Control of Mint Rust on Peppermint Epidemiology and Chemical Control

A report for the Rural Industries Research and Development Corporation by J. Edwards University of Melbourne

October 1999

RIRDC Publication No 99/122

RIRDC Project No UM-16A

ii

©1999 Rural Industries Research and Development Corporation. All rights reserved ISBN 0 642 579954 ISSN 1440-6845 Control Of Mint Rust of Peppermint - Epidemiology And Chemical Control Publication No. 99/122 Project No. UM-16A The views expressed and the conclusions reached in this publication are those of the author and not necessarily those of persons consulted. RIRDC shall not be responsible in any way whatsoever to any person who relies in whole or in part on the contents of 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 Jacqueline Edwards Institute of Land and Food Resources The University of Melbourne PARKVILLE VIC 3052 Phone: 03 9344 6458 Fax: 03 9349 4518 email: j.edwards@landfood.unimelb.edu.au RIRDC Contact Details Rural Industries Research and Development Corporation Level 1, AMA House 42 Macquarie Street BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: 02 6272 4539 Fax: 02 6272 5877 Email: rirdc@rirdc.gov.au Internet: http://www.rirdc.gov.au/ Published in October 1999 Printed on environmentally friendly paper by Canprint

iii

Foreword Essential oil production is an emerging new industry in the river valleys of north east Victoria. Commercial sales of peppermint oil from this region began in 1991, and have increased annually since that time. The long term viability of the industry, however, is dependent upon production of an oil of consistent quality and reliable yield. The disease mint rust, caused by the fungus Puccinia menthae, has caused major problems for the peppermint growers. If uncontrolled, mint rust reduces oil yield by 50% or more and also reduces oil quality. Control measures to date have been based on those effective in the USA, yet little is known about the behaviour of the disease under local conditions. This publication considers the epidemiology of the disease as it occurs in north east Victoria. It examines the environmental requirements of the uredinial stage of the fungus. An action threshold for timely initiation of chemical control is developed, and several fungicides are screened for their effectiveness. The use of flaming for mint rust control is investigated. Variation within the pathogen population is examined, and two distinct groups are revealed suggesting that the taxonomy of Puccinia menthae needs to be re-examined. This project is part of RIRDC’s Essential Oils and Plant Extracts Program which aims to obtain efficiency in production and to improve Australia’s presence in and share of world markets. Most of our diverse range of over 400 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/pub/cat/contents.html

Peter Core Managing Director Rural Industries Research and Development Corporation

iv

Acknowledgements

The work presented here is part of a PhD study undertaken at The University of Melbourne. I would like to thank my PhD supervisors Dr. Doug Parbery, Dr. Gerald Halloran and Dr. Peter Taylor for their valuable advice.

Many people contributed to the successful completion of this study. At the Ovens Research Station, Myrtleford, I would particularly like to thank the Director, Mike Morgan, for allowing me to collaborate with his staff, Fred Bienvenu, Helen Morgan and Brendan Ralph, and to use the facilities of the Research Station. In the Institute of Land and Food Resources at The University of Melbourne, Dr. Glyn Rimmington allowed me use of his computer, Dr. Paul Taylor, Rebecca Ford, Allison Croft, Dr. Peter Ades and Dr. Pauline Gere provided assistance with the genetic analysis of Puccinia menthae. Jocelyn Carpenter of the Botany School, The University of Melbourne, carried out the scanning electron microscopy. Dr. Robert Beresford, HortResearch, New Zealand, gave me permission to use his disease assessment key developed specifically for peppermint rust.

v

Contents

Foreword iii

Acknowledgements iv

Executive Summary vi

Introduction 1

Objectives 2

I. Epidemiology of Mint Rust on Peppermint 3 A. The Field Disease Cycle 3 B. The Uredinial Cycle 7 C. Studies of Mint Rust under Field Conditions 19

2. Effect of Mint Rust Infection on Peppermint Growth and Yield 23

3. Investigation of Methods for Control of Mint Rust 32 A. Development of an Action Threshold for Effective Chemical Control 32 B. The Use of Flaming in Peppermint Crops 37 C. The Effect of the Herbicide Paraquat on the Viability of Puccinia menthae Urediniospores 40 D. Evaluation of Fungicides to Control Mint Rust 43

4. Variation in the Pathogen, Puccinia menthae 51

Conclusions 61

Bibliography 66

vi

Executive Summary

Mint rust, caused by Puccinia menthae, is a disease limiting the development of a

successful essential oil industry in north east Victoria. This publication documents a study

of the disease on peppermint, Mentha x piperita, from 1994 to 1998, and methods for

control.

This study reports a four-year study of the disease cycle on peppermint growing at the

Ovens Research Station, Myrtleford, north east Victoria, and it showed that the full life

cycle of the rust does not occur on peppermint growing in this region. No spermogonia or

aecia were ever seen on peppermint, whereas urediniospores persisted all year round.

A study of the infection process of urediniospores on peppermint found that, under artificial

conditions, the minimum, optimum and maximum temperatures for infection were <5, 20

and 27°C respectively, and the minimum leaf wetness duration was six hours. This

explained the persistence of urediniospores over the winter months in this region. These

findings suggested that control measures would be best targeted during October -

November, when temperatures are still less than optimum for the fungus, and the crop

canopy is still open. Once the canopy is dense and closed, the fungus is protected from

dessication and high temperatures.

An investigation of the effect of rust infection on the growth and yield of peppermint

showed that mint rust significantly increased leaf loss, and significantly reduced leaf area,

leaf fresh weight, oil content, stem and root dry weights, and the numbers of stolons per

plant. These effects were seen to have implications for the long-term health of the crop, and

a plot of peppermint where mint rust was allowed to develop without control died out after

three and a half years.

Under field conditions, daily maximum temperatures above 35°C caused disease levels to

fall. A strong relationship was found to exist between disease incidence and disease

severity for mint rust, and it was proposed that this may be used to develop an action

threshold for the initiation of chemical control of the disease. The incidence - severity

relationship indicated that when the disease reached a mean disease incidence of 60% or

vii

less infected leaves per shoot, fungicide should be applied. This was borne out in the field,

when applications of the fungicide propiconazole when disease incidence was 60%

eradicated from the crop, but applications when disease incidence was 80% merely slowed

disease progress.

Autumn flaming of peppermint resulted in increased oil yields, but spring flaming did not.

Six currently available fungicides were tested for mint rust control on M. x gracilis cv.

Scotch spearmint over two years. Bitertanol and tebuconazole were identified as

significantly more efficient than the other fungicides in controlling the disease, which

resulted in higher oil yields.

Variation within the pathogen population was examined by comparing host ranges,

teliospore morphology and RAPD-PCR profiles. It became evident that there are two types

of Puccinia menthae found in Victoria which differ in morphology, host range and disease

cycles, and they appear to be genetically isolated from each other. One type infects

peppermint and related hosts, and the other type infects spearmint and its relatives. These

results raise doubts about the current taxonomic classification of this rust as a single

species.

1

Introduction

Peppermint is grown in north east Victoria for essential oil production. Research into

essential oil production began at the Ovens Research Station (ORS), Agriculture Victoria,

near Myrtleford, in 1976 when rootstock of the main varieties of peppermint and spearmint

were imported from the USA through the Australian Government Quarantine Station,

arriving at ORS twelve months later (Bienvenu 1993). The peppermint cultivar Todd’s

Mitcham was chosen as the one best-suited for the region and commercial plantings were

established in 1990. Two and a half tonnes of Victorian peppermint oil were sold for export

in 1991, increasing annually to 16 tonnes by 1996, at $43,000/tonne (Mr. Fred Bienvenu,

pers. comm. 1997).

The economic viability of the peppermint oil industry depends upon its capacity to expand

the area of plantings and maintain high average yields of top quality oil. A limiting factor

has been inadequate control of the disease mint rust (Peterson 1995). In the absence of mint

rust, growers in north east Victoria can achieve yields in excess of 90 kg/ha of top quality

oil, but unfortunately the disease has commonly reduced it to less than 40 kg/ha of an oil of

inferior quality (Bienvenu 1992). This is comparable with yield losses reported from the

USA and the USSR (Fletcher 1963). An example of the devastating effects that inadequate

control of the disease can have was given in New Zealand. The New Zealand peppermint

oil industry began in 1970 and was rust-free until 1978. After the first epidemic during the

season of 1978/79, Harvey (1979) predicted that unless effective control measures were

found, the emerging industry would not survive. Unfortunately his prediction was correct;

by the mid-1980s production ceased, mainly because of mint rust (Dr. Robert Beresford,

HortResearch NZ, pers. comm.).

Mint rust is caused by the fungus Puccinia menthae Pers., which attacks many members of

the herb family Lamiaceae, including peppermint. Control measures in the USA aim to

break the life cycle of the fungus before the destructive urediniospore stage is reached.

Non-chemical means, such as cultivation or the use of flame, have been used, but generally

the use of fungicides, the simplest procedure, is reasonably effective. Fungicides are also

used against infection that may develop during the growing season. Currently, Victorian

growers use recommendations adapted from those developed in the USA, but the reliance

2

on chemicals for control is restricted by its high cost, concern over contamination of the oil

and concern that resistance will develop to the chemicals. There is also growing awareness

internationally that Australia is striving for a “clean and green” agriculture for marketing its

produce. An understanding of how the rust fungus interacts with its mint host under local

conditions, and how such infection affects oil yield, is important to facilitate the efficient

use of a range of control measures.

Objectives

Until the present investigation was undertaken, no work on the biology and pathology of

mint rust had been done in Australia. The objectives of this study, therefore, were:

1. To determine and demonstrate an effective strategy for the control of peppermint rust

based on minimal use of chemical sprays;

2. To identify fungicides which will effectively control Puccinia menthae without

contamination of distilled oil;

3. In conjunction with (1) and (2) to conduct detailed studies of the epidemiology of mint

rust eg. life cycle and the influence of temperature and humidity on its capacity for growth

and reproduction.

3

1. Epidemiology of Mint Rust on Peppermint

A. The Field Disease Cycle

Introduction

Puccinia menthae is a rust fungus with a complicated life cycle, involving five spore stages

produced in sequence on its host (Figure 1). During winter, the fungus survives on debris as

tough-walled black teliospores. These teliospores germinate in spring, producing tiny

colourless basidiospores which infect young host tissue as it emerges through the soil

surface. Spermagonia result from this infection, and if cross-fertilisation occurs, yellow

aeciospores are produced. These become foci of disease in the field, infecting nearby mint

leaves, and resulting in the production of brown urediniospores, which are aerially

dispersed and can travel large distances. Urediniospores infect mint leaves through the

stomata and within two weeks, if conditions are favourable, thousands of new

urediniospores are released from uredinia (rust pustules) formed on the undersides of the

leaves. The urediniospore stage is the very destructive summer phase of the disease.

The sequence of these events and how they are influenced by the seasons of the year is

known as the disease cycle. The disease cycle as it occurs in the field under north-east

Victorian conditions was unknown when this project began, therefore a study was initiated

to monitor the spore stages as they developed on peppermint growing at the Ovens

Research Station near Myrtleford.

Methodology

One of the tobacco seedbed plots (1m x 15m) at the Ovens Research Station, Myrtleford,

was planted with peppermint cv. Todd’s Mitcham in September 1993. This seedbed plot of

peppermint was regularly watered and fertilised, but no measures were taken to control

mint rust. Sampling was done by cutting 20 mint stems at ground level along a longitudinal

transect of the plot. This sample size was determined to be adequate for statistical analysis

by comparison of the mean and standard deviation with sample size, as described by Kranz

(1988) and Neher and Campbell (1997).

4

Sampling was carried out weekly during the growing season (September to February) and

fortnightly from autumn through to spring (March to August) from November 1993 to

December 1997.

Urediniosporecycle repeatedmany times

Autumn-winterUredinia convert to teliaproducing dark-brown teliospores

Late winter - early springTeliospores germinate to produccolourless basidiospores whichinfect the emerging mint shoots

SpringRed 'blisters' containing spermogoniaappear on leaves and petioles

SpringRed 'blisters' on the stemsbreak open releasing large numbersof yellow aeciospores

SpringAeciospores infect mint leavesand produce brown urediniospores

Summer

5

Figure 1 Diagram of the disease cycle of Puccinia menthae as reported for mint-growing

areas of the USA.

The peppermint samples were examined for the presence of P. menthae using a Wild Leitz

dissecting microscope (x16 - x50). The spore stages observed were recorded, and the

percentage of shoots bearing each spore stage was calculated.

Results

Urediniospores were always present, although less abundant over the winter months (June

to August). Some teliospores were produced in autumn and winter, but no spermogonia or

aecia were ever observed (Figure 2). The exact timing of teliospore production varied from

year to year, but was between April and August.

0255075

100

% u

redi

nia

1994 1996 19971995

010203040

% a

ecia

020406080

% sp

erm

agon

ia

0255075

100

% te

lia

Telia

Spermagonia

Aecia

Uredinia

N D M J J AJ F M A S O N D M J J AJ F M A S O N D M J J AJ F M A S O N DM J J AJ F M A S O N D

N D M J J AJ F M A S O N D M J J AJ F M A S O N D M J J AJ F M A S O N DM J J AJ F M A S O N D

N D M J J AJ F M A S O N D M J J AJ F M A S O N D M J J AJ F M A S O N DM J J AJ F M A S O N D

N D M J J AJ F M A S O N D M J J AJ F M A S O N D M J J AJ F M A S O N DM J J AJ F M A S O N D

Figure 2 Disease cycle of Puccinia menthae on peppermint growing in the Ovens Valley,

north east Victoria.

Urediniospores were collected fortnightly during the winter months of 1994 to 1997, plated

onto water agar and incubated for 24 hours at 20°C to test for viability. The percentage of

germinating urediniospores was never less than 30%, indicating that viable urediniospores

6

are present all year round and that they, rather than teliospores, carry the disease from one

growing season to the next.

Discussion

Prior to this study, the disease cycle of Puccinia menthae in the field in north-east Victoria

was unknown, but was assumed to be similar to that reported in the USA. As a

consequence, control measures were being recommended to local growers that aimed to

eliminate the aecial stage in spring. The results presented here show that the full life cycle

does not occur on peppermint grown under local conditions. This is not an unexpected

phenomenon, as the cereal rust Puccinia graminis also persists as urediniospores all year

round in Australia, despite being macrocyclic in the cold climate conditions of the northern

hemisphere wheat-growing regions (Brown 1997).

At the same time, however, spermogonia and aecia were observed every spring on both

commercial spearmint species (Mentha spicata cv. Native and M. x gracilis cv. Scotch

speamint) also growing at the Ovens Research Station (data not shown). In the case of the

spearmints, therefore, the teliospores are the main source of overwintering inoculum as

reported in other mint growing areas, but on peppermint the urediniospores carry the

disease through the winter.

7

B. The Uredinial Cycle

Introduction

Because viable urediniospores were present throughout the year, a series of controlled

environment experiments was conducted to investigate environmental influences on the

uredinial cycle. Each experiment was conducted at least twice, unless otherwise stated.

Methodology

i. Effect of temperature on urediniospore germination

Fresh urediniospores were collected by tapping infected leaves over petri dishes containing

2% water agar, and were spread evenly over the surface of the agar using a sterile bent

glass rod. The petri dishes were placed in darkness at temperatures of 5, 10, 15, 20, 25, 30

and 35°C, and left for 24 hours to ensure maximum germination. The petri dishes were then

examined to determine urediniospore germination. There were six replicates (petri dishes)

per treatment (temperature) in each of which 100 urediniospores were inspected.

ii. Effect of temperature on the rate of urediniospore germination

Urediniospores were collected, plated on water agar and incubated as described above. In

this case, germination was determined hourly for the first six hours of incubation to follow

the kinetics of the germination process. There were three replicates for each combination of

incubation time and temperature and, as previously, 100 urediniospores were examined per

replicate.

iii. Effects of leaf wetness duration and temperature on germination and

penetration processes

Three-week-old peppermint cuttings were inoculated by spraying the leaves to run-off with

a spore suspension containing 104 urediniospores / ml. They were then randomly divided

into five groups of 24 cuttings, and each group placed into a plastic box with a tight-fitting

lid. Immediately after sealing, each box was placed in the dark at one of five different

temperatures (5, 10, 15, 20 and 27°C). Four plants were randomly removed from each box

after 0, 3, 6, 9, 12 and 24 hours respectively, but kept at the same temperature in the dark

for 24 hours after inoculation. All moisture had evaporated from the leaves within 10

8

minutes of removal from the boxes. There were four replicates (plants) for each

combination of leaf wetness duration and temperature. After the 24 hour period, all plants

were returned to the glasshouse.

Disease severity (a visual estimate of the percent leaf area per plant covered by lesions)

was assessed 17 days after inoculation using a peppermint rust key developed by Beresford

and Mulholland (1987). Each leaf was assessed and the mean of the leaf values was taken

as the disease severity value for the plant. Absolute disease severity values differed

between repeats of the experiment, so were normalised by taking the highest value in each

repeat as 1.00 and expressing the other values as proportions of 1.00. These normalised

values did not allow valid statistical analysis as they are not independent of each other.

iv. The infection process under optimum conditions

A 12-week-old plant, growing in a 10 cm pot in the glasshouse, was inoculated and kept in

a sealed plastic box for 24 hours at 20°C. It was assumed that these conditions would

produce the highest levels of infection. At intervals (1, 2, 4, 6, 12 and 24 hours) after

inoculation, a young fully-expanded leaf was detached and immediately submerged in

Chlorazole Black E clearing and staining solution. After seven days, the leaves were

washed with deionised water and transferred to a saturated solution of chloral hydrate for a

further two days. This procedure stains the fungal structures grey-black and clears the leaf

tissue (Keane et al. 1988). Leaves were then mounted in lactic acid and examined using a

Dialux 20 microscope (x100, x400 and x1000). At least 50 spores (where available) were

examined per leaf, and the percentage of spores producing germ tubes, appressoria,

substomatal vesicles, infection hyphae or haustoria was calculated for each time interval.

This experiment was not repeated as considerable numbers of spores were examined for

each time period.

v. Effect of temperature on the latent period of infection

Peppermint cuttings were inoculated and kept at 20°C for 24 hours to ensure maximum

penetration. The plants were then placed into six controlled environment cabinets, five at

constant temperatures of 5, 10, 15, 20, 27°C and one at 24°C day/20°C night. There were

six replicates for each of the six temperature treatments. Plants were inspected daily and the

9

number of days taken to reach flecking (the first visible symptoms of disease) and the first

opening of uredinia were recorded. Daily counts of the number of open uredinia per leaf

were made until all uredinia had opened.

The definition of the ‘latent period’ of rust infection cycles varies between research

workers (Teng and Close 1978). In this experiment, the latent period is defined as the time

from inoculation to when 50% of uredinia had opened. This is in accord with previous

latent period studies on P. menthae (Beresford and Mulholland 1987, Johnson 1995).

Disease severity was assessed 23 days after inoculation when all uredinia had opened, with

the exception of the 5°C treatment which was recorded as zero.

vi. Effect of temperature on urediniospore production

Plants were inoculated and maintained in the glasshouse until the start of uredinium

opening. Leaves were then detached and single leaves were placed, abaxial surface

upwards, in petri dishes containing 2% water agar, amended with 0.5 gL-1 benzimidazole,

to postpone senescence of the leaf tissue (Chandrashekar 1982, Shtienberg and Vintal

1995). The petri dishes were sealed and incubated at six different temperatures (5, 10, 15,

20, 25 and 27°C). There were three replicates (leaves) per treatment (temperature). After

five days, the number of uredinia per leaf was recorded and each leaf was carefully

transferred to a glass test tube. Two mls of sterile deionised water containing a surfactant

(1% Tween 20) was added to each leaf. The tubes were shaken for 24 hours in darkness at

4°C to dislodge the mature spores, following which the leaves were discarded. The number

of spores in each tube was counted using a haemocytometer and the mean number of spores

produced per uredinium per day was calculated. This experiment was repeated twice. The

second repeat, however, was slightly modified, with six replicates (leaves) per treatment

(temperature), and incubation for seven instead of five days. In this case, the benzimidazole

was omitted from the medium with no deleterious effect, as it had not been effective in

postponing leaf senescence in the earlier repeats.

vii. Data analysis

Appropriate curves were fitted to the data sets using the curve-fitting options of both

software packages, Sigma Plot and Cricket Graph. Differences between treatments were

10

assessed, where applicable, using analysis of variance within the statistical software

package Minitab.

Results

i. Effect of temperature on urediniospore germination

The effect of temperature on urediniospore germination was best described by a 3rd-order

polynomial regression (Figure 3). The optimum temperature was 20°C and the maximum

30°C. Maximum germination occurred at 20°C and there was little difference between 10,

15 and 25°C, but was low at 5°C. Only three of 600 spores germinated at 30°C, and there

was no germination at 35°C. After completion of the germination counts, plates from 5, 30

and 35°C were incubated for a further two hours at 20°C and then re-examined. Most of the

ungerminated spores from the 5°C treatment germinated during this period, but those from

30 and 35°C did not.

0

20

40

60

80

Ger

min

atio

n (%

)

0 5 10 15 20 25 30 35

Temperature (°C)

y = 7.44 + 3.1x+ 0.28x2-0.01x3 R2 = 0.97

11

Figure 3 Effect of temperature on P. menthae urediniospore germination after 24 hours

incubation on water agar.

12

ii. Effect of temperature on the rate of urediniospore germination

Most germ-tube emergence occurred during the first three hours of incubation, reaching a

maximum within four hours (Figure 4). The rate of urediniospore germination was highest

at 15 and 20°C, reaching a maximum of 63%, while at 10 and 25°C the maximum was

45%. The germination pattern was the same at 10 and 25°C, and similar at 15 and 20°C.

Maximum germination at 5°C (11%), however, was reached after two hours incubation. At

30°C, only one spore out of 1,800 germinated (0.06%) and this was noted after one hour of

incubation. No spores germinated at 35°C. Once again, plates from the 5, 30 and 35°C

treatments were removed to 20°C and re-examined after two hours, and again most

ungerminated spores from the 5°C treatment germinated during this period, but those from

30 and 35°C did not.

Figure 4 Effect of temperature on the rate of P. menthae urediniospore germination.

13

iii. Effects of leaf wetness duration and temperature on germination and

penetration processes

No infection occurred under three hours of leaf wetness, regardless of temperature. Trace

infection (1 - 2 uredinia) on some individuals at each temperature was noticed after six

hours of leaf wetness. After nine hours, the wetness effect could be ranked

10>15>20>5>27°C.

At all temperatures except 27°C, disease severity increased as the duration of leaf wetness

increased (Figure 5). At 27°C, maximum severity was reached following nine hours of leaf

wetness. The optimum temperature range was 10 - 20°C, the minimum temperature was

less than 5°C and the maximum was 27°C.

Figure 5 Effects of leaf wetness duration and temperature on disease severity (data have

been normalised) of P. menthae on peppermint.

14

iv. The infection process under optimum conditions

Urediniospores began to germinate on the leaf surface within two hours, but appressoria

and sub-stomatal vesicles were not observed until six hours after inoculation (Table 1). The

percentage of spores producing appressoria and sub-stomatal vesicles continued to increase

with time until observations ceased at 24 hours. At this time infection hyphae and haustoria

inside the leaf tissue were observed. The mean germ-tube length continued to increase over

the 24 hour period.

Table 1 Percentage of Puccinia menthae urediniospores that had germinated and produced

infection structures (appressoria, substomatal vesicles, infection hyphae and

haustoria) on peppermint leaves at time intervals after inoculation.

Infection Time after inoculation (h)

structure 1 2 4 6 12 24

Ungerminated urediniospores

100 73 14 14 15 5

Germ tube - 27 86 86 85 95

Appressorium - - - 3.6 7.0 20.0

Substomatal vesicle

- - - 1.2 2.4 17.4

Infection hypha - - - - - 14

Haustorium - - - - - 6

Mean germ tube length (µm)

- 29.0 56.6 71.1 192.6 254.0

The study of the infection process was undertaken to gain a better understanding of the

host-parasite relationship and to help place other observations into context. Urediniospores

began germinating within two hours of inoculating the leaf surface, and no more

germination occurred after four hours, corresponding to observations of urediniospore

germination on water agar. Appressoria and substomatal vesicles began to form within six

15

hours, indicating that the fungus was penetrating inside the leaf and was no longer

vulnerable to the external environment, and six hours was shown to be the minimum period

of leaf wetness allowing infection to occur. The steady increase in disease severity

observed as the duration of leaf wetness increased can be attributed to the continued

elongation of the germ tubes over 24 hours, which increases the chances of the fungus

finding an open stoma through which to enter. The exception to this was 27°C, when the

percentage of leaf area affected was negligible and did not increase significantly with an

increased duration of leaf wetness. It is evident that germ-tubes exposed to extended

periods of high temperature did not survive, whereas mycelium within the leaf survived a

little better.

v. Effect of temperature on the latent period of infection

Temperature had a significant effect (P<0.001) on latent period which decreased as

temperature increased to 20°C. This relationship can be described by the hyperbolic

equation LP = 253.291temp-1.056 (Figure 6).

0

20

40

60

Late

nt p

erio

d (d

ays)

0 5 10 15 20 25

Mean temperature (°C)

y = 253.291x-1.056 R2 = 0.98

Figure 6 Effect of mean temperature on the latent period of P. menthae on peppermint.

16

No difference was evident between the constant 20°C and the alternating 24/20°C (mean

22°C) treatments. No infection developed at 27°C. The response curve between severity

and temperature was a 3rd-order polynomial, with a maximum at 20°C and no disease at

27°C (Figure 7). Although no infection was evident at the time severity assessments were

made, plants growing at 5°C developed sori after a latent period of 50 days, at which point

many teliospores were present amongst the urediniospores. Teliospores were also observed

in sori 42 days after inoculation on the plants growing at 10°C.

0

0.2

0.4

0.6

0.8

1

Dis

ease

sev

erity

(nor

mal

ised

)

0 5 10 15 20 25 30

Mean temperature (°C)

y = 0.160 - 0.091x + 0.015x2 - 0.000x3

R2 = 0.96

Figure 7 Effect of post-penetration temperature on the disease severity of P. menthae on

peppermint after 23 days. Data have been normalised.

17

0

200

400

600

800

1000

1200

Ure

dini

ospo

res /

soru

s / d

ay

0 10 20 30

Temperature (°C)

y = 8.771 + 12.691x + 7.815x2 - 0.295x3

R2 = 0.92

Figure 8 Effect of temperature on the daily production of P. menthae urediniospores per

uredinium.

vi. Effect of temperature on urediniospore production

Urediniospore production data were transformed for normality using a log10 transformation

prior to analysis. Temperature had a significant effect (P<0.001) on urediniospore

production, with the response curve again being 3rd-order polynomial (Figure 8). Daily

urediniospore production increased by 353% as the temperature increased from 5 to 20°C,

and was inhibited significantly at 27°C.

Discussion

In summary, the optimum temperature for the development and function of the uredinial

stage of P. menthae is 20°C, with a near-optimum range of 10 - 25°C and a maximum

temperature around 27°C. The lower temperature threshold, however, is below 5°C, which

explains the presence of viable urediniospores on peppermint growing during winter in

north-east Victoria, when daily average temperatures are between 5 - 10°C. It can be seen

from these results that low temperatures merely slow down infection and sporulation

processes, allowing the fungus to persist until conditions become favourable.

This data point was assumed and added to the analysis

18

In order to produce high-quality peppermint oil, a photoperiod response (long days and

short nights) is required to induce flowering and a diurnal temperature fluctuation in the

month preceding harvest is essential, restricting production to climates characterised by

clear hot days and clear cool nights in the summer (Bienvenu 1993). The river valleys of

north-east Victoria are particularly suitable for essential oil production, with plenty of

irrigation water available during summer. These conditions, however, also favour the

development of rust epidemics. The maximum daily temperature in summer is often above

30°C, but it is usually accompanied by overnight minima of 8 - 14°C. Overhead irrigation

of the crop late in the day or at night provides conditions that are conducive to infection,

and there will most certainly be areas within the dense peppermint canopy that remain

shaded, relatively cool and moist even during the hottest part of the day. It would appear

from this study, therefore, that control measures would be best applied during spring

growth when temperatures are between 5 - 15°C, before the canopy has closed over and

conditions have become ideal for rapid generation of the fungus.

19

C. Studies of Mint Rust under Field Conditions

Introduction

In order to develop and implement a successful plant disease management strategy, it is

necessary to understand as much as possible about the factors that contribute to the

development of the plant disease epidemics. A common method for quantifying the

progressive changes in disease levels in the field is to plot the amount of disease present at

consecutive time intervals. This then serves as a summary of the disease present in a

particular crop during a particular season, and can be compared with other plots of the same

disease in other regions or seasons, or with concurrent variation in temperature and rainfall,

in order to identify the effects of climatic variables, fungicide treatments, etc. on the course

of the epidemic.

The amount of disease present is usually assessed by either or both of two methods:

(1) measurement of the incidence of disease and (2) measurement of the severity of disease.

Disease incidence is calculated as the percentage of the total leaves on a stem with one or

more uredinia. It is quick and simple, while being objective with little opportunity for user-

error. The drawback, however, is that once all leaves are infected, no further increase in

disease can be measured, limiting its value for investigating crop losses or disease

management strategies.

Disease severity is a measure of the leaf area loss due to the disease. Estimations of how

much leaf area is affected by mint rust are made by comparing leaves with a visual disease

assessment key. Disease severity in this study was calculated as the mean percentage loss

of leaf area per mint stem, using a visual assessment key adapted for peppermint rust by

Beresford and Mulholland (1987). Unless otherwise stated, one leaf was assessed from

each third of the main stem and the mean of the three measurements was taken as the

disease severity value of the stem. Missing leaves were ignored.

As far as can be determined, there is no such published information for disease epidemics

of mint rust. The following experiment was undertaken to study the influence of climate on

the progress of the disease on peppermint growing in north east Victoria.

20

Methodology

Samples were collected weekly from the peppermint seedbed plot, Ovens Research Station,

during September to February, 1994 to 1998. Twenty shoots were taken at each sampling

time and assessed for disease incidence and disease severity as previously described, and

the mean values were calculated for each sample. Disease progress was monitored by

plotting the means of the disease measurements (incidence and severity) against time.

Meteorological data were collected using the Hardi Metpole installed at the Ovens

Research Station. The minimum and maximum daily temperatures were measured at 20 cm.

above ground level, which was within the crop canopy. Minimum and maximum daily

relative humidity was measured at the same place as temperature, and total daily rainfall,

total daily solar radiation and maximum daily wind speed were also recorded.

Results and Discussion

Disease incidence:

Disease incidence began at less than 10% in all seasons (Figure 9A). During 1994/5 and

1995/6, disease incidence increased at a linear rate from early October to 80% by the end of

December, but in 1996, disease incidence only reached 55% when the combination of hot

summer weather and three years of continual rust infection severely checked the growth of

the plants and they did not recover. The following season, 1997/8, there were very few

shoots of peppermint to be found and no rust was observed. Weekly sampling had to be

abandoned due to lack of shoots.

Disease severity:

Disease severity was negligible until the beginning of November in each season (Figure

9B). In 1994/5, it increased rapidly to 15% by the end of November 1994, then fluctuated

between 5 and 10% during December 1994, increasing to 15% again by February 1995. In

1995/6, disease severity increased in steps, reaching a plateau of 5% in December 1995. In

1996/7, disease severity was again 5% by December 1996, but then the rust disappeared as

the peppermint died off, as explained when discussing disease incidence, and no rust was

observed on the meagre growth of peppermint in 1997/8.

21

0

5

10

15

20

Dis

ease

seve

rity

(% le

af a

rea

affe

cted

)

Sept Oct Nov Dec Jan Feb

97/896/795/694/5

0

20

40

60

80

100D

isea

se in

cide

nce

(% le

aves

infe

cted

) A

B

Figure 9 Changes in disease levels for peppermint growing in the seedbed plot - A disease

incidence, B disease severity.

The influence of climatic variables on disease

There were no obvious correlations between relative humidity, solar radiation, rainfall and

wind speed and disease. Temperature, however, had a direct effect on the levels of disease,

particularly disease incidence (Figure 10). The peppermint plants died out during the

summer of 1996, so there is no 1997/8 graph. On 5th September 1995, a severe frost (-5°C)

caused high levels of defoliation in all the crops and reduced the amount of disease present

at the beginning of the season. In late spring and summer, days with maximum

temperatures above 35°C caused an immediate reduction in the amount of rust present, but

22

the levels rose again if the maximum temperature dropped below 35°C for a few days eg

late January 1994/5.

(10)0

102030405060708090

100

Tem

pera

ture

(°C

)0

20

40

60

80

100

Dis

ease

inci

denc

e(%

leav

es in

fect

ed) 1995/6

0102030405060708090

100

Tem

pera

ture

(°C

)

0

20

40

60

80

100

Dis

ease

inci

denc

e(%

leav

es in

fect

ed) 1996/7

(10)0

102030405060708090

100

Tem

pera

ture

(°C

)

0

20

40

60

80

100

Dis

ease

inci

denc

e(%

leav

es in

fect

ed) 1994/5

min T

max T

Oct Nov Dec Jan Feb

Sept Oct Nov Dec Jan Febmin T

max T

max T

min T

Sept Oct Nov Dec Jan Feb

Sept

Figure 10 Effect of temperature on the incidence of mint rust on peppermint in the seedbed

plot, Ovens Research Station.

23

2. Effect Of Mint Rust Infection On Peppermint Growth And Yield

Introduction

It is well established that mint rust reduces the quantity of essential oil produced by

commercially-grown peppermint, but there is little information on its impact on the growth

of the plant itself. The objective of the following experiments, therefore, was to quantify

the effects of P. menthae on the growth and yield of peppermint.

Methodology

On 1st September 1995, 84 rooted tip cuttings of Todd’s Mitcham peppermint were

transplanted singly into 10 cm pots containing potting mix. The plants were kept in an

outdoor area of the System Garden at The University of Melbourne. On 22nd September

1995, they were divided randomly into two groups, each of 42 plants. One group was

inoculated with 5 x 103 urediniospores/ml. The plants were returned to the outdoor area for

the duration of the experiment and maintained under natural temperature and photoperiod

with supplementary daily watering. They were fertilised monthly with the complete slow-

release fertiliser, Osmocote. The control group was sprayed with propiconazole (as Tilt

250EC) at a rate of 0.4 ml/L as soon as rust was detected (spray dates: 21-11-95, 27-11-95,

21-12-95 and 4-1-96).

The experiment was repeated in 1997/8, with slight modifications. On 12th September

1997, 40 rooted tip cuttings of Todd’s Mitcham peppermint were transplanted singly into

10 cm pots as previously, and on 8th October 1997 the resultant plants were divided

randomly into two groups of 20 plants each. One of the groups was inoculated as described

above, but with a higher inoculum dose of 105 spores/ml. In 1995/6, the pots were watered

daily but unavoidably dried out between watering during periods of hot weather. In 1997/8,

the pots were placed in plastic trays to retain the run-off water, ensuring that the plants

remained free from any water stress. The control group was again kept ‘rust-free’ with

applications of propiconazole as described above (spray dates: 13-10-97, 21-11-97, 23-12-

97 and 30-12-97).

The first experiment was harvested during April 1996, and the second experiment was

harvested during March 1998. At harvest, the following measurements were made per

24

plant: oil content, shoot number, stolon number, number of leaf nodes per shoot and

number of missing leaf pairs per shoot, total leaf area, leaf fresh weight, stem fresh weight

and root fresh weight. Stem and root samples were oven-dried at 80°C for one week, then

reweighed to obtain dry weights. Leaf samples were frozen immediately after weighing

and stored at -20°C until the harvest was complete. The frozen leaf samples were then

steam-distilled, the oil content measured and the oil analysed. The protocol for oil

extraction prevented measurements of leaf dry weight. Total leaf area was measured using

a LI-3100 Area Meter (LICOR, Lincoln, Nebraska, USA).

Disease incidence was measured as the percentage of leaves that were infected per plant,

disease severity was visually assessed as previously described, and defoliation was

measured as the percentage of leaf nodes per stem that had no leaves.

Oil content was determined by steam-distillation of all leaves from each plant. The

extracted oil was bulked into four samples, two control samples and two rusted samples,

and sent to the Ovens Research Station for oil quality analysis. A Shimadzu Gas

Chromatograph 9A fitted with a 30 mm x 0.25 mm free fatty acid phase column was used

to measure the percentage of the five major oil components (menthone, menthol, menthyl

acetate, menthofuran and isomenthone) in each oil sample, according to the standard

procedure used by peppermint oil producers in the region. The bulking of the samples was

necessary as the apparatus could not analyse quantities of oil less than 1ml.

The data were analysed using the statistical software package Minitab. Where the

assumptions for normality were met, the two-sample t-test was applied to test for

significant differences between the two treatments at the 5% level. In the cases of stem and

root dry weights (1995/6) and oil content, stem dry weight, leaf fresh weight and leaf area

(1997/8), the data were transformed using a log10(x+1) transformation before analysis by

two-sample t-tests. In the case of stolon number (1995/6), no transformation could be found

that would overcome the problem of unequal variances, so differences between the control

and rusted treatments were assessed using the non-parametric Mann-Whitney U-test, which

allows for unequal variances between the two groups, and does not assume normality.

25

Regression analysis was used to examine relationships between the disease measurements

and oil yield of the plants. Curves were fitted using the graphical software package Cricket

Graph.

Table 2 Effects of rust infection on growth and yield components of peppermint cv. ‘Todd’s Mitcham’. Means with 95%

confidence intervals are presented, with P-values calculated from the two sample t-test, except a which are medians, with

P-values calculated from the Mann-Whitney U-test.

1995/6 1997/8

Control Rusted P-value

Change due to rust

Control Rusted P-value Change due to rust

Oil/plant (µl) 64.4 ± 7.5 36.1 ± 4.1 <0.01 −44% 131.0 ± 15.7 80.1 ± 8.4 <0.01 −39%

Stem number/plant 53.9 ± 2.5 30.0 ± 3.1 <0.01 −44% 54.9 ± 9.2 51.6 ± 5.6 0.53 n.s

Stolon number/plant 5.0 a 0.0 a <0.01 a −100% 24.7 ± 3.8 8.7 ± 1.9 <0.01 −65%

Total nodes/stem 14.9 ± 0.5 11.1 ± 0.6 <0.01 −26% 15.2 ± 1.2 15.5 ± 1.1 0.65 n.s

Leaf fresh wt. (g) 17.43 ± 1.27 11.19 ± 1.27 <0.01 −36% 46.64 ± 5.98 21.03 ± 3.01 <0.01 −55%

Stem dry wt. (g) 8.46 ± 0.96 3.63 ± 0.43 <0.01 −57% 47.75 ± 7.00 18.28 ± 2.45 <0.01 −62%

Root dry wt. (g) 6.08 ± 0.92 2.41 ± 0.51 <0.01 −61% 21.08 ± 3.32 10.58 ± 2.45 <0.01 −50%

Leaf area (cm2) 826.3 ± 56.8 620.6 ± 69.4 <0.01 −25% 1922.1 ± 229.9 1031.8 ± 141.7 <0.01 −46%

Defoliation (%) 50.0 ± 1.9 45.5 ± 3.7 0.043 −9% 37.3 ± 3.8 62.4 ± 3.5 <0.01 +67%

Disease incidence (%) 13.2 ± 6.4 63.5 ± 8.7 <0.01 55.2 ± 9.1 89.4 ± 4.2 <0.01

Disease severity (%) 0.8 ± 0.4 7.8 ± 1.7 <0.01 3.6 ± 1.0 22.6 ± 2.4 <0.01

27

Results

A small amount of rust was observed in the control group by the end of harvest, but

because the plants were free of rust for most of the experiment, such infection was accepted

as insufficient to invalidate the comparisons. Disease severity and incidence were low in

the control compared to the rusted group.

There were significant differences (P<0.05) between the rusted and control groups in each

of the variables measured (Table 2). Mint rust significantly reduced leaf nodes per shoot,

shoot number, stolon number, total leaf area, leaf, stem and root weights, and oil content

per plant. During a period of very hot weather in January 1996 many older shoots on the

rusted plants wilted irreversibly and died, and this leaf loss is not accounted for in the

defoliation measurements. Therefore it appears as if defoliation was higher in the control

plants than the rusted plants in 1995/6, but this is an aberration of the data. There was no

difference between treatments in oil content per leaf fresh weight (data not shown).

Table 3 Effect of rust infection on the relative proportions of the major oil components

(measured by gas chromatography). Means for each treatment are presented.

Oil constituents Experiment 1 (1995/6) Experiment 2 (1997/8)

(% total oil) Control Rusted Control Rusted Typicala

Menthone 13.7 15.8 6.3 11.3 18-22

Menthol 39.1 46.6 54.9 53.6 43-46

Menthyl acetate 16.3 9.4 7.8 5.8 3-6

Menthofuran 11.8 9.9 8.3 11.7 2-4

Isomenthone 1.7 2.0 1.4 1.9 2-3 atypical: the range expected for oil of this type, grown and harvested under optimum conditions.

Results of the gas chromatography analysis of the extracted oil are shown in Table 3. None

of them conformed to the range typical of oil produced under optimum conditions at

Myrtleford, but the oil composition from the rusted plants indicated that these plants had

not reached the same level of maturity as the control plants.

28

0

0.05

0.1

0.15

0.2

0.25

Oil

yiel

d (m

l)

0.000 0.500 1.000 1.500 2.000

Disease incidence (transformed)

B y = 0.201 - 0.088xR2 = 0.42, p < 0.01

A y = 0.063 - 0.024xR2 = 0.29, p < 0.01

B

A

0

0.05

0.1

0.15

0.2

0.25

Oil

yiel

d (m

l)

0.000 0.100 0.200 0.300 0.400 0.500 0.600

Disease severity (transformed)

A y = 0.061 - 0.077x R2 = 0.26, p < 0.01

B y = 0.162 - 0.158xR2 = 0.50, p < 0.01

B

A

Figure 11 Relationships between oil yield and measures of disease incidence and severity.

Original disease values were transformed using an arsine transformation. Open

circles (o) and equations A refer to 1995/6, and closed circles (•) and equations

B refer to 1997/8.

29

The disease incidence and severity levels were plotted against oil yields from both

experiments and although the actual levels of disease were higher in 1997/8, the same trend

of decreasing oil yield with increasing disease levels was evident in each (Figure 11). The

disease data for each experiment were normalised using an arcsine transformation and

linear regression was applied. The equations were significant (P<0.05), but the coefficients

of determination (R2) were low in 1995/6.

0.00

0.05

0.10

0.15

0.20

0.25

Oil

yiel

d (m

l)

0 20 40 60 80 100

Defoliation (%)

y = 0.203 - 0.002xR2 = 0.60, p < 0.01

Figure 12 Relationship between oil yield per plant and amount of defoliation (1997/8 data

only).

Defoliation measurements in 1995/6 were confounded by adverse hot weather and

associated moisture stress. Consequently, only defoliation data from 1997/8 were examined

for a relationship with oil yield. The data did not need to be transformed prior to analysis. It

was assumed that if defoliation was 100%, oil yield would be zero, and this point was

added to the analysis. There was a strong negative linear relationship (R2 = 0.60) between

oil yield and the amount of defoliation (Figure 12). Defoliation exhibited a positive linear

relationship (R2 = 0.55) with disease incidence (Figure 13), and positive logarithmic

relationship (R2 = 0.69) with disease severity (Figure 14).

this data point assumed and

added to the analysis

30

0

10

20

30

40

50

60

70

80

Def

olia

tion

(%)

10 20 30 40 50 60 70 80 90 100

Disease incidence (%)

y = 15.166 + 0.479x R2 = 0.55

Figure 13 Relationship between amount of defoliation and disease incidence per plant

(1997/8 data only).

0

10

20

30

40

50

60

70

80

Def

olia

tion

(%)

0 5 10 15 20 25 30 35

Disease severity (%)

y = 26.868LOG(x) + 24.846 r2 = 0.692

Figure 14 Relationship between amount of defoliation and disease severity per plant

(1997/8 data only).

31

Discussion

Excessive leaf drop has been observed in the field in north east Victoria during years of severe rust infection. In the present study, the comparison of defoliation in healthy and rusted peppermint was confounded in 1995/6 by the death of a high proportion of shoots in the rusted plants caused by water stress during hot weather. In 1997/8, however, with unlimited water availability, leaf drop was increased by 67% in the plants infected with mint rust compared to the control. Since the essential oil of peppermint is produced in glands on the surfaces of the leaves, the increased defoliation explains the significant reduction in oil yield observed, as the number of shoots per plant and the number of leaf nodes per shoot were not affected by mint rust infection. The negative correlation between oil yield and level of defoliation was adequately described by a simple linear equation, and the effect of rust infection on the level of defoliation could also be quantified. Leaves were shed rapidly as disease severity increased from zero to approximately 5% leaf area affected, reaching 40% shoot defoliation, but further increases in disease severity had less effect on the rate of defoliation. These results indicate that even a small amount of rust infection will cause peppermint cv. Todd’s Mitcham to shed its leaves, emphasising the importance of effective control of the disease.

Infection courts of biotrophic parasites, such as rust fungi, are known to act as sinks for assimilate in competition with the infected host plant, resulting in reduced plant growth rate, size and weight compared to healthy individuals (Parbery 1996). In the present study, mint rust caused reductions of 50-60% in stem and root dry weights, and stolon production was severely affected. The reduction in growth accompanying such infection, particularly that of roots and storage organs such as rhizomes and stolons, is likely to have profound effects on the ability of a perennial crop such as peppermint to yield well and persist over several years. This was demonstrated in the peppermint seedbed plot where mint rust was allowed to develop and progress unhindered by any control measures (see previous section). Within four years, the peppermint plants had died out. Peppermint crops in north east Victoria are currently replanted every three to five years, but growers see economic advantage in extending this period to possibly ten years. The implications from the results of the present study are that unless adequate rust control can be developed, the likelihood of longer cropping periods seems remote.

32

3. Investigation Of Methods For Control Of Mint Rust

A. Development of an Action Threshold for Effective Chemical Control

Introduction

At the present time, no chemical is registered for control of mint rust in Australia. Local

growers, however, currently use the fungicide propiconazole (as Tilt 250EC, manufactured

by Novartis) during the growing season to control rust on peppermint, as it is known to be

effective against other rust fungi in many agricultural and horticultural crops. Growers wait

until the disease can be readily observed, usually mid to late November, and then apply two

to three spray treatments of the fungicide at fortnightly intervals.

Effective integrated disease management depends upon a system of crop monitoring and

recognition of a critical level of disease as an ‘action threshold’ for the initiation of

fungicide treatment, resulting in minimal yet effective use of chemicals for control. In the

past, such action thresholds have relied upon recognition of a particularly critical level of

disease severity (Brown and Holmes 1983, Zadoks 1985, Royle 1994, Shtienberg 1995,

Brown and Keane 1997), but visual estimates of severity vary considerably between

assessors (Zadoks 1985, Nutter and Schultz 1995, Parker et al. 1995, Brown and Keane

1997, Nutter 1997). If it is possible to replace the assessment of disease severity with the

more objective assessment of disease incidence, the use of an action threshold is more

likely to be adopted (Zadoks 1985).

To do this, it is necessary to first establish whether there is a quantifiable relationship

between disease incidence and disease severity for the disease in question. Therefore it was

decided to investigate whether an incidence - severity relationship exists for mint rust, and

if so, whether it might be exploited to develop an action threshold for peppermint growers

to use as a decision tool for the initiation of fungicide applications.

33

Methodology

Part 1

The disease incidence and severity measurements from both the field experiments (p.17)

and the pot experiments (p.21) were plotted against each other using the graphical software

package Cricket Graph, and any relationship between them was examined both visually and

by using the coefficient of determination, as proposed by Fowler and Cohen (1990).

Part 2

Samples of peppermint were collected weekly as previously described. The samples were

taken from:

(i) a farmer’s paddock (O’Sullivan) near Myrtleford, Ovens Valley, September 1994 to

February 1995. Three applications of propiconazole (‘Tilt 250EC’ at a rate of 500 ml/ha)

were applied on November 20th, December 4th and December 11th 1994. Treatment was

initiated when disease incidence had reached 80%.

(ii) a research plot on the Ovens Research Station, September 1996 to February 1997.

Propiconazole (‘Tilt 250EC’ at a rate of 500 ml/ha) was applied on October 23rd,

November 7th, December 9th and January 9th. Treatment was initiated when disease

incidence had reached 60%.

(iii) a research plot on the Ovens Research Station, September 1997 to February 1998.

Propiconazole was applied on November 7th, November 28th and December 12th.

Treatment was initiated when disease incidence had reached 80%.

Twenty shoots were taken at each sampling time and assessed for disease incidence and

disease severity as previously described and the mean values were calculated for each

sample. Disease progress was monitored by plotting the means of the disease measurements

(incidence and severity) against time.

34

Results

Part 1

Visual examination of the disease incidence and severity data for each of the experiments

revealed a relationship that was very similar for each data set. Therefore, the data for all

experiments were combined and plotted as one (Figure 15). The curve could be described

by the exponential equation:

Y = 0.264 x 100.02X (R2 = 0.82)

where Y = disease severity (%) and X = disease incidence (%). The relationship between

the two measurements appeared to be linear at levels of incidence below 60%.

0

10

20

30

40

Dis

ease

seve

rity

(% le

af a

rea

affe

cted

)

0 10 20 30 40 50 60 70 80 90 100Disease incidence (% leaves infected)

y = 0.264 * 100.021x R2 = 0.82

pot experiment data

field experiment data

Figure 15 Incidence - severity relationship for mint rust of peppermint. Data from field

experiments and pot experiments were pooled together.

35

Part 2

The fungicide treatments applied to the peppermint crops in 1994/5, 1996/7 and 1997/8 all

reduced disease incidence and severity considerably, and the time of application had an

effect on disease progress (Figure 16). The earlier the applications began, the less disease

developed. The earliest treatment, applied in October 1996 when disease incidence was

60%, resulted in eradication of rust from the crop, lasting for two months until inoculum

was carried in from elsewhere. There were some strong winds in late December 1996

which could have blown in fresh urediniospores.

0102030405060708090

Dis

ease

inci

denc

e (%

leav

es in

fect

ed)

0

5

10

15

20

25

30

Dis

ease

seve

rity

(% le

af a

rea

affe

cted

)

Sept Oct Nov Dec Jan Feb

O'Sullivan 94/5paddock 97/8paddock 96/7

A

B

Figure 16 Changes in disease levels for peppermint cv. Todd’s Mitcham growing in the ORS paddock plot and O’Sullivans farm - A disease incidence, B disease severity. Tilt was applied on 20th November, 4th and 11th December 1994; 23rd October, 7th November, 9th December 1996 and 9th January 1997; and 7th and 28th November and 12th December 1997.

36

Discussion

In the present study an incidence - severity relationship was found for mint rust on

peppermint that could be described by an exponential equation, with a linear phase at low

levels of disease. Furthermore, the relationship was consistent across experiments carried

out in different years, in pots and in the field. Such an exponential relationship, with a

linear phase at the lower incidence levels, has been reported for other diseases (James and

Shih 1973, Seem 1984, Dillard and Seem 1990a, 1990b, Beresford and Royle 1991).

Dillard and Seem (1990b) successfully used this relationship to identify a level of disease

incidence that could be recommended to maize growers as an action threshold for the

initiation of fungicide treatment to prevent the onset of common maize rust epidemics. In

the light of this, therefore, the preliminary finding of a strong incidence - severity

relationship for the mint rust pathosystem was encouraging, and possibly helpful in

determining a critical time for the initiation of fungicide applications to control the disease.

The action threshold recommended by Dillard and Seem (1990b) for common maize rust

was the level of disease incidence reached at the point where the relationship stops being

linear. After this point, disease severity increases very rapidly and the disease becomes

increasingly difficult to control. In the case of mint rust, this point is at 60% disease

incidence, when disease severity is approximately 5%.

Does this theory work in practice in a peppermint crop? With this question in mind, data

collected over three years in the Ovens Valley was closely examined. In 1996/7, fungicide

treatment was initiated when disease incidence was approximately 60%, and resulted in

complete control of mint rust until late in the season. A preliminary recommendation,

therefore, could be made to peppermint growers that they monitor the disease level in their

crops by regular sampling of stems, and as the percentage of infected leaves per shoot

reaches 60%, begin their fungicide applications. This could be expected to be in middle to

late October, when the canopy is still relatively open and average temperatures are cooler

than the optimum for the uredinial cycle of the rust (see section IB). This would be earlier

than the current practice, and the indications from this study so far are that it could lead to

more effective control of the disease.

37

B. The Use of Flaming in Peppermint Crops

In the major peppermint growing areas of the USA, flaming is a recommended practice in

programs for controlling mint rust (Lewis McKellip, Wm. Leman Company, Bremen,

Indiana, USA, personal communication 1996). Spring flaming of peppermint crops in the

USA is used to destroy the aecial stage of the fungus (Horner 1965, Maloy and Skotland

1969, Hardison 1976) and has been recommended for the same purpose in Australia (Small

1986). Autumn flaming is used as a post-harvest clean-up measure to destroy regrowth and

debris.

Control ‘Tilt’ - autumn Flamed - autumn

no treatment early and late spring ‘Tilt’ no treatment

late spring flame early spring ‘Tilt’

late spring flame

late spring flame

early and late spring flame early and late spring flame early and late spring flame

early and late spring flame early and late spring flame early and late spring flame

late spring flame early spring ‘Tilt’

late spring flame

late spring flame

no treatment early and late spring ‘Tilt’ no treatment

Figure 17 Plot layout, showing where the autumn and spring treatments were applied.

Autumn treatments were applied down (columns in table), and spring

treatments were applied across (rows in table), the trial area.

The use of flaming technology as a potential control method for mint rust in Victorian

peppermint crops was investigated during 1994/5 in conjunction with another RIRDC-

funded project, DAV71A. Autumn flaming, spring flaming, and combinations of both were

compared with the use of propiconazole (as Tilt 250EC) for their effect on disease levels

38

and oil yields of peppermint growing under local conditions (Figure 17). The methodology

and results have already been reported in ‘Evaluation of a chemical free method of

controlling pest and diseases in Australian peppermint crops’ RIRDC Report Project

Number DAV 71A, so only a discussion of the findings will be presented here.

Spring flaming caused considerable disruption to plant growth, resulting in large bare

patches of soil and much weed infestation, and therefore produced comparatively poor

yields of oil, even though the levels of disease and defoliation were low compared to the

other treatments. Effective spring flaming requires a knowledge of the life cycle of

Puccinia menthae under local climatic conditions, which was not available for north east

Victoria until the present study was conducted, when it was discovered that the aecial stage

is not produced. The effectiveness of spring flaming for rust control in Australian

peppermint crops as recommended by Small (1986) was therefore unsubstantiated. In

addition, Grey and Welty (1995) reported that in the USA, while the frequency of rust was

considerably reduced by spring flaming, the peppermint crop subsequently suffered

sporadic plant death and reduced plant vigour. In view of the absence of the aecial stage on

peppermint grown in north east Victoria, and the adverse effects on peppermint growth and

yield, it would be inadvisable to recommend the use of spring flaming for control of mint

rust under north east Victorian conditions.

The autumn-flamed treatments produced significantly higher oil yields than the other

treatments despite carrying high levels of mint rust, and it is important to elucidate the

reasons for its effect. The reduced impact on plant assimilates and shoot damage by pests

and diseases over the subsequent winter may have allowed more shoots to develop to

maturity, therefore increasing biomass. Quadrat counts of plant numbers or measurements

of fresh weight would have indicated whether this was the case. In order to propose any

hypothesis with confidence, it would be necessary to repeat the experiment and include

some measurements of biomass. This was not possible during the present study as the

flamer was not available for a second trial.

There have been several reports over the years recommending autumn flaming as a clean-

up measure for use in peppermint crops (Drummond-Goncalves 1943, Murray 1961,

Horner and Dooley 1965, Maloy and Skotland 1969, Fox 1995, Whitten 1997).

39

Unfortunately, there are two serious impediments to this recommendation in north east

Victoria. First, there is the legal barrier to cross before flaming can be used on a

commercial scale in Australia. It is currently illegal to transport filled bulk liquid petroleum

gas (LPG) tanks as seen in static industrial situations. Such tanks must be transported

empty. The exception is where specially designed recessed valves and controls are fitted to

the trucks. The current sizes of these tanks which are available are far larger than required

for a flamer. The second impediment is that autumn flaming in north east Victoria would

coincide with the high fire risk period, and the special permission which would be required

from the Country Fire Authority would almost certainly be denied.

The application of flaming on peppermint at the Ovens Research Station, October 1994.

40

C. The Effect of the Herbicide Paraquat on the Viability of Puccinia menthae

Urediniospores

Introduction

Several peppermint growers in north east Victoria currently use the contact desiccant

herbicide paraquat (1.5L/ha as Gramoxone, supplied by Crop Care Australasia) as an

autumn clean-up measure and the question is raised as to whether this is an effective

alternative to flaming. The previous experiment established that autumn flaming killed

above-ground insect pests and weeds, and incinerated foliage and debris that served as a

reservoir of overwintering rust inoculum. Paraquat is equally effective against weeds, but

would not affect the insect population or remove debris in the same manner. There have

been some reports of the use of contact herbicides to control mint rust (Campbell 1954,

1955, 1956), including a study by Horner (1965) that compared the effectiveness of such a

herbicide with propane gas flaming, but the focus was always to kill developing aecia and

thus break the life cycle of the fungus. The study of the disease cycle under local conditions

showed that urediniospores carry the disease through winter in north east Victoria where

aecia have not been observed. The objective of this experiment was to examine whether

urediniospore viability was affected by treating peppermint foliage with paraquat.

Methodology

Sixteen 12-week-old peppermint cuttings were inoculated with 5x104 spores/ml and

maintained in the glasshouse for a further six weeks, by which time the leaves were well-

covered with sporulating uredinia. The plants were then randomly divided into four groups,

each containing four plants, and the following treatments were applied:

Treatment C - no treatment (control);

Treatment S - the shoots were cut off at the base of the stem and placed on

aluminium foil on a glasshouse bench;

Treatment L - the individual leaves were cut off the shoots and placed on

aluminium foil on a glasshouse bench;

Treatment P - paraquat (2.5ml/L as Gramoxone) was sprayed with a

hand-held spray unit on to the plants until run-off.

41

The rate of paraquat used was equivalent to that applied in the field (250ml/100L as

Gramoxone).

Urediniospores were collected from each plant by lightly pressing the abaxial surface of an

infected leaf onto 1.5% water agar, then, once the leaf was removed, remaining groups of

urediniospores were spread uniformly across the surface of the agar with a bent glass rod.

The agar plates were incubated for 24 hours in darkness at 20°C, then examined to

determine the percentage of spores which had germinated. At least 400 spores per replicate

were examined, and the mean percentage germination was calculated for each treatment.

This was done immediately prior to application of the treatments and then daily until 20

days after treatment. The experiment was repeated. Since the results were similar for both,

the data sets were combined and the daily means for each treatment were calculated. As the

daily fluctuation of urediniospore viability per se was not of interest in the present study,

the urediniospore germination for each treatment was expressed as a percentage of the

control.

Results

The leaves from Treatments S and L shrivelled and dried within a couple of days. The

paraquat-treated leaves (Treatment P) wilted and died within three days, but did not fall off

the stems. Eight days after the paraquat treatment, the plants began to recover, with the

emergence of new shoots and also new leaves from the nodes of the existing shoots. These

new leaves had to push past the dead leaves still hanging on the stems as they grew and

expanded.

Urediniospore viability remained similar for all treatments for the first six days (Figure 18).

After that, the viability of the paraquat-treated spores was less than that of the other

treatments. By 15 days, the paraquat-treated spores were only 20% as viable as the control

spores, and the viability remained at this level to the end of the experiment (Figure 19).

Spores from dead leaves which had not been killed with paraquat (Treatments S and L)

began to die 15 days after treatment, and then dropped to 60% of the control by 20 days.

42

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

200%

220%

240%

Ure

dini

ospo

re v

iabi

lity

(exp

ress

ed a

s %

con

trol)

0 2 4 6 8 10 12 14 16 18 20

Days after treatment

Treatment P - paraquat

Treatment L - cut leaves

Treatment S - cut stems

Treatment C - control

C

L

S

P

Figure 18 Effect of the treatments on the level of urediniospore viability. Data are

expressed as a percentage of the control treatment.

0%

20%

40%

60%

80%

100%

120%

140%

Ure

dini

ospo

re v

iabi

lity

(exp

ress

ed a

s %

con

trol)

0 2 4 6 8 10 12 14 16 18 20

Days after treatment

y = 1.231 * 10 -0.041x R2 = 0.80

Figure 19 Effect of paraquat on the level of urediniospore viability, expressed as a

percentage of the control treatment.

43

Discussion

These results indicate that the use of paraquat is not equivalent to the use of autumn

flaming. Although urediniospore viability was reduced by paraquat compared to the

control, new shoots and leaves were beginning to emerge within eight days of herbicide

application while the spore viability was still 60%. It is quite possible that as the new

foliage grew passed the dead leaves remaining on the stems, viable urediniospores could be

transferred and infect them or be blown on to them.

As already discussed, however, autumn flaming is not an option due to current legal

constraints concerning bushfire risk. In this experiment, urediniospores taken from

paraquat-treated foliage were less viable than those taken from foliage which had died after

cutting, such as may be left lying on the ground after harvest. Therefore using paraquat as a

clean-up measure, while not as effective as flaming, would substantially reduce the

inoculum level that would be carried over the winter into the next season, and may reduce

its infection potential even more.

D. Evaluation of Fungicides to Control Mint Rust

Introduction

Research is currently being carried out at the Ovens Research Station to evaluate the

potential of Scotch spearmint as a complementary essential oil crop for the region (RIRDC

Project Number DAV 101A ‘Economic feasibility of Native and Scotch Spearmint

production in Tasmania and Victoria’). An objective of this current project (UM 16A) was

to identify fungicides which will effectively control Puccinia menthae without

contamination of distilled oil. Due to the lack of available land at the Ovens Research

Station, however, it was not possible to use field-grown peppermint for a fungicide

screening trial, so the decision was made to collaborate with RIRDC Project DAV 101A

and use rust-infected Scotch spearmint for the fungicide screening trial. The aim, therefore,

was to compare six commercially available fungicides for use against mint rust on Scotch

spearmint grown at the Ovens Research Station across two seasons.

44

Methodology

Scotch spearmint rhizomes were transplanted from a seedbed plot on the Ovens Research

Station to the research plot area in May 1996. A randomised complete block design with

seven treatments and four replicates per treatment was applied to the trial site. The

treatments were:

- control (no fungicide),

- bitertanol (1.7ml/L as Baycor 300EC, manufactured by Bayer),

- triadimenol (0.7ml/L as Bayfidan 250EC, manufactured by Bayer),

- fluquinconazole (30g a.i./100L as Castellan 250WP, manufactured by Agrevo),

- tebuconazole (300ml/ha as Folicur 250, manufactured by Bayer),

- myclobutanil (1.2g/L as Systhane 400WP, manufactured by Hoescht),

- propiconazole (500ml/ha as Tilt 250EC, manufactured by Novartis).

Each row contained seven 1m x 5m plots, and the four rows were separated from each other

by an untreated row of Scotch spearmint, one metre wide. This was to ensure that there was

no drift of fungicide across the rows, and to keep inoculum pressure high. The trial was not

inoculated with urediniospores prior to treatment as the Scotch spearmint was already well

infected with rust from field inoculum.

In May 1997, rhizomes from the first year’s trial site were transplanted to an adjacent area

of paddock, and again a randomised complete block design was used with the same plot

sizes as described above.

The fungicide treatments were applied using a 15L knapsack sprayer with the addition of

1ml/L wetting agent (as Agral 600, manufactured by Crop Care Australasia) and the dates

of application were:

trial 1: 25th November, 12th and 27th December 1996;

trial 2: 1st, 14th and 28th December 1997.

Ten samples per plot were assessed for disease incidence and severity as described

previously. Pre-treatment assessments of each plot in both seasons confirmed that there

45

were no significant plot differences in disease levels before the fungicides were applied

(data not shown).

Assessment of the final disease levels and defoliation levels was undertaken in the week

prior to harvest. At this time, however, the plants were large and bushy, and cutting the

stems at ground level would have been very destructive, removing a large proportion of the

biomass in the plots. In trial 1, twenty samples were taken per plot as ten side-shoots from

the upper half of the plants and ten side-shoots from the lower half of the plants. Mean

disease severity measures were calculated for the upper and lower side-shoot groups both

separately and combined, and the three different measures were compared against the

respective final oil yields per treatment. The disease severity levels on the upper side-

shoots showed the strongest negative relationship with oil yield, and therefore only data

from the upper shoots were analysed further. In trial 2, ten upper side-shoots per plot were

sampled at the final assessment date.

Although the plots were 1m x 5m, only one m2 was harvested from each, which was taken

near the centre of each plot to minimise interplot interference. Measurements of fresh

weight per m2 and oil yield, converted to kg/ha, were made for each plot.

Disease incidence, severity, shoot defoliation, fresh weight and oil yields per treatment

were compared using one-way analysis of variance, and differences were compared using

Fisher’s least significance difference (P < 0.05). Disease severity data were transformed

using a log10 transformation prior to analysis in order to satisfy the requirement for

normality.

Results

Bitertanol and tebuconazole were consistently the most effective fungicides and

fluquinconazole was consistently the least effective in each trial (Table 4 and Figure 20),

even though the level of disease was much higher in the first trial.

Environmental conditions in the first year (trial 1) were extremely favourable for disease

development. Under the consequent high inoculum level, bitertanol significantly reduced

the disease incidence, but in trial 2 with less inoculum potential, disease incidence was the

same across all treatments (Table 4). Disease severity was significantly reduced by all of

46

the fungicides, except fluquinconazole, in both trials, with the most effective, bitertanol,

reducing severity to approximately 5% (Figure 20).

Under conditions conducive to severe disease (trial 1), all fungicides reduced the level of

shoot defoliation compared to the control, but under less disease pressure (trial 2)

fluquinconazole was not effective. Once again, bitertanol was significantly better than all

the rest over both years.

Fresh weights did not differ (P < 0.05) across treatments in trial 1, but there were

significant differences between oil yields, with twice as much oil produced under the

bitertanol and tebuconazole treatments than under the fluquinconazole treatment or the

control. In trial 2, treatments with bitertanol and tebuconazole resulted in significantly

higher yields both in terms of biomass and oil production. The other fungicides were not

significantly better than the control in that year.

There was a clear negative correlation between the final level of disease severity on the

upper side-shoots and the oil yield (Figure 20). The performance of the fungicides was

fairly consistent for both years, although triadimenol was more effective and myclobutanil

less effective when the conditions were conducive to severe disease (trial 1).

47

Table 4 Effects of a range of fungicides on fresh weight, oil yield, disease severity,

incidence and shoot defoliation of Scotch spearmint infected with mint rust.

TRIAL 1 (1996/7)

Fresh weight

(mg)

Oil

(kg/ha

)

Disease

incidence

(%)

Disease

severity

(%)

Defoliation

(%)

Control 2526a 49c 99.0a 66.9e 57.6d

Bitertanol 3609a 104a 54.9b 5.3a 24.2a

Triadimenol 3459a 86b 96.7a 51.2bc 38.6b

Fluquinconazole 2485a 54c 97.9a 65.9e 49.4c

Tebuconazole 3516a 100a 96.8a 27.1b 35.6b

Myclobutanil 3057a 63bc 96.8a 67.9d 42.1bc

Propiconazole 3414a 78ab 98.1a 56.1cd 38.8b

TRIAL 2 (1997/8)

Fresh weight

(mg)

Oil

(kg/ha

)

Disease

incidence

(%)

Disease

severity

(%)

Defoliation

(%)

Control 1868bc 64b 59.5a 26.6d 31.0d

Bitertanol 3294a 135a 56.7a 5.6a 1.1a

Triadimenol 2016bc 64b 68.9a 18.3bc 15.5bc

Fluquinconazole 1744c 59b 62.7a 24.7d 24.6d

Tebuconazole 3248a 113a 60.3a 13.7b 6.4b

Myclobutanil 2560ab 85b 63.9a 20.2c 19.7c

Propiconazole 2512b 82b 61.1a 17.6c 11.8c

Values in the same column sharing a common letter are not significantly different at P = 0.05.

48

Trial 1 (1996/7)

0

20

40

60

80

100

120

control fluquinconazole myclobutanil propiconazole triadimenol tebuconazole bitertanol

Oil

(kg/

ha)

0

10

20

30

40

50

60

70

Dis

ease

sev

erity

(% le

af a

rea)

oil yielddisease severity

Trial 2 (1997/8)

0

20

40

60

80

100

120

140

control fluquinconazole myclobutanil propiconazole triadimenol tebuconazole bitertanol

Oil

(kg/

ha)

0

5

10

15

20

25

30

Dis

ease

sev

erity

(% le

af a

rea)

oil yielddisease severity

Figure 20 Oil yields and final disease severity levels of Scotch spearmint infected with

mint rust after different fungicide treatments. The data presented is the mean

of the replicates.

49

Discussion

The fungicide bitertanol gave the most effective control of mint rust and resulted in the

highest oil yields of the six fungicides evaluated on Scotch spearmint at the Ovens

Research Station. This was consistent over two seasons with different levels of disease in

each. Bitertanol had shown potential in 1989/90 on peppermint and in 1990/91 on Scotch

spearmint (Bienvenu 1992) and was nominated as a fungicide that should be investigated

further. It has also been reported as effective against mint rust on peppermint and corn mint

(Mentha arvensis) in Bulgaria (Margina and Zheljazkov 1994).

Tebuconazole was also effective, its use resulting in high levels of disease control and good

oil yields, particularly under the high inoculum potentials of the first year. It was identified

as a cheaper alternative to propiconazole for use on peppermint crops in Tasmania in 1995

(Warner and Bailey, unpublished data), and has been reported as effective on peppermint in

the USA (Grey and Welty 1995).

The other two fungicides nominated by Bienvenu (1992), propiconazole and myclobutanil,

were not as effective at controlling mint rust as expected. Propiconazole is the most

commonly used fungicide on several mint species and is the currently used fungicide in

both Victorian and Tasmanian peppermint crops. In this experiment, while it was more

effective than the control, it was not as effective as the fungicides discussed above.

Myclobutanil was also not very effective, which was surprising in view of it being the best

fungicide in the 1990/91 screening trial (Bienvenu 1992) and was reported as being

particularly effective against mint rust on peppermint in the USA (Grey and Welty 1995,

1997). In the present study, its effectiveness broke down in trial 1 when the environment

was especially favourable for disease development.

The practice for controlling rust among north east Victorian peppermint growers is to rely

on a single fungicide, and currently it is propiconazole. This practice carries the risk of

developing a fungicide-resistant pathogen population (Skylakakis 1987). The results of this

trial show that several readily-available fungicides are more effective at controlling mint

rust on Scotch spearmint than propiconazole, with a corresponding increase in oil yield. For

example, bitertanol reduced rust severity from 67% to 5% in the first year and from 27% to

6% in the second year, compared to propiconazole which reduced it from 67% to 56% in

50

year 1 and 27% to 18% in year 2. In addition, bitertanol increased oil yield by 112% and

111% respectively, whereas propiconazole increased it by only 57% and 28% respectively.

In view of this, it could be recommended that growers use a combination of fungicides,

alternating their use to reduce the likelihood of fungicide resistance arising in the local

pathogen population. The fungicides identified in this study that could be used in such a

manner were bitertanol and tebuconazole, with triadimenol in seasons particularly

favourable to the development of the disease. Unfortunately, they are all demethylation-

inhibiting (DMI) fungicides which inhibit C-14 sterol demethylation ie. they all have a

similar mode of action, and changes in pathogen sensitivity to one may affect sensitivity to

the others (Heaney 1988). In practice, however, the risk of resistance building up in

response to DMI fungicides is considered moderate to low, as DMI fungicides have been

widely used since the 1970s yet few cases of resistance have been clearly identified

(Georgopoulos 1987, Scheinpflug 1988a, 1988b). Scheinpflug (1988b) recommends that to

minimise the development of DMI fungicide resistance the following management

strategies should be followed:

(1) DMI fungicides should not be used continuously throughout the growing season. The

number of foliar applications should be limited to a maximum of four to eight sprays per

season;

(2) a minimum of two to three consecutive (block) treatments is necessary to achieve the

maximum performance benefit from DMI fungicides. If more than this is required, two or

more blocks of two to four treatments is the preferred method.

Clearly, further screening trials should be undertaken to investigate fungicides with other

modes of action. Incorporation of two or more fungicides from different chemical groups

into the growers’ mint rust control programme should be promoted to ensure that their

effectiveness is not broken down.

51

4. Variation in the Pathogen, Puccinia menthae.

Introduction

Although outside the objectives of this project, a study was also included to investigate the

extent of variation in the Puccinia menthae population present in Victoria. Samples of mint

rust were collected from mint species growing in Melbourne and north east Victoria, and

18 single-spore isolates were established and maintained in the glasshouses at the

University of Melbourne. These isolates of rust were compared on the basis of host range

differences, teliospore morphology, and genetic differences as revealed by a molecular

biology technique known as the random amplified polymorphic DNA (RAPD) assay.

Methodology

Host range differences

Host range differences between P. menthae isolates were examined in the glasshouse by

inoculating 30 clones from 13 Mentha species, and recording the host reactions.

Inoculations were made by spraying sets of three-week-old rooted tip-cuttings with

urediniospore suspensions (104 spores/ml) of each isolate and incubating in darkness at

100% humidity and 20°C for 24 hours. Inoculations were made only during April to

October, when the glasshouse temperatures could be consistently maintained at 14/22°C, as

temperatures above this could lead to unreliable symptom expression. Each isolate was

tested on at least two separate occasions by inoculating sets of plants containing three

cuttings per host clone. Each time, a control set inoculated with water and wetting agent

only was maintained to check for contamination by foreign rust spores. Uredinia appeared

on susceptible hosts after 10 to 12 days, and results were recorded after 18 to 20 days.

Host reactions were recorded on a scale of 0-4 based on that used by Baxter and Cummins

(1953) and Fletcher (1963), and adapted by Beresford (1982), for P. menthae:

0 Immune to extreme resistance. No visible reaction, or small necrotic spots but no

uredinia formed.

1 Highly resistant. Uredinia few, small, in necrotic spots most failing to rupture the

epidermis.

52

2 Moderately resistant. Uredinia fairly abundant, some failing to rupture the

epidermis; in necrotic spots.

3 Moderately susceptible. Uredinia abundant of moderate size, surrounded by

chlorotic spots which later turn necrotic.

4 Highly susceptible. Uredinia abundant, large, with little or no chlorosis. Uredinia

often produced on the upper leaf surface as well as on the lower leaf surface.

‘+’ was used to record reactions varying between two adjacent reaction types. Reaction

types 0 to 2 were considered resistant reactions, and types 3 to 4 were considered

susceptible reactions.

Teliospore morphology

Teliospores were collected from four Mentha species in 1994, 1996 and 1997. Five

collections were from M. x piperita, host of peppermint rust, and five were from M.

spicata, M. x cordifolia and M. suaveolens Ehrh. (applemint), all hosts of spearmint rust.

Teliospores were lifted from the sori with a fine needle, mounted in lactic acid and

examined under bright field magnification (x1000).

From each collection, 50 mature teliospores that had broken off from their points of

attachment were measured for the following characteristics:

- apical cell width and length, measured to the outside of the wall;

- basal cell width and length, measured to the outside of the wall;

- range of wall thickness, measured at the thickest and the thinnest parts;

- septum thickness, measured at the centre of the septum;

- pedicel length;

- apical papilla thickness, measured from the centre of the pore in the spore

wall to the top of the papilla;

- apical papilla width at the base.

Scanning electron microscopy (SEM) was used to examine teliospore wall ornamentation.

Fragments of leaves from M. x piperita cv. ‘Todd’s Mitcham’ and M. spicata cv. ‘Native’

on which teliospores were being produced were air-dried, mounted on an aluminium stub

with a carbon sticky tape and sputter-coated with gold in an Edwards Sputter-coater 5150B.

53

The teliospores were then examined using a Philips XL30 FEG scanning electron

microscope over a range of magnifications from x750 to x 3600.

RAPD analysis

DNA was extracted using the adapted CTAB method of Taylor et al. (1995). 113 decamer

oligonucleotide primers were tested for amplification of P. menthae genomic DNA.

Reaction volumes of 25 µl were optimised to contain 40 ng of DNA template, 0.8 units of

Taq DNA polymerase (Boehringer Mannheim Biochemica, Germany), 0.24 mmol each of

dATP, dCTP, dGTP and dTTP, 0.2 µM of primer, and PCR buffer with a final

concentration of 0.01 M Tris-HCl / 3 mM MgCl2 / 0.05 M KCl / 0.1 mg/ml gelatin, pH 8.3

(Boehringer Mannheim Biochemica, Germany). PCR was performed with a PTC-100

thermocycler (MJ Research, Inc., USA) programmed as follows: initial denaturation step at

94°C for 1 min followed by 35 cycles of 94°C for 10 sec, 40°C for 30 sec and 72°C for 1

min, with a final extension at 72°C for 5 min. Amplification products were separated by

electrophoresis on 1.4% agarose gel in TAE buffer, stained with ethidium bromide and

visualised with UV illumination. The molecular weight marker used was Lambda-DNA

digested with EcoRI and HindIII (Promega Inc., USA). The PCR with each primer was

performed at least twice to check consistency of results.

PCR products from each DNA sample were scored as either present (1) or absent (0). Only

distinct, reproducible bands were scored. Data analysis was conducted using SAS (SAS

Institute 1996) to compute simple matching distances (Sneath and Sokal 1973). UPGMA

cluster analysis was carried out using Proc Cluster in SAS Release 6.12 and a dendrogram

was produced from the output of Proc Cluster using Proc Tree.

Results

Host range differences

The reactions of the 30 host genotypes to the rust isolates differed (Table 5). Some isolates

produced only slight differences in infection types but others produced distinctly different

infection types on several Mentha clones. Fifteen races were separated within the 18

isolates on the basis of resistant or susceptible reactions. There was a clear demarcation

between ‘peppermint rust’ races and ‘spearmint rust’ races, according to the host from

54

which the isolate had originated. Among 12 isolates from M. x piperita and M. x piperita

ssp. citrata, nine ‘peppermint rust’ races were differentiated, and the six isolates from M. x

cordifolia, M. x gracilis and M. spicata each differentiated into separate ‘spearmint rust’

races.

55

Table 5 Host reactions (0-4 s cale) p roducedby18 iso lates o fPucciniamenthae. I so lates followedby the s ame letter ato missing data iso lates 93-03 and 93-06 cou ld not be conclus ively d if feren tiated and cou ld be either a o r b .

Isolat es of P. ment hae[DNA s ampl e]

( host from whi ch origi nal uredi ni os pore sample was coll ect ed)93-08a 93-03ab 93-06ab 93-09b 95-02c 95-21c 93-01d 93-02e 93-05f 95-16g 95-17h 95-14i 93-12j[Mp-04] [Mp-01] [Mp-10] [Mp-05] [Mp-07] [Mp-09] [Mp-08] [Mp-11] [Mpc] [Ms-02]

Hosts inoculated (M. x pi per it a) ( M. x pi per it as sp. cit rat a)

(M. s pi

M. aquat ica 0 0 0 0 1 1 0 0 0 0 0 0 0M. ar vensi s 4 3+ 4 3+ 3 3 3+ 3+ 3+ 3 3 3+ 3M. x cor di fol ia I 0 0 0 0 0 0 0 0 0 1 1 2 4

II - - - - 1 - - 0 - 0 1+ 0 3M. di emeni ca 0 0 0 0 0 0 0 0 0 0 0 1 0M. x gr aci li s I 4 4 3+ 4 4 4 4 3 2 3+ 4 4 4

II 3+ - - 3+ 4 3 4 3 3 3 3+ 4 4III 2 - - 3 3 3 4 2 - 3 3 3 4IV 0 - - 0 0 0 0 0 - 0 0 1 4

M. l axif lor a 3 - - 2 3 3 0 3+ - 3 3 1+ 0M. x maxi mi lianea 0 - - 0 0 0 0 0 0 0 0 0 0M. x pi per it a I 4 4 4 4 4 4 4 4 4 4 4 4 1

II 4 4 4 4 4 4 4 4 4 4 4 4 1III 4 4 4 4 4 4 4 4 4 4 4 4 1IV 4 4 4 4 4 4 4 4 4 4 4 4 0V 4 4 4 4 4 4 4 4 4 4 4 4 0VI 1+ 1+ 1 1+ 1 2 0 1 1 0 4 1 0

M. x pi per it a s sp. ci trat a I 3+ 3+ 3 3+ 1 1 2+ 4 2+ 2 1 2 0+II 0 - - 0 1 0 0 1 - 1+ 0+ 0+ 1III 4 4 3 4 1 1+ 2+ 2 4 2 1 2 1

M. pul egium I 2 1 0 2 2+ 2 1 0 1 3 3 2+ 0II 3 4 3+ 4 4 4 4 4 2 4 4 4 0

M. r ubra r ar ipi la 1 1+ 1 1+ 1 2 0 1+ 0 1+ 0 0 0M. s pi cata I 1 0 0 0 2 1 1 0 0 1 1 0 4

II 1 1+ 2 1 2 1 3 1 1 2 2 0 3III 4 4 4 4 4 4 4 4 4 4 3 4 4

M. s pi cata s sp. cr ispa 0 0 1 0 0 0 0 1 0 0 0 3 4M. s uaveol ens I 0 0 0 0 0 0 0 0 0 0 0 0 0

II 0 - - 0 0 0 0 0 - 0 0 0 0M. x ver ti ci ll at a 0 - - 0 0 0 0 0 0 0 - - -Meli ssa of fi ci nali s 0 0 0 0 0 - - - 0 - - - -Or iganum vul gare 0 0 0 0 0 - - - 0 - - - -Sat ur ej a mont ana 0 0 0 0 0 - - - 0 - - - -- = no inocul at ion made; 0 = i mmune/ extreme res is tance; 1 = highl y resi st ant ; 2 = moderat el y resi st ant ; 3 = moderat ely s us cepti bl e; 4 = hig+ = react ion varying bet ween that shown and t he next hi gher react ion type.

56

M. x gracilis cv. ‘Scotch spearmint’ was susceptible to all isolates tested except 93-05.

There were also differences between the reactions of weed relatives found growing in the

region. Mentha x cordifolia was susceptible to ‘spearmint rust’ races, M. pulegium (II) was

susceptible to ‘peppermint rust’ races and M. spicata (III) was susceptible to all the isolates

tested.

Teliospore morphology

The teliospores could be separated into two distinct groups, peppermint rust and spearmint

rust, according to the host they were taken from (Table 6). Peppermint rust teliospores have

thicker cell walls and septa, longer pedicels and wider apical papillae than the spearmint

rust teliospores. They also have two unequally-sized cells, the apical cell being larger and

thicker-walled than the basal cell, while the spearmint rust teliospores have equally-sized

cells with uniform wall thickness.

Table 6 Comparison of teliospore morphology of peppermint and spearmint rust.

Teliospore character Peppermint rust (µm) Range Mean

Spearmint rust (µm) Range Mean

P-value

Width of spore 21 - 28 23.8 19 - 26 22.1 0.40

Length of spore 24 - 37 31.2 25 - 34 29.1 0.07

Wall thickness

- Thickest part

- Thinnest part

1.4 - 3.2

0.5 - 1.8

2.4

1.0

0.9 - 2.8

0.5 - 1.4

1.8

0.7

0.00

0.00

Width of apical cell 20 - 28 23.6 17 - 25 21.3 0.01

Length of apical cell 14 - 23 18.0 12 - 18 15.6 0.00

Width of basal cell 16 - 26 21.2 18 - 25 21.4 0.82

Length of basal cell 6 - 16 12.8 10 - 18 13.7 0.18

Pedicel length 3 - 64 23.1 4 - 28 14.4 0.05

Septum thickness 0.5 – 2.0 1.1 0.5 – 0.9 0.7 0.00

Apical papilla

- Width at base

- Thickness

8 – 13

2 - 5

10.4

3.5

8 – 11

3 - 5

9.7

3.6

0.05

0.59

57

Differences in teliospore wall ornamentation between the two rust groups were most

apparent under SEM (Figures 21 and 22). The spearmint rust teliospores are uniformly

verrucose, while the peppermint rust teliospores have smooth basal cells and mostly

smooth apical cells.

Figure 21 Wall ornamentation of peppermint rust teliospores revealed by SEM

(bar=10µm).

Figure 22 Wall ornamentation of spearmint rust teliospores revealed by SEM

(bar=10µm).

58

RAPD analysis

Six out of 113 primers amplified distinct and reproducible marker profiles from the 18 P.

menthae DNA samples. Each primer amplified between nine and 18 scoreable DNA

markers, which ranged in size from 100 bp to 3000 bp. In total, 83 bands were scored, of

which 58 were polymorphic (70%).

The dendrogram constructed using the UPGMA algorithm (Figure 23) clearly differentiated

two non-overlapping clusters. One cluster contained the isolates from M. x piperita and the

other, those from M. spicata, M. x gracilis and M. x cordifolia.

Mp -0 1

Mp -0 4

Mp -0 5

Mp -0 7

Mp -0 8

Mp -1 0

Mp -1 1

Mp cPeppermint rust

Mp -0 9

Mg

Ms -0 2

Mc-0 3

Mc-0 2

Mc-0 1Spearmint rust

Ms -0 1

0.3 2 0 .24 0.16 0.08 0 .0 0

Figure 23 Dendrogram constructed using UPGMA cluster analysis from simple matching distances (Sneath and Sokal 1973), using RAPD data from 15 P. menthae isolates. Isolates Mp-01 to Mp-11 were from M. x piperita, Mpc was from M. x

59

piperita ssp. citrata, Mg was from M. x gracilis, Ms-01 and Ms-02 were from M. spicata, and Mc-01 to Mc-03 were from M. x cordifolia.

Discussion

Fifteen physiologic races were identified from 18 isolates of mint rust examined in this

study, indicating the existence of a great deal of variation present within mint rust in

Australia. The lack of internationally accepted differential host clones precludes detailed

comparisons with overseas races, but certain comparisons can be made with the results of

workers in other countries. In France, Cruchet (1907, in Beresford 1982) found eight races

from 10 collections of P. menthae; Niederhauser (1945) identified six races out of six

isolates from the north east USA; in the north western states of the USA, Baxter and

Cummins (1953) found 15 races from 22 isolates and Johnson (1995) differentiated 11

races among 17 collections; Fletcher found nine races in 15 collections from around

England, and Beresford (1982) identified three races in New Zealand. The 15 races found

in the present study can constitute two distinct groups dependent upon the Mentha species

from which they originated. This was expected, as so-called ‘peppermint rust’ and

‘spearmint rust’ have been reported as distinct on numerous occasions in the literature. Of

particular interest in this case, however, was that this separation was also clearly evident at

the morphological and molecular levels.

Two distinct groups were revealed by RAPD analysis, one containing the isolates from M.

x piperita, and the other, isolates from M. spicata, M. x gracilis and M. x cordifolia. The

two groups corresponded to peppermint and spearmint rusts as mentioned previously. The

distinct clustering of the two rust types revealed by the RAPD profiles, with no evidence of

intermediates, is a strong indication that the two groups are genetically isolated.

The two rust groups are also clearly different based on the more traditional comparison of

teliospore morphology. Although the range of values for any one trait may overlap, the

means are usually different, and when all the traits are considered together the two groups

are quite distinct. As presently defined, P. menthae encompasses many varieties and forms

based on teliospore morphology and host range. Sydow (1904, cited in Fletcher 1963) and

Grove (1913) considered P. menthae to be a collective species and proposed division into

several independent species on the basis of such morphological differences as the presence

60

or absence of warts on the teliospores, variations in the apical papillae, and lengths of the

pedicels. The present study shows significant differences between the two groups in each of

these teliospore characters. The spearmint rust teliospores described in the present study fit

within the description of P. menthae var. menthae (Baxter 1959), but the peppermint rust

teliospores described do not.

Differences in host range alone generally characterise formae speciales, whereas

morphological differences can indicate varietal distinction. Until now, the morphological

differences observed within P. menthae have not been deemed sufficient to warrant further

separation of the species. With the addition of the molecular data, there is now a strong

case for some degree of taxonomic separation of these two rusts, but the question posed is

at what taxonomic level? In view of the consistent pattern of difference in the suite of

characters ie. RAPD profile, teliospore morphology and host specificity, in the mint rusts of

Victoria, it appears that peppermint rust deserves at least separate varietal, or specific, rank

from spearmint rust. If the biological definition of a species is genetic isolation, then the

molecular data points in that direction. Given the insights now possible using molecular-

based techniques, it appears that the taxonomy of P. menthae as a whole should be re-

examined.

61

Conclusions

At the outset of the present investigation, there was little known about the behaviour of

mint rust in south eastern Australia, about the extent of variation within Puccinia menthae

in the region, or about the effectiveness of the various control options either already being

adopted or available but untested. The objectives of the present investigation were to

determine and demonstrate an effective strategy for the control of peppermint rust based on

minimal use of chemical sprays, to identify fungicides which will effectively control

Puccinia menthae without contamination of distilled oil, and to conduct detailed studies of

the epidemiology of mint rust eg. life cycle and the influence of temperature and humidity

on its capacity for growth and reproduction. To a large extent these objectives have been

met.

Results of a four-year study showed the full life cycle of the rust does not occur on

peppermint grown locally. No spermogonia or aecia were ever seen on peppermint where

urediniospores persisted all year round. In view of this, considerable emphasis was placed

on the effects of environmental factors on components of the uredinial cycle of P. menthae

on peppermint. Under artificial conditions, urediniospore germination required at least six

hours leaf wetness and lasted four to six hours at temperatures of 5-25°C. Appressoria,

substomatal vesicles and haustoria were produced six, six and 24 hours after inoculation,

respectively, at 20°C. The latent period of infection ranged from 50 days at 5°C to 10 days

at 22°C. Sporulation occurred over a wide range of temperatures (5-27°C), the optimum

being 15-20°C. The minimum, optimum and maximum temperatures for infection were <5,

20 and 27°C respectively.

Infection with mint rust substantially reduced peppermint growth and oil yield. Leaf loss

was significantly increased, which as a consequence significantly reduced leaf area, leaf

fresh weight and oil content per plant. Apparent retention of synthate by the infection

courts was associated with significantly reduced stem and root dry weight and numbers of

stolons. The results suggested that infection with mint rust was likely to reduce the

persistence of peppermint as a perennial crop, which was borne out by the death of the

62

seedbed plot of peppermint after three and a half years of continuous rust infection. Under

disease-free conditions such beds would be expected to persist indefinitely.

Field studies of rust on peppermint crops showed that during summer, the level of disease

was influenced considerably by temperature. Whenever the daily maximum temperature

reached 35°C, the level of disease in crops fell within a few days. This correlated with the

results from the controlled environment experiments on the urediniospore cycle, where it

was seen that high temperatures adversely affected all stages of the process.

There has been recent interest in the disease incidence - severity relationship of several rust

diseases (Seem 1984). This relationship was quantified for common maize rust as an

exponential curve (Dillard and Seem 1990a). The point of inflection in the curve was

determined to indicate a critical level of disease incidence, above which disease severity

increased rapidly. It was considered that when disease incidence in the maize crop reached

this critical level (indicated by the point of inflection in the incidence - severity curve),

chemical control should be initiated. This ‘action threshold’ was recommended to maize

growers, who were then able to successfully control common maize rust under field

conditions (Dillard and Seem 1990b). The incidence - severity relationship for mint rust

was examined in the present study where it was found that it could be described by a single

equation for both field- and pot-grown peppermint. The relationship was exponential, the

point of inflection in the curve occurring when disease incidence reached 60%. This

information provides a basis for developing an action threshold for determining the most

propitious time to apply chemical sprays in Victoria, as was described for common maize

rust in the USA (Dillard and Seem loc. cit.).

The only mint rust control measure currently employed by local peppermint growers is

spraying with the fungicide propiconazole. Spraying is begun when growers perceive that

they have a rust problem, which is usually late in November. They know that they cannot

use the fungicide within 14 days of harvest because of the risk of oil contamination, so that

peppermint crops usually receive three applications of fungicide separated by 14 days, each

at the recommended rate of 500 ml/ha as ‘Tilt 250EC’, during November and December

each year. The investigation of the effectiveness of propiconazole in controlling mint rust

demonstrated that if propiconazole treatment was applied before the disease incidence

63

exceeded 60%, rust was eradicated from the crop, but if applied after the incidence had

exceeded 60%, the level of rust was reduced, but it was not eradicated. This confirmed the

importance and value of 60% disease incidence as an action threshold point for initiating

control programs.

Effective control of mint rust on peppermint crops in the USA depends on breaking the life

cycle of Puccinia menthae at the aecial stage in early spring. This is commonly achieved by

burning the new mint growth with a tractor-drawn propane-gas flamer which destroys any

aecia present and stimulates a second flush of growth into a subsequently rust-free field.

This method of control was investigated for use in north east Victoria, but was found to

cause a severe setback to the spring growth of the peppermint, resulting in reduced oil yield

and quality. The study of the disease cycle also showed that no aecia are produced on

peppermint growing in north east Victoria, which meant that there was no point in targeting

this stage of the life cycle for control purposes.

Flaming the plots in autumn, however, resulted in high yields of peppermint oil in the

following growing season. It is evident from this that if peppermint is allowed to

overwinter free from pests and disease, healthy regrowth results in the subsequent spring.

Unfortunately it is not possible to recommend autumn flaming to growers in north east

Victoria because of the risk of bushfire which has resulted in legal constraints being placed

on gas transportation in addition to legislated fire restrictions.

The use of a contact desiccant herbicide, paraquat, as an early autumn treatment for

controlling the disease was investigated but, although it reduced the viability of

urediniospores on the treated peppermint foliage, the shoots recovered and produced new

leaves while many urediniospores were still viable on the dead foliage. It was concluded,

therefore, that the use of paraquat in autumn was not an effective rust control measure.

Many growers, however, use paraquat in late autumn to ‘clean’ the fields of peppermint

regrowth and weeds prior to winter. This practice would certainly reduce the inoculum load

carried over winter and should therefore slow the onset of rust in the following season.

Six fungicides, namely bitertanol, fluquinconazole, myclobutanil, propiconazole,

tebuconazole and triadimenol, were compared for effectiveness at controlling mint rust on

Scotch spearmint. Two of them, bitertanol and tebuconazole, proved particularly effective

64

against the rust. Both are readily available and cheaper than propiconazole, the currently

used fungicide. Although it is evident that their immediate use in preference to

propiconazole is warranted, it is unfortunate that they are both demethylation-inhibiting

(DMI) compounds, as is propiconazole. Continued use of them poses a slight risk that

fungicide-resistant strains of mint rust may evolve in time (Heaney 1988). It is important,

therefore, that fungicidal compounds with different chemical modes of action are also

evaluated for their effectiveness against mint rust and incorporated into a rotational spray

regime.

Isolates of mint rust collected from four different Mentha species from different regions of

Victoria allowed comparisons to be made of host range, morphology and molecular

differences in the rust. Isolates clearly differentiated into two distinct groups, ‘peppermint

rust’ and ‘spearmint rust’, according to the host species from which they had been

collected. Two groups of mint rust have often been mentioned in the literature, but are both

widely considered to belong to one variety of rust, P. menthae var. menthae described by

Baxter (1952, 1959), which also contains the type specimen for the species. The evidence

presented here strongly suggests that the two groups in fact represent separate species of

rust. While there is overlap of the host ranges on a common host, M. x gracilis cv. ‘Scotch

spearmint’, for the most part they occur on separate groups of hosts. Even on the common

hosts they behave differently and at the molecular level, there is little evidence of gene flow

between them. The two groups exhibit different disease cycles under the same field

conditions, with spearmint rust undergoing annual sexual reproduction while peppermint

rust remains asexual. In addition to these differences, teliospore morphology is consistently

distinct between the two groups.

The taxonomic status of P. menthae has been controversial for many years. Several species

were initially described, and subsequently incorporated into the one species, P. menthae,

making it a species with one of the widest host ranges in the Uredinales. In the 1950s,

Baxter (1952, 1959, 1960) subdivided the species into varieties on the basis of host range

differences and teliospore morphology. The teliospore morphology of the spearmint rust

described herein is consistent with Baxter’s description of P. menthae var. menthae, but the

morphology of the peppermint rust teliospore is outside any existing descriptions, and has

65

only been noted in passing by Walker and Conroy (1969). In view of this, it appears that

peppermint rust should be separated from P. menthae var. menthae, but should it be

regarded as a new variety of P. menthae or as a new species altogether? A comparison of

Victorian isolates of peppermint rust with isolates collected from peppermint grown in

regions where the full life cycle occurs, such as in the USA and the UK, would be desirable

before a decision is made.

66

Bibliography

Baxter, J. W. (1952). A biological study of the mint rust organism, Puccinia menthae Pers. PhD Thesis, Purdue University, West Lafayette, USA. 47 pages.

Baxter, J. W. (1959). Morphologic variation in Puccinia menthae. Lloydia 22: 242-246. Baxter, J. W. (1960). A note on varieties of Puccinia menthae. Mycologia 52: 807. Baxter, J. W. and Cummins, G. B. (1953). Physiologic specialization in Puccinia menthae

Pers., and notes on epiphytology. Phytopathology 43: 178-180. Beresford, R. M. (1982). Races of mint rust (Puccinia menthae Pers.) on cultivated

peppermint and other hosts in New Zealand. New Zealand Journal of Agricultural Research 25: 431-434.

Beresford, R. M. and Mulholland, R. I. (1987). Mint rust on cultivated peppermint in Canterbury: disease cycle and control by flaming. New Zealand Journal of Experimental Agriculture 15: 229-233.

Beresford, R. M. and Royle, D. J. (1991). The assessment of infectious disease for brown rust (Puccinia hordei) of barley. Plant Pathology 40: 374-381.

Bienvenu, F. E. (1992). Development of a viable peppermint oil industry in south eastern Australia. Rural Industries Research and Development Corporation Report DAV 24A, 22 pages.

Bienvenu, F. E. (1993). The emergence of a peppermint oil industry in north east Victoria. The Australian Institute of Food Science and Technology 26th Annual Convention incorporating Australian Essential Oils Industry Seminar, Adelaide, 2 - 7 May 1993.

Brown, J. F. (1997). Fungi with septate hyphae and a dikaryophase. In: Plant pathogens and plant diseases. (eds. Brown, J. F. and Ogle, H. J.), APPS, Rockvale Publications, Armidale NSW, Australia; pages 60-85.

Brown, J. F. and Keane, P. J. (1997). Assessment of disease and effects on yield. In: Plant pathogens and plant diseases. (eds. Brown, J. F. and Ogle, H. J.), APPS, Rockvale Publications, Armidale, Australia; pages 315-329.

Brown, J. S. and Holmes, R. J. (1983). Guidelines for use of foliar sprays to control stripe rust of wheat in Australia. Plant Disease 67: 485-487.

Campbell, L. (1954). Experiments on mint rust control. Phytopathology 44: 484. Campbell, L. (1955). Control of mint rust by premerge treatment with dinitro amine. Down

to Earth 10: 6-7. Campbell, L. (1956). Control of plant diseases by soil-surface treatment. Phytopathology

46: 635. Chandrashekar, M. (1982). Effect of some chemicals employed in the detached leaf culture

of Populus on the infection of Melampsora larici-populina. European Journal of Forest Pathology 12: 301-308.

Dillard, H.R. and Seem, R.S. (1990a) Incidence-severity relationships for common maize rust on sweet corn. Phytopathology 80: 842-846.

Dillard, H.R. and Seem, R.S. (1990b) Use of an action threshold for common maize rust to reduce crop loss in sweet corn. Phytopathology 80: 846-849.

67

Drummond-Goncalves, R. (1943). Ferrugem da hortela pimenta. In Review of Applied Mycology 23: 189.

Fletcher, J. T. (1963). Physiologic specialization in Puccinia menthae. Transactions of the British Mycological Society 46: 345-354.

Fowler, J. and Cohen, L. (1990). Practical statistics for field biology. John Wiley and Sons, Inc., New York, USA; 227 pages.

Fox, R. T. V. (1995). Fungal foes in your garden 30. Mint rust. The Mycologist 9: 84. Georgopoulos, S. G. (1987). The development of fungicide resistance. In: Populations of

plant pathogens - their dynamics and genetics. (eds. Wolfe, M. S. and Caten, C. E.), Blackwell Scientific Publications; pages 239-251.

Grey, W. E. and Welty, L. E. (1995). Effect of fungicide and flaming for control of peppermint rust, 1994. Fungicide and Nematicide Tests 50: 407.

Grey, W. E. and Welty, L. E. (1997). Effectiveness of fungicides for control of peppermint rust (Puccinia menthae). Fungicide and Nematicide Tests 52: In press.

Grove, W. B. (1913). The British Rust Fungi (Uredinales). Their biology and classification. Cambridge University Press, London; 412 pages.

Hardison, J. R. (1976). Fire and flame for plant disease control. Annual Review of Plant Pathology 14: 355-379.

Harvey, I. C. (1979). The impact of rust on peppermint crops in Canterbury 1978/79. Australasian Plant Pathology 8: 44-45.

Heaney, S. P. (1988). Population dynamics of DMI fungicide sensitivity changes in barley powdery mildew. In: Fungicide resistance in North America. (ed. Delp, C. J.), The American Phytopathology Society, Minnesota, USA; pages 89-92.

Horner, C. E. (1965). Control of mint rust by propane gas flaming and contact herbicides. Plant Disease Reporter 49: 393-395.

Horner, C. E. and Dooley, H. L. (1965). Propane flaming kills Verticillium dahliae in peppermint stubble. Plant Disease Reporter 49: 581-582.

James, W. C. and Shih, C. S. (1973). Relationship between incidence and severity of powdery mildew and leaf rust on winter wheat. Phytopathology 63: 183-187.

Johnson, D. A. (1995). Races of Puccinia menthae in the Pacific Northwest and interaction of latent period of mints infected with rust races. Plant Disease 79: 20-24.

Keane, P. J., Limongiello, N. and Warren, M. A. (1988). A modified method for clearing and staining leaf-infecting fungi in whole leaves. Australasian Plant Pathology 17: 37-8.

Kranz, J. (1998). Measuring plant disease. In: Experimental techniques in plant disease epidemiology. (eds. Kranz, J. and Rotem, J.) Springer-Verlag, Heidelberg; pages 35-50.

Maloy, O. C. and Skotland, C. B. (1969). Diseases of mint. College of Agriculture Extension Circular 357, Washington State University.

Margina, A. and Zheljazkov, V. (1994). Control of mint rust (Puccinia menthae Pers.) on mint with fungicides and their effect on essential oil content. Journal of Essential Oil Research 6: 607-615.

68

Murray, M. J. (1961). Spearmint rust resistance and immunity in the genus Mentha. Crop Science 1: 175-179.

Neher, D. A. and Campbell, C. L. (1997). Determining sample size. In: Exercises in plant disease epidemiology. (eds. Frankl, L. J. and Neher, D. A.) The American Phytopathology Society, St. Paul, Minnesota, USA; pages 12-15.

Niederhauser, J. S. (1945). The rust of greenhouse grown spearmint and its control. Cornell University Agricultural Experiment Station Memoir 263, Ithaca, New York.

Nutter, F. W. Jr. (1997). Disease severity assessment training. In: Exercises in plant disease epidemiology. (eds. Francl, L. J. and Neher, D.A.), The American Phytopathology Society, St. Paul, Minnesota, USA; pages 1-7.

Nutter, F. W. Jr. and Schultz, P. M. (1995). Improving the accuracy and precision of disease assessments: selection of methods and use of computer-aided training programs. Canadian Journal of Plant Pathology 17: 174-184.

Parbery, D. G. (1996). Trophism and the ecology of fungi associated with plants. Biological Reviews 71: 473-527.

Parker, S. R., Shaw, M. W. and Royle, D. J. (1995). The reliability of visual estimates of disease severity on cereal leaves. Plant Pathology 44: 856-864.

Peterson, L. (1995). Experience of the Essential Oils of Tasmania Company in developing new crops. The Australian New Crops Newsletter 4: 2-3.

Royle, D. J. (1994). Understanding and predicting epidemics: a commentary based on selected pathosystems. Plant Pathology 43: 777-789.

SAS Institute Inc. (1996). SAS/STAT Software, Changes and Enhancements through Release 6.11 Cary, NC.

Scheinpflug, H. (1988a). History of DMI fungicides and monitoring for resistance. In: Fungicide resistance in North America. (ed. Delp, C. J.), The American Phytopathology Society, Minnesota, USA; pages 77-78.

Scheinpflug, H. (1988b). Resistance management strategies for using DMI fungicides. In: Fungicide resistance in North America. (ed. Delp, C. J.), The American Phytopathology Society, Minnesota, USA; pages 93-94.

Seem, R. C. (1984). Disease incidence and severity relationships. Annual Review of Phytopathology 22: 133-150.

Shtienberg, D. and Vintal, H. (1995). Environmental influences on the development of Puccinia helianthi on sunflower. Phytopathology 85: 1388-93.

Skylakakis, G. (1987). Changes in the composition of pathogen populations caused by resistance to fungicides. In: Populations of plant pathogens - their dynamics and genetics. (eds. Wolfe, M. S. and Caten, C. E.), Blackwell Scientific Publications; pages 227-238.

Small, B. E. J. (1986). Mint oils. Agfact P6.4.1. Department of Agriculture, New South Wales, Agdex 184.

Sneath, P. H. A. and Sokal, R. R. (1973). Numerical Taxonomy. W.H. Freeman: San Francisco.

69

Taylor, P. W. J., Fraser, T. A., Ko, H.-L. and Henry, R. J. (1995). RAPD analysis of sugarcane during tissue culture. In: Current issues in plant molecular and cellular biology. (eds. M. Terzi, R. Cella and A. Falavigna), Kluwer Academic Int., Dordrecht, The Netherlands; pages 241-246.

Teng, P. S. and Close, R. C. (1978). Effect of temperature and uredinium density on urediniospore production, latent period, and infectious period of Puccinia hordei Otth. New Zealand Journal of Agricultural Research 21: 287-96.

Walker, J. and Conroy, R. J. (1969). Puccinia menthae Pers. in Australia. The Australian Journal of Science 32: 164-165.

Warner, J. and Bailey, L. (1995). Rust control in peppermint (Mentha x piperita). Unpublished report for Essential Oils of Tasmania, 4 pages.

Whitten, G. (1997). Herbal harvest: commercial production of quality dried herbs in Australia. Agmedia, Daratech Pty. Ltd., Melbourne; 555 pages.

Zadoks J. C. (1985). On the conceptual basis of crop loss assessment: the threshold theory. Annual Review of Phytopathology 23: 455-473.