and mycoherbicides as
alternative strategies for weed control. In Microbes and microbial products as herbicides, ed.
R. E. Hoagland. American Chemical Society: Washington DC, pp. 240-259.
Auld, B. A. (1993). Mass production of fungi for biopesticide. Plant Protection Quartely, 8(1), pp.
7-9.
Auld, B. A. (1990). Mycoherbicides: one alternative to chemical control of weeds. In Alternatives to
the chemical control of weeds, eds C. Basset, L. J. Whitehouse and J. A. Zabkiewicz.
Ministery of Forestry: Rotorua, New Zeland, pp. 71-73.
Auld, B. A., K. M. Menz, and C. A. Tisdell. (1987). Weed control economics. Academic press:
London, U K.
Auld, B. A. & Morin, L. (1995). Constraints in the development of bioherbicides. Weed Technology,
9, pp. 638-652.
Ayesu-Offei, E. N. & Clare, B. G. (1970). Processes in the infection of barley leaves by
Rhynchosporium secalis. Australian Journal of Biological Science. 23, pp. 299-307.
Ayesu-Offei, E. N. & Clare, B. G. (1971). Symptoms of scald disease induced by toxic metabolites of
Rhynchosporium secalis. Australian Journal Biological Science, 24, pp. 169-174.
Baker, C. A. & Henis J. M. S. (1990). Commercial production and formulation of microbial biocontrol
agents. In New directions in biological control: Alternatives for suppressing agricultural pests
and diseases, ed. R. R. Baker. Alan R Liss: New York, pp. 333-344.
Baltazar, A. M. & DeDatta, S. K. (1992). Weed management in rice. Weed Abstracts, 41(11), pp.
495-07.
Barker, R. & Hayami, Y. (1983). Weed control practices as component of rice production systems In
Proceedings of the conference on weed control in rice, International Rice Research institute
Beecher, H. G., Thompson, J. A., Bacon, P. E. & Heenan, D. P. (1994). Effect of cropping sequences
on soil nitrogen levels, rice growth, and grain yields. Australian Journal of Experimental
Agriculture, 34, pp. 977-86.
Bewick, T. A. (1996). Technological advancement in biological control with microorganisms: An
introduction. Weed Technology, 10, p. 600.
Blair, B. D. & Parochetti, J. V. (1982). Extension and implentation of integrated pest management
systems. Weed Science, 30, pp. Supplement 48-53.
Bos, L. & Parlevliet, J. E. (1995). Concepts and terminology on relationships: Toward consensus in
57
plan pathology and crop protection. Annual Review Phytopathology, 33, pp. 269-102.
Boyette, C. D., Quimby Jr., P. C., Caesar, J. A., Birdsall J.L., Connick Jr., W. J., Daigle, D. J.,
Jackson, M. A., Egley, G. H. & Abbas, H. K. (1996). Adjuvants, formulations, and spraying
systems for improvement of mycoherbicides. Weed Technology, 10, pp. 637-644.
Boyette, C. D., Quimby Jr. P. C., Connick Jr W. J., Daigle. D.J. & Fulgham F.E. (1991). Progress in
the production, formulation and application of mycoherbicides. In Microbes and Microbial
Products as Herbicides, ed. R. E. Hoagland. American Chemical Society: Washington DC,
pp. 209-222.
Braun, U. (1995). A monograph of Cercosporella, Ramularia and allied genera (Phytopathogenic
Hyphomycetes). IHV-Verlag: Munchen.
Braun, U. (1993). New genera of phytopathogenic Deuteromycetes. Cruptogamic Botany, 4, pp.
107-114.
Brennan, J. P., Lewin, L. G., Marshall, G. R. & Reinke, R. F. (1994). An economic assessment of rice
breeding in New south Wales. Australian Journal of Experimental Agriculture, 34, pp.
889-96.
Bruzzese, E. (1993). The role of biological in weed management - what is success. Plant Protection
Quarterly, 8(4), pp. 147-150.
Burge, M. N. (1988). The scope of fungi in biological control. In Fungi in biological control systems,
ed. M. N. Burge. Manchester University Press: Manchester, pp. 1-107.
Carey, V. F., Smith, R. J. & Tablert, R. E. (1994). Interference duration of bearded sprangletop
(Leptochloa fascicularis in rice (Oryza sativa). Weed Science, 42, pp. 180-183.
Cate, J. R. (1990). Biological control of pests and diseases: integrating a diverse heritage. In New
directions in biological control: Alternatives for suppressing agricultural pests and disease,
ed. R. R. Baker. Alan R Liss: New York, pp. 23-43.
Charters, Y. M., Robertson, A., Wilkinson, M. J. & Ramsay, G. (1996). PCR analysis of oilseed rape
cultivars (Brassica napus L. ssp. oleifera) using 5' anchored simple sequence repeat (SSR)
primers. Theory and Application of Genetics, 92, pp. 442-447.
Charudattan, R. (1985). The use of genetically altered strains of pathogens for weed control. In
Biological control in agricultural IPM systems, eds M. A. Hoy and D. C. Herzog. Academic
Press: London, pp. 347-372.
Charudattan, R. (1986). Integrated control of waterhycinth (Eichhornia crassiopes) with pathogen,
insects, and herbicides. Weed Science, 34, p. supplement 1: 26-30.
Charudattan, R. (1988). Inundative control of weeds with indigenous fungal pathogens. In Fungi in
biological control systems, ed. M. N. Burge. Manchester University Press: Manchester, pp.
85-109.
Charudattan, R. (1989). Assessment of efficacy of mycoherbicide candidates In Proceedings VII
international symposium on biological control of weeds, E. S. Delfosse. pp. 455-464.
Charudattan, R. (1991). The mycoherbicide approach with plant pathogens. In Microbial control of
weeds, ed. D. O. TeBeest. Chapman & Hall: New York, pp. 24-56.
Charudattan, R. (1990a). Pathogens with potential for weed control. In Microbes and microbial
products as herbicides, ed. R. E. Hoagland. American Chemical Society: Washington DC, pp.
132-154.
Charudattan, R. (1990b). Prospects for biological control of weeds by plant pathogens. Fitopatologia
Brasileiro, 15(1), pp. 13-19.
Charudattan, R. & DeLoach C. J. (1988). Management of pathogens and Insects for weed control in
agroecosystems. In Weed Management in agroecosystems: ecological approaches, eds M. A.
58
Altieri and M. Liebman. CRC press. Boca Raton: Florida, pp. 245-264.
Charudattan, R., DeValerio J. T. & Prange V.J. (1990). Special problems associated with aquatic weed
control. In New direction in biological control: Alternatives for suppressing agricultural pests
and diseases, ed. R. R. Baker. Alan, R. Liss: New York, pp. 287-303.
Chisaka, H. (1977). Weed damage to crops: Yield loss due to weed competition. Integrated control of
weeds, eds J. D. Fryer and S. Matsunaka. Japan Scientific Society Press: Tokyo, pp. 1-16.
Combellback, J. H. (1992). An appraisal of opportunities to reduce herbicide use. Plant Protection
Quarterly, 7(2), pp. 66-69.
Cother, E. J. (1994). Challenging current perceptions of biological control: A pathologists perspective
of the future of bioherbicides in weed management in Asian rice production In Proceedings of
international conference on herbicide use in Asian rice production, International Rice
Research Institute: Manila,
Cother, E. J. (1999). Host range studies of the mycoherbistat fungus Rhynchosporium alismatis.
Australasian Plant Pathology, 28, pp. 149-155.
Cother, E. J., Gilbert, R. I. & Pollock, D. C. (1994). First record of Rhynchosporium alismatis on
Alisma lanceolatum and Damasonium minus. Australasian Plant Pathology, 23, pp. 46-49.
Cother, E. J. & Gilbert, R. L. (1994a). Efficacy of a potential mycoherbicide for control of Alisma
lanceolatum and Damasonium minus in Australian rice crops. Australian Journal of
Experimental Agriculture, 34, pp. 1043-50.
Cother, E. J. & Gilbert, R. L. (1994b). Pathogenicity of Rhynchosporium alismatis and its potential as
a mycoherbicide on several weed species in the Alismataceae. Australian Journal of
Experimental Agriculture, 34, pp. 1039-42.
Cother, E. J. & van de Ven, R. (1999). The influence of nutrition on conidial production by
Rhynchosporium alismatis and their subsequent infectivity to Alisma lanceolatum. Biocontrol
Science and Technology, 9, pp. 395-407.
Cox, B. 1984. "Weeds." In Rice growing in New South Wales, ed.
Agriculture New South Wales and Rice Research Committee:
A. Currey. Department of
Craig, S. & Beaton, C. D. (1996). A simple cryo-SEM method for delicate plant tissues. Journal of
Microscopy, 182, pp. 102-105.
Crump, N. S., Cother, E. J. & Ash, G. J. (1999). Clarifying the nomenclature in microbial weed
control. Biocontrol Science and Technology, 9, pp. 89-97.
Cullen, J. M. (1992). Prediction and practice in biological control In Proceedings of the eight
international symposium on biological control of weeds, E. S. Delfosse and R. R. Scott.
CSIRO: Melbourne, pp. 103-109.
Cullen, J. M. & Hasan S. (1988). Pathogens for the control of weeds. In Biological control of pests,
pathogens and weeds: developments and prospects, eds R. K. S. Woods and M. J. Ways. The
Royal Society: London, pp. 213-222.
Daigle, D. J., Connick Jr, W. J., Quimby, C. P., Evans, J., Trask-Morrell, B. & Fulgham, F. E. (1990).
Invert emulsions: carrier and water source for the mycoherbicide, Alternaria cassiae. Weed
Technology, 4, pp. 327-331.
Daniel, J. T., Templeton, G. E., Smith, R. J. & Fox, W. T. (1973). Biological control of northern
jointvetch in rice with an endemic fungal disease. Weed Science, 21(4), pp. 303-307.
Debach, P. (1964). The Scope of biological control. In Biological control of insect pests and weeds,
eds P. Debach and E. I. Sclinger. Reinhold: New York,
Debach, P. & Rosen, D. (1991). Biological control by natural enemies, Cambridge university Press:
Cambridge
59
Dhingara, D. O. & B. J. Sanders. (1985). Basic plant pathology. CRC Press: Boca Raton, Florida.
Dilday, R. H., Lin., J. & Yan, W. (1994). Identification of allelopathy in the USDA-ARS rice
germplasm collection. Australian Journal of Experimental Agriculture, 34, pp. 907-910.
Ditommaso, A. & Watson, A. K. (1997). Effect of the fungal pathogen, Colletotrichum coccodes, on
Abutilon theophrasti height hierarchy development . Journal of Applied Ecology, 34(2), pp.
518-529.
Ditommaso, A. & Watson, A. K. (1995). Impact of a fungal pathogen, Colletortichum coccodes on
growth and competitve ability of Abutilon theophrasti. New Phytologist, 131, pp. 51-60.
Ditommaso, A., Watson, A. K. & Hallet, S. G. (1996). Infection by the fungal pathogen
Colletotrichum coccodes affects velvetleaf (Abutilon theophrasti)-soybean competition in the
field. Weed Science, 44(4 ), pp. 924-933.
Ehler, L. E. (1992). Evolutionary history of pest-enemy associations. In Proceedings of the eight
international symposium on biological control of weeds, E. S. Delfosse and R. R. Scott.
CSIRO: Melbourne, pp. 83-91.
Elmore, C. L. (1996). A reintroduction to integrated weed management. Weed Science, 44, pp.
409-412.
Emmett, R. W. & Parbery, D. G. (1975). Appressoria. Annual Review Phytopathology, 13, pp.
147-167.
Evans, G., Rowland, P. C. & McLean, G. D. (1996). Herbicide dependent agriculture: Have we gone
too Far? In Proceedings of the eleventh Australian weed conference, R. C. H. Shepherd. Weed
Science Society of Victoria, pp. 506-510.
Fischer, A. J., Lozano, J., Ramirez, A. & Sanint, L. R. (1993). Yield loss prediction for integrated
weed management in rice. International Journal of Pest Management , 39(2), pp. 175-180.
Fischer, A. J. & Ramirez, A. (1993). Red rice (Oryza sativa): competition studies for management
decisions. International Journal of Pest Management, 39(2), pp. 133-138.
Fox, K. M. (1995). Influence of infection by Rhynchosporium alismatis on the seed production of
starfruit (Damasonium minus). Charles Sturt University, Wagga Wagga, NSW.
Franz, J. E., M. K. Mao, and J. A. Sikorski. (1997). Glyphosate: A unique global herbicide. American
Chemical Society: Washington, DC.
Glauninger, J. & Holzner W. (1982). Interference between weeds and crops: A review of literature. In
Biology and ecology of weeds, eds W. Holzner and N. Numata. Junk W Publisher: The Hague,
pp. 149-157.
Graham, R. J., Pratley, J. E., Slater, P. D. & Baines, P. R. (1996). Herbicide resistant aquatic weeds, a
problem in New South Wales In Proceedings of 11th Australian weeds conference , R. C. H.
Shepherd. Weed science Society of Victoria: Melbourne, pp. 156-158.
Guntli, D., Burgos, S., Kump, I., Heeb, M., Pfirter, H. A. & Defago, G. (1999). Biological control of
hedge bindweed (Calystegia sepium) with Stagonospora convolvuli strain LA39 in
combination with competition from red clover (Trifolium pratense). Biological Control, 15(3),
pp. 252-258.
Hill, G. D. (1982). Herbicide technology for integrated weed management systems. Weed Science, 30,
pp. Supplement 35-39.
Hill, J. E., De Datta, S. K. & Real, J. G. (1990). Echinochloa competition in rice: A comparison of
studies from direct-seeded and transplanted flooded rice Proceeding of the symposium on
weed management, B. A. Auld, R C. Umaly, and S. S. Tjitrosomo. Seameo-Biotroppp.
115-129.
Hill, J. E., Smith, Jr. R. J. & Boyer, D. E. (1994). Rice weed control: current technology and emerging
60
issues in temperate rice. Australian Journal of Experimental Agriculture, 34, pp. 1021-1029.
Holt, J. S. & Lebaron, H. M. (1990). Significance and distribution of herbicide resistance. Weed
Howard, R. J. (1997). Breaching the outer barriers - cuticle and cell wall penetration. In The mycota. A
comprehensive treatise on fungi as experimental systems for basic and applied research, vol.
V, eds K. Esser and P. A. Lemke. pp.43-60.
Huber, L. & Gillespie, T. J. (1992). Modelling leaf wetness in relation to plant disease epidemiology.
Annual Reviews Phytopathology, 30, pp. 553-577.
Huffaker, C. B. (1957). Fundamentals of biological control of weeds. Hilgardia, California
agricultural experiment station, 7(3), pp. 101-153
Jackson, M. A. & Schisler, D. A. (1992). The composition and attributes of Colletortichum truncatum
spores are altered by the nutritional environment. Applied Environmental Microbiology, 58(7),
pp. 2260-2265.
Jackson, M. A., Schisler, D. A., Slininger, P. J., Boyette, C. D., Silman, R.W. & Bothast, R. J. (1996).
Fermentation strategies for improving the fitness of a bioherbicide. Weed Technology, 10, pp.
645-650.
Jahromi, F. G., Ash, G. J. & Cother, E. J. (1998). Influence of cultural and environmental conditions
on conidial production, conidial germination and infectivity of Rhynchosporium alismatis, a
candidate mycoherbicide. Australasian Plant Pathology, 27(180-185).
Joye, G. F. (1990). Biological control of aquatic weeds with plant pathogens. In Microbes and
microbial products as herbicide, ed. R. E. Hoagland. American chemical Society:
Washington, DC., pp. 155-174.
Jutsum, A. R. (1988). Commercial application of biological control, status and prospects. In Biological
control of pests, pathogens and weeds: developments and prospects, eds R. K. S. Woods and
M. J. Way. The Royal Society: London, pp. 357-373.
Kenney, D. S. (1986). DeVine-The way it was developed-An industrialists view. Weed Science, 34, p.
Supplement 1:15-16.
Klein, T. A. & Auld B.A. (1995). Influence of spore dose and water volume on a mycoherbicides
efficacy in field trials. Biological Control, 5(173-178).
Klerk, R. A., Smith, R. J. & TeBeest, O. D. (1985). Integration of a microbial herbicide into weed and
pest control programs in rice (Oryza sativa). Weed Science, 33, pp. 95-99.
Kremer, R. J. & Kennedy, A. C. (1996). Rhizobacteria as biocontrol agents of weeds. Weed
Kropff, M. J. (1993). Eco-Physiological models for crop-weed competition. In Modelling crop-weed
interactions, eds M. J. Kropff and Van Laar H.H. X: X, pp. 25-32.
Kutcher, H. R. & Mortensen K. (1999). Genotypic and pathogenic variation of Colletotrichum
gloeosporioides f. sp. malvae. Canadian Journal of Plant Pathology, 21(1), pp. 37-41.
Lattimore, M.-A. E. (1994). Pastures in temperate rice of south-eastern Australia. Australian Journal
of Experimental Agriculture, 34, pp. 959-65.
Lee, T. D. & Bazzaz, F. A. (1980). Effects of defoliation and competition on growth and reproduction
in the annual plant Abutilon theophrasti. Journal of Ecology, 68, pp. 813-821.
Levesque, C. A. & Rahe, E. J. (1992). Herbicide interactions with fungal root pathogens, with special
reference to Glyphosate. Annual Reviews Phytopathology, 30(579-602).
Lewin, L. & Heenan D. (1984). The rice plant. In Rice growing in New South Wales, ed. A. Currey.
Department of Agriculture New South Wales and Rice Research Committee.
61
Lovett, J. V. & Knights, S. E. (1996). Where in the world is weed science going? In Proceedings of
eleventh Australian weed conference, R. C. H. Shepherd. Weed Science Society of Victoria,
pp. 2-13.
Luo, Y. & TeBeest, D. O. (1999). Effect of temperature and dew period on infection of northern
jointvetch by wild-type and mutant strains of Colletotrichum gloeosporioides f. sp.
aeschynomene. Biological Control, 14, pp. 1-6.
Makowski, R. M. D. (1993). Effect of inoculum Concentration, Temperature, dew period, and plant
growth stage on disease development of round leaved mallow and velvetleaf by
Colletotrichum gloeosporioides f. sp. malvae. Phytopathology, 83(11), pp. 1229-1234.
Makowski, R. M. D. & Mortensen, K. (1992). The first mycoherbicide in Canada: Colelletotrichum
gloeosporioides f. sp. malvae for round-leaved mallow control In Proceedings of the first
international weed control congress, Vol. Vol II, J. H. Combellack, K. J. Levick, J. Parsons,
and R. G. Richardson. Weed Science Society of Victoria, pp. 298-300.
Massion, C. L. & Lindow, S. E. (1986). Effects of Sphacelotheca holci infection on morphology and
competitiveness of johansongrass (Sorghum halepense). Weed Science, 34, pp. 883-888.
Mazars, C., Rossignol, M., Marquet, P. & Auriol, P. (1989). Reassessment of the toxic glycoprotein
isolated from Rhynchosporium secalis (Oud.) Davis culture filtrates: Physicochemical
properties and evidence of its presence in infected barley. Plant Science, 65, pp. 165-174.
McDonald, D. J. (1994). Temperate rice technology for the 21st century: an Australian example.
Australian Journal of Experimental Agriculture, 34, pp. 877-88.
McIntyre, S. & Barret, S. C. H. (1985). A comparison of weed communities of rice in Australia and
California. Proceedings of Ecological Society of Australia, 14, pp. 237-250.
McIntyre, S., Finlayson, C. M., Ladiges, P. Y. & Mitchell, D. S. (1991). Weed community
composition and rice husbandry practices in New South Wales, Australia. Agriculture,
Ecosystems and Environment, 35, pp. 27-45.
McIntyre, S., Ladiges, P. Y. & Adams, G. (1988). Plant species-richness and invasion by exotics in
relation to disturbance of wetland communities on the Riverina Plain, NSW. Australian
Journal of Ecology, 13, pp. 361-373.
McRae, C. F. (1988). Classical and inundative approaches to biological weed control compared. Plant
Protection Quarterly, 3(3), pp. 124-127.
McRae, C. F. & Auld, B. A. (1988). The influence of environmental factors on anthracnose Xanthium
spinosum. Phytopathology, 78(9), pp. 1182-1186.
McWhorter, C. G. & Shaw, W. C. (1982). Research needs for integrated weed management systems.
Weed Science, 30, pp. Supplement 40-45.
Medd, R. W. (1992). Directions for bioherbicide research in Australia. Plant Protection Quarterly,
7(4), pp. 151-153.
Moore-Landecker, E. (1990). Fundamentals of the Fungi. Third edition ed. Prentice Hall, Englewood
cliffs: New Jersey.
Morin, L. (1993). The prospects for biological control of the noogoora burr complex with naturally
occurring fungi in Australia. PHD Thesis, University of New England, Armidale NSW.
Morin, L., Watson, A. K. & Reeleder, R. D. (1989). Efficacy of Phomopsis convolvulus for control of
field bindweed (Convolvulus arvensis). Weed Science, 37(830-835).
Mortensen, K. (1986). Biological control of weeds with plant pathogens. Canadian Journal of Plant
Pathology, 8, pp. 229-231.
Mullerscharer, H. & Rieger, S. (1998). Epidemic spread of the rust fungus Puccinia lagenphorae and
its impact on the competitive ability of Senecio vulgaris in celeriac during early development.
62
Biocontrol Science & Technology, 8(1), pp. 59-72.

Nissen, S. J., Master, R. A., Lee, D. J. & Rowe, M. L. (1995). DNA-based marker systems to
determine genetic diversity of weedy species and their application to biocontrol. Weed
Science, 43, pp. 504-513.
Norris, R. F. (1992). Have ecological and biological studies improved weed control strategies? In
Proceedings of the first international Weed Control Congress, Vol. 1, Levick K. J. P. J. a. R.
R. G. Combellack J.H. Weed Science Society of Victoria, pp. 7-33.
Paul, N. D. & Ayres, P. G. (1987). Effects of rust infection of Senecio vulgaris on competition with
lettuce. Weed Research, 27, pp. 431-441.
Pennell, D. J. (1994). Economic justification for government involvement in weed management: a
catalogue of market failures. Plant Protection Quarterly, 9(4), pp. 131-137.
Penner, D. (1982). Integrated approach to teaching integrated pest management. Weed Science, 30, pp.
Supplement 46-47.
Podila, G. K., Roger, L. M., Kolattukudy, P. E., Roger, L. M. & Kolattukudy, P. E. (1993). Chemical
signals from avocado surface wax trigger germination and appressorium formation in
Colletotrichum gloeosporioides. Plants Physiology, 103, pp. 267-272.
Potter, L. R. (1987). Effect of crown rust on regrowth, competitive ability and nutritional quality of
prennial and Italian ryegrasses. Plant Pathology, 36, pp. 455-461.
Powell, K. A. & Faull J. L. (1989). Commercial approaches to the use of biological control agents. In
Biotechnology of fungi for improving plant growth, eds J. M. Whipps and R. D. Lumsden.
Cambridge University Press: Cambridge, pp. 259-275.
Powell, W., Machray, G. C. & Provan, J. (1996). Polymorphism revealed by simple sequence repeats.
Trends in Plant Science, 1(7), pp. 215-221.
Powles, S. B. & Howat, P. D. (1990). Herbicide resistance weeds in Australia. Weed Technology, 4,
pp. 178-185.
Prasad, R. (1994). Influence of several pesticides and adjuvants on Chondrostereum purpureum-A
bioherbicide agent for control of forest weeds. Weed Technology, 8, pp. 445-449.
Punithalingam, E. (1988). Ascochyta II. species on Monocotyledons (excluding grasses), cryptogams
and gymnosperms. Mycological Papers, 159, pp. 157-162.
Ricegrowers' Co-operative Ltd. (n.d.). Rice sales and marketing. Leeton, NSW, Australia
Roderick, H. W. & Thomas, B. J. (1997). Infection of ryegrass by three rust fungi (Puccinia coronata,
P. graminis and P. loliina) and some effects of temperature on the establishment of the disease
and sporulation. Plant Pathology, 46, pp. 751-761.
Roger, P. A., Simpson, I., Oficial R., Ardales, S. & Jimenez R. (1994). Effects of pesticides on soil
and water microflora and mesofauna in wetland rice fields: a summary of current knowledge
and extrapolation to temperate environments. Australian Journal of Experimental Agriculture,
34, pp. 1057-1068.
Rohringer, R., Kim, W. K. & Samborski, D. J. (1979). A histological study of interactions between
avirulent races of stem rust and wheat containing resistance genes Sr5, Sr6, Sr8, or Sr22.
Canadian Journal of Botany, 57, pp. 324-331.
Ross, M. A. & Lembi, C. A. (1985). Applied weed science, Burgess Publishing Company:
Minneapolis, Minnesota
Sanders, B. A. (1994). The life cycle and ecology of Cyperus difformis (rice weed in temperate
Australia. Australian Journal of Experimental Agriculture, 34, pp. 1031-1038.
Shabana, Y. M., Charudattan, R. & Elwakil, M. A. (1995). Identification, pathogenicity and safety of
Alternaria eichhorniae from Egypt as a bioherbicide agent for Waterhyacinth. Biological
63
Control, 5, pp. 123-135.

Shaw, W. C. (1982). Integrated weed management systems for pest management. Weed Science, 30,
pp. Supplement 3-11.
Shepherd, R. C. H. (1993). The role of biological control in weed management. Plant Protection
Quarterly, 8(4), pp. 144-146.
Smith, D. & A. H. S. Onions. (1994). The preservation and maintenance of living fungi. Second
edition ed. CAB International: Wallingford, UK.
Smith, H. S. (1919). On some phases of insect control by biological method. Journal of Economic
Entomology, 12, pp. 288-292.
Smith, Jr. R. J. (1983). Weeds of major economic importance in rice and yield losses due to weed
competition Proceeding of the conference on Weed Control in Rice, International Rice
Research Institute, pp. 19-36.
Smith, Jr. R. J. (1986). Biological control of northern jointvetch (Aeschynomene virginica) in rice
(Oryza sativa) and soybean (Glycine max) - A researchers view. Weed Science, 34, p.
Supplement 1 17-23.
Smith, Jr. R. J. (1991). Integration of biological control agents with chemical pesticides. In Microbial
Control of Weeds, ed. D. O. TeBeest. Chapman and Hall Inc: London,
Stant, M. Y. (1963). Anatomy of the Alismataceae. Journal of the Linnean Society, 59, pp. 1-42.
Staples, R. C. & Hoch H. C. (1997). Physical and chemical cues for spore germination and
appressorium formation by fungal pathogens. In The Mycota. a comprehensive treatise on
fungi as experimental systems for Basic and applied research, vol. V, eds K. Esser and P. A.
Lemke. Springer: Berlin, pp. 27-40.
Strobel, G., Sugawara F. & Clardy J. (1987). Phytotoxins from plant pathogens of weedy plants. In
Allelochemicals: Role in agriculture and forestry, ed. G. R. Waller. American Chemical
Society: pp. 516-523.
Swanton, C. J. & Murphy, S. D. (1996). Weed science beyond the weeds: the role of integrated weed
management in agroecosystem health. Weed Science, 44, pp. 437-445.
TeBeest, D. O. (1990). Conflicts and strategies for future development of mycoherbicides. New
ed. R. R. Baker. Alan R Liss, Inc: New York, pp. 323-332.
TeBeest , D. O. (1992). Status of research on bioherbicides. Plant Protection Quartely, 7(4), p. 142.
TeBeest, D. O. (1993). Biological control of weeds with fungal plant pathogens. In Exploitation of
microorganisms, ed. G. Jones. Chapman and Hall: London, pp. 1-17.
TeBeest, D. O. (1996). Issues and prospects confronting biological control of weeds with plant
pathogens in the future. Phytopathology, 24(2), pp. 91-95.
TeBeest, D. O., Templeton, G. E. & Smith, J. R. J. (1978). Temperature and moisture requirements for
development of anthracnose on northern jointvetch. Phytopathology, 68, pp. 389-393.
TeBeest, D. O., Yang, X. B. & Cisar, C. R. (1992). The status of biological control of weeds with
fungal pathogens. Annual Review Phytopathology, 30, pp. 637-57.
Templeton, G. E. (1983). Integrating biological control of weeds in rice into a weed control program.
Weed control in rice, International Rice Research Institute: Manilla, pp. 219-225.
Templeton, G. E. (1982). Biological herbicides: discovery, development, deployment. Weed Science,
30, pp. 430-433.
Templeton, G. E. (1986). Mycoherbicide research at the university of Arkansas - past, present, and
future. Weeds Science, 34, p. supplement 1: 35-37.
64
Templeton, G. E. (1992a). Some orphaned mycoherbicides and their potential for development. Plant
Protection Quarterly, 7(4), pp. 149-150.
Templeton, G. E. (1992b). Potential for developing and marketing mycoherbicides In Preceedings of
the first international weed control congress, Vol. 1, J. H. Combellack, K. J. Levick, J.
Parsons, and R. G. Richardson. Weed Science Society of Victoria, pp. 264-268.
Templeton, G. E. (1990). Weed control with pathogens, future needs and direction. In Microbes and
microbial products as herbicide, ed. R. E. Hoagland. American chemical Society:
Washington, DC, pp. 320-329.
Templeton, G. E. & Heiny D. K. (1989). Improvement of fungi to enhance mycoherbicides potential.
In Biotechnology of fungi for improving plant growth, eds J. M. Whipps and R. D. Lumsden.
Cambridge University Press: Cambridge, pp. 127-151.
Templeton, G. E. & Heiny D. K. (1990). Mycoherbicides. in New Directions in Biological Control:
Alternatives for suppressing agricultural pests and diseases, ed. R. R. Baker. Alan R. Liss
Inc.: New York, pp. 279-286.
Tomely, A. J. & Hardwick, G. (1996). Bulking up, field distribution and establishment of rubbervine
rust, Maravilia cryptostegiae in far north Queensland In Proceedings of eleventh australian
weed conference , R. C. H. Shepherd. Weed Science Society of Victoria: Melbourne, pp. 237238.
Trujillo. E.E. (1992). Bioherbicides. In Frontiers in Industrial Mycology, ed. G. F. Leatham. Chapman
and Hall: London, pp. 197-211.
Waage, J. (1992). Classical biological control In Proceeding of the first international weed control
congress, Vol. Vol 1, J. H. Combellack, K. J. Levick, J. Parsons, and R. G. Richardson. Weed
Science Society of Victoria: Melbourne, pp. 240-249.
Wardle. (1992). Side effects of herbicides on the soil microbial biomass In Proceedings of the first
international weed control congress, Vol. Vol. II, Weed Science Society of Victoria, pp. 561563.
Watson, A. K. (1992). Biological and other alternative control measures. Proceedings of the first
international weed control congress, Vol. Vol.1, Weed Science Society of Victoria:
Melbourne, pp. 298-300.
Watson, A. K. (1991). The classical approach with plant pathogens. In Microbial control of weeds, ed.
O. T. TeBeest. Chapman & Hall: London, pp. 3-23.
Watson, A. K. & Wymore L. A. (1990). Identifying limiting factors in the biocontrol of weeds. In New
ed. R. R. Baker. Alan R Liss, Inc.: New York, pp. 305-316.
Williams, R. J. (1992). Management of weeds in the year 2000. Pest management and the
environment in 2000, eds S. A. A. Aziz, H. Kadir and H. S. Barlow. CAB International: U.K.,
Womack, J. G., Eccleston, G. M. & Burge, M. N. (1996). A vegetable oil-based invert emulsion for
mycoherbicide delivery. Biological Control, 6(23-28).
Woods, R. K. S. & Way M. J. (1988). Preface. In Biological control of pests, pathogens and weeds:
developments and prospects, eds R. K. S. Wood and M. J. Way. The Royal Society: London,
Wymore, L. A., Watson, A. K. & Gotlieb, A. R. (1987). Interaction between Colletotrichum coccodes
and Thidiazuron for control of Velvetleaf (Abutilon theophrastis). Weed Science, 35, pp.
377-383.
Yang, X. B. & TeBeest, D. O. (1992). Rain dispersal of Colletotrichum gloeosporioides in simulated
rice field conditions. Phtopathology, 82(10), pp. 1219-1222.
Yates, I. E. & Cason, K. M. T. (1996). Pecan scab fungus lacks substrate specificity for early
infection stages. Journal of the American Society for Horticultural Science, 121(5), pp.
65
948-953.
Yim, K. O. & Bayer, D. E. (1996). Root growth inhibition of rice by bensulfuron. Weed Research, 36,
pp. 49-54.
Zimdahl, R. L. (1988). The concept and application of the critical weed-free period. Weed
management in agroecosystems: Ecological Approaches, eds M. A. Altieri and Liebman M.
CRC Press: Boca Raton, Florida.
Zoschke, A. (1990). Yield loss in tropical rice as influenced by the competition of weed flora and the
timing of its elimination. Pest management in rice, eds B. T. Grayson, M. B. Green and L. G.
Copping. Elsevier Science: London, pp. 301-313.
66

Weed Control in Rice

Uploaded by

Copyright:

Available Formats

Weed Control in Rice

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Weed Control in Rice

Uploaded by

Copyright:

Available Formats

Weed Control

Suitability of Rhynchosporium alismatis

A report for the Rural Industries Research

By Farzad Jahromi, Eric Cother and

RIRDC Publication No 01/39

2001 Rural Industries Research and Development Corporation.

RIRDC Contact Details

Published in May 2001

A special stipend from Charles Sturt University.

3.4.2.4 Effect of water depth ............................................................................ 21

Effect of temperature on growth and appressorium formation on cellophane paper

2.1 Rice (Oryza sativa)

2.2 The Australian rice industry

2.3 Weed management in rice

2.4 Losses due to weed competition

2.5 Weed control

2.6 Problems associated with chemical herbicides

2.7 Integrated weed management systems

2.8 Biological weed control

Definitions and terminology

History and general principles

Biological control of weeds with pathogens

Biocontrol strategies using pathogenic fungi

2.8.4.2 Augmentative approach

2.8.4.3 Inundative approach

2.9 The host target and its pathogen

2.9.2 The fungus (Rhynchosporium alismatis)

Table 2.1. Different taxonomical classification of Rhynchosporium alismatis

2.9.2.2 Distribution and host range

Environmental and Cultural Factors

3.2 Materials and methods

Fungal isolation and storage

3.2.2.1 Effect of media

3.2.3.1 Effect of media

3.2.3.3 Effect of temperature during conidial production

Leaf disc bioassays for lesion development

3.2.4.1 Effect of media

3.4.1.1 Effect of media

3.4.1.2 Effect of light

Conidial germination experiments

3.4.2.1 Effect of media

Difco lima bean agar

3.4.2.2 Effect of temperature

Leaf disc bioassay for lesion development

3.4.3.1 Effect of media

Factors affecting disease development

4.2 Leaf disc bioassays

Materials and methods

4.2.1.1 Isolate virulence

4.2.2.3 Isolate virulence

4.2.2.6 Effect of post inoculation temperature

4.3 Glasshouse experiments

4.3.2 Effect of plant growth stage at constant water level on susceptibility to

Above ground biomass

Factors affecting disease development on the floating leaf stage

4.3.4.1 Conidial density

4.3.5.1 Conidial density

Green leaf area

Above ground biomass

Factors affecting disease development at the juvenile stage