Mimosa Pigra Dieback in the Northern Territory, Australia: Investigation Into Possible Causes

Mimosa pigra dieback in the Northern Territory, Australia: Investigation into possible causes

Aniline D. Sacdalan Master of Science in Biology

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2015

School of Agriculture and Food Sciences

Abstract

In this study, Mimosa pigra dieback is described as plants or population of plants which appeared to suffer from ill health, death of their growing points and the emergence of green reshoots coming from the base. Dieback is an important phenomenon, not only in many crops and natural forests worldwide, but also in many invasive trees and shrubs in Australia. The dieback phenomenon is thought to be caused by complex and interacting factors and causes are often difficult to disentangle. In some cases it is a result of an interaction between both biotic and abiotic factors. Since 1988, dieback has been widely observed in Mimosa pigra L. (Fabaceae), a woody shrub that is one of the most serious exotic invaders of floodplains and riparian areas in the Top End of the Northern Territory. However, there has been no conclusive evidence of links between M. pigra dieback and specific causal factors. In addition, the dieback phenomenon has not been previously quantified.

This thesis aimed to describe M. pigra dieback phenomenon at plant and population levels and its effect on plant growth, survival and demography. Specifically, the aim of this thesis was to evaluate the role of biotic factors (insect borers and fungal pathogens), and abiotic factors (soil pH and salinity), in relation to the development of M. pigra dieback. In addition, the thesis examines the pathogenicity of the fungal endophytes/pathogens associated with dieback- affected and healthy M. pigra. M. pigra has been a focus of biological control programs for more than 30 years, but it still remains a serious threat in the wetlands and conservation areas in the Top End of the Northern Territory. While extensive studies have been conducted to evaluate the impact of biological control programs, little is actually known about dieback and its causal factors.

To evaluate the causal factors of M. pigra dieback, repeated ground surveys, plant health assessment and specimen collection of stems and soils, were conducted in the period 2011 to 2013 in established M. pigra sites on Melaleuca Station, Mary River, Northern Territory. Insect borers were quantified and putative fungal pathogen abundance were surveyed. A combined morphological and phylogenetic approach was used for fungal pathogens/endophyte identification. Using a subset of isolates, trials were performed to examine differences in the pathogenicity of the recovered fungal pathogens/endophytes.

Results of field research showed that dieback in M. pigra is highly dynamic and fluctuates in its severity. Different sites showed increases and/or decreases in plant health over time.

Dieback affects plants in all age classes, from seedlings, juveniles to adult plants, resulting in partial mortality of plants in all age categories over the trial period. Contrary to expectations, none of the biotic and abiotic factors examined showed a direct relationship to the changing population dynamics of M. pigra dieback over time. Nevertheless, fungal isolations demonstrated that M. pigra is a host to a diverse range of endophytes, with the most frequently isolated taxa belonging to the Botryosphaeriales, Xylariales and Diaporthales.

Results of the fungal identification studies showed that M. pigra harbours diverse endophyte communities. The use of morphological characters was crude and unreliable. However, species identification using the ITS sequence data was also insufficiently discriminatory while inclusion of sequence data based on EF1-α allowed the discrimination between sibling species, especially among the confusing taxa within the Botryosphaeriaceae. The pathogenicity trials undertaken revealed the potential of fungal species from Botryosphaeriaceae to kill seedlings and cause lesion development in juvenile and adult plants in both controlled and field conditions.

Overall, the research findings highlight the challenges in attributing a single factor as being responsible for the dieback phenomenon. This is the first study to investigate M. pigra dieback and the results reveal that M. pigra dieback has potential dramatic plant population-level effects. The results also provide evidence of the ubiquity of the phytophagous insects and fungal endophytes across all sites and seasons. Furthermore, among the 1,352 fungal isolations, it is likely that new fungal species will be identified. However, a number of factors were able to be ruled out in relation to potential causes of M. pigra dieback. Continued efforts are needed to focus on interaction studies of biotic and abiotic factors in relation to M. pigra dieback. The fact that population (stand) dieback in M. pigra occurs in multiple catchments in the Northern Territory, and over large areas, makes the dieback phenomenon to be of considerable ecological importance.

iii

Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

Publications during candidature

Sacdalan, A., Galea, V., Goulter, K., Elliott, L., and van Klinken R.D. (2012).Preliminary investigations of the Mimosa pigra dieback phenomenon. In: Valerie Eldershaw, Proceedings of the 18th Australasian Weeds Conference. 18th Australasian Weeds Conference 2012, Melbourne Australia, (356-358). 8-11 October 2012.

Sacdalan, A., Galea, V., Goulter, K., Elliott, L., and van Klinken R.D (2013). Dieback- affected and healthy Mimosa pigra stands have similar endophyte communities in the Northern Territory, Australia. In: Abstract and Programme Book: 19th Australasian Plant Pathology Conference, Auckland, New Zealand.

Publications included in this thesis

No publications included.

Contributions by others to the thesis

Galea, V. J (Thesis Principal Supervisor) for the experimental design, guidance and critical revision of the whole thesis.

Goulter, K.C (Thesis Advisor) for the guidance, training and support in my laboratory, glasshouse and sequencing work and critical revision of the whole thesis.

van Klinken, R. D (Thesis Advisor) for the guidance in the logic, set-up, data entry, data analysis and writing of my field ecology chapter and for the critical revision of the whole thesis.

Lukitsch, B. (NT-DLRM) our dieback manager, for the organisation of fieldwork logistics from permit acquisition to field data collection.

Elliott, L. (NT-DLRM) for the assistance in the logic of the thesis, suggestions for the field design and data collection.

Bissett, A. (CSIRO) for the molecular sequencing identification of the first batch of mimosa isolates.

Statement of parts of the thesis submitted to qualify for the award of another degree

None.

Acknowledgements

First and foremost, I thank my supervisors: Associate Professor Vic Galea, Dr Ken Goulter and Dr Rieks van Klinken. This work would have not been realised without their guidance and constant support. I truly appreciate that they had encouraged me to do my best and for being very understanding when it mattered.

I would like to acknowledge the Australia Leadership Awards (AusAID) for the very generous 4 years of financial support and scholarship at the University of Queensland- Gatton Campus. Special thanks to Catherine Fitzgerald, Sue Quarton and Cheryl Brugman of UQ International.

To the School of Agriculture and Food Science (SAFS) of the Faculty of Science staff: Allan Lisle, Victor Robertson, Mel, Lochie, John, Brigitte, Ross, Heidi, Katherine Raymont, Ross Bourne, Kylie and Kaye for the support given.

I am indebted to the Department of Land and Resource Management (DLRM), Northern Territory Government. To: Bert, Lou, Tash, Susanne, and Phil. Thank you for all the time and assistance given during my fieldwork.

To Tony Searle, for allowing us to do fieldwork on his property and for sharing his knowledge and expertise on mimosa in the area.

Life as a graduate student can be very challenging, but I am glad to have amazing friends who cheered me up when I felt low. To Susanne Casanova and Michele Foley, thanks for all the wonderful times we spent together. I will surely miss you both.

I am forever grateful to my beloved hubby, Alwen Jess, who has been very understanding and supportive of my research endeavour all these years. Life has been so much more enjoyable since you came along.

To my parents and three sisters for the encouragement.

Above all, I am thankful to the Lord Almighty for giving me the strength, courage and knowledge that I needed to finish my PhD journey.

vii

Keywords

Mimosa pigra, dieback phenomenon, fungal endophytes, biological control, mycoherbicide, Lasiodiplodia theobromae, pathogenicity.

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 060704 Plant Pathology, 60%

ANZSRC code: 070308 Crop and Pasture Protection (Pests, Diseases and Weeds), 40%

Fields of Research (FoR) Classification

FoR code: 0607 Plant Biology 60%

FoR code: 0703 Crop and Pasture Production 40%

viii

Table of Contents

Abstract ...... ii Declaration by author ...... iv Publications during candidature ...... v Publications included in this thesis ...... v Contributions by others to the thesis ...... v Statement of parts of the thesis submitted to qualify for the award of another degree ...... vi Acknowledgements ...... vii Keywords ...... viii Australian and New Zealand Standard Research Classifications (ANZSRC) ...... viii Fields of Research (FoR) Classification...... viii

Table of Contents ...... ix List of Figures ...... xiii List of Tables ...... xviii CHAPTER 1 ...... 1 1.1 Research background ...... 1 1.2 Thesis aim and objectives ...... 2 1.3 Thesis outline ...... 3

CHAPTER 2 ...... 5 2.1 The Host Plant Mimosa pigra L...... 5 2.2 The Biology of Mimosa pigra L...... 5 2.3 Distribution of Mimosa pigra in Australia and its Impact on Agriculture ...... 7 2.4 Mimosa pigra Risk and Control ...... 11 2.4.1 Physical and Mechanical Control ...... 11 2.4.2 Chemical Control ...... 11 2.4.3 Biological Control ...... 12 2.5 Classical Biological Control of Mimosa pigra ...... 14 2.5.1 Insect Biocontrol...... 14 2.5.2 Fungal Biocontrol ...... 16 2.6 Dieback...... 17 2.6.1 Dieback Definition ...... 17 2.6.2 Dieback Causes...... 17 2.6.3 Mimosa pigra Dieback ...... 18

2.6.4 Dieback in other Weeds of Australia ...... 20 2.7 Impact Evaluation of Biological Control Programs ...... 21 2.7.1 The Importance of Impact Evaluation ...... 21 2.7.2 Challenges of Impact Evaluation of Biocontrol Programs ...... 21 2.7.3 Different Methods of Evaluating the Impact or Causality of Biocontrol Programs 21 2.7.4 Evaluation of Mimosa pigra Biocontrol Programs ...... 22 2.8 Possible Role of Biotic Factors in the Development of Dieback in M. pigra ...... 23 2.8.1 Endophytes ...... 23 2.8.2 Fungal Pathogens ...... 28 2.8.3 Pathogenicity and Virulence Defined ...... 29 2.8.4 Native Borers ...... 29 2.9 Possible Role of Environmental Factors in the Development of Dieback in M. pigra .. 30 2.9.1 Floodplain Habitat ...... 30 2.9.2 Temperature ...... 30 2.9.3 Soil Acidity ...... 31 2.9.4 Waterlogging and Salinity ...... 31

CHAPTER 3 ...... 33 3.1 Abstract ...... 33 3.2 Introduction ...... 33 3.3 Method ...... 36 3.3.1 Study System ...... 36 3.3.2 Describing M. pigra dieback phenomenon...... 40 3.3.3 Insect Activity Assessment ...... 42 3.3.4 Endophyte Community Assessment ...... 42 3.3.5 Abiotic Conditions ...... 43 3.3.8 Data Analysis ...... 43 3.4 Results ...... 44 3.4.1 Plant Health ...... 44 3.4.2 Plant Growth ...... 46 3.4.3 Plant Demography ...... 48 3.4.4 Abiotic Factors ...... 50 3.4.5 Biotic Factors: Stem Borers ...... 51 3.4.6 Biotic Factors: Fungal endophytes/pathogens ...... 54

3.5 Discussion ...... 57 3.6 Conclusion ...... 59

CHAPTER 4 ...... 60 4.1 Abstract ...... 60 4.2 Introduction ...... 61 4.3 Methods ...... 62 4.3.1 Study Area and Field Sampling Design ...... 62 4.3.2 Specimen Collection ...... 62 4.3.3 Fungal Isolations...... 63 4.3.4 Identification of Recovered Isolates ...... 64 4.3.4.1 Morphological Identification ...... 65 4.3.5 Microscopy ...... 66 4.3.6 Phylogenetic Analysis ...... 68 4.3.7 Final Designation of the Isolates ID Using Morphology and Genetic Data...... 68 4.3.8 Storage and Maintenance ...... 69

4.4 Results ...... 69 4.4.1 Fungal Endophytes of M. pigra ...... 69 4.4.2 Agreement between Morphological and Phylogenetic Characters ...... 71 4.4.3 Taxa of Mimosa pigra Endophytes ...... 72 4.5 Discussion ...... 77 4.6 Conclusion ...... 79

CHAPTER 5 ...... 80 5.1 Abstract ...... 80 5.2 Introduction ...... 80 5.3 Materials and Methods ...... 82 5.3.1 Collection of Samples ...... 82 5.3.2 Fungal Isolations...... 83 5.3.4 Preliminary Screening of Isolates and Trial Assessment ...... 85 5.3.5 Confirmation of Infection in Seedling Trials ...... 86 5.3.6 Identification of Isolates ...... 86 5.3.7 Growing of M. pigra Plants at the UQ Gatton Plant Nursery Unit ...... 86 5.3.8 Study Design and Treatments ...... 87 5.3.9 Inoculum Preparation and Plant Inoculation ...... 87

5.3.10 Plant Assessment ...... 87 5.3.11 Confirmation of Infection ...... 88 5.3.12 Data Analysis...... 88 5.3.13 Plot Selection, Design and Treatments ...... 88 5.3.14 Inoculum and Capsule Preparation ...... 89 5.3.15 Plant Inoculation ...... 90 5.3.16 Trial Assessment...... 90 5.3.17 Confirmation of Infection for Field Samples ...... 90 5.3.18 Data Analysis ...... 90 5.4 Results ...... 90 5.5 Discussion ...... 96

CHAPTER 6 ...... 98 6.1 Recommendations for future research work: ...... 102 6.2 Conclusions ...... 102 7.0 References ...... 104

Appendices ...... 128 Appendix A: Permit to use a declared weed from Northern Territory Government (Department of Land Resource Management)...... 128 Appendix B: Ecological Studies Data Sheets ( Chapter 3) Data sheet(Mimosa pigra) ...... 130 Appendix C: Soil pH and Electrical Conductivity (EC) Determination (1:5 soil/water method) ...... 132 Appendix D: Representative conidia of Botryosphaeriaceae morphotypes ...... 133 Appendix E: Rationalisation table used to compare different characters...... 134 Appendix F: Maximum parsimony of Pesalotia EF sequences...... 140 Appendix G: Nursery Propagation Media ...... 141 Appendix H: Sequence Data Identification and its corresponding MP (Mimosa pigra) Codes...... 142 Appendix I: List of isolates used (numbers and descriptions) for each of the three experiments ...... 144

xii

List of Figures

Figure 2. 1 Mimosa pigra photographed at Melaleuca Station, Northern Territory: (A) main infestation, (B) adult plants, (C) mature pods, (D) flower ------6

Figure 2.2 Map Showing the Northern Territory M. pigra Catchments (Source: Weed Management Branch, Northern Territory). ------9

Figure 2.3 Adult Carmenta mimosa (A); Adult Neurostrota gunniella (B) Photos from Weeds Branch, Northern Territory. ------16

Figure 2.4 Decline Disease Spiral (Manion, 1992). ------19

Figure 2.5 Mimosa pigra population dieback at Melaleuca Station, characterised by stem death but with the presence of green, young reshoots coming from the base and main stems ------20

Figure 2.6 The nuclear ribosomal RNA gene cluster is comprised of a Non-Transcribed Spacer (NTS), External Transcribed Spacer (ETS), 18S rRNA, Internal Transcribed Spacer 1 (ITS1), 5.8S, Internal Transcribed Spacer 2 (ITS2) and 28S rRNA genes. IGS stands for the intergenic spacer region (adapted from Mitchell, 1995). ------29

Figure 3.1 An example of a healthy class of M. pigra population (usually located on the edge of a stand) (A); an example of M. pigra population showing a partial dieback class (B). ------38

Figure 3.2 An example of a recovering class of M. pigra population (A); an example of M. pigra population showing a “dead” class (B). ------38

Figure 3.3 The study site: Melaleuca Station, Mary River, Northern Australia (inset) A, and map of Australia (B) ------39

Figure 3.4 An example of a 5 meter by 5 meter plot, with pickets on each corner and boundaries being set using a tape measure. The site was defined before any measurement and census of seedlings, juvenile and adult plant counts in the plot. ------40

Figure 3.5 Measuring of canopy width using a height rule (A). Volume calculation used the formula of a cone (B). ------42

Figure 3.6 Plant health through time for four replicate plots initially assessed as healthy (a), dieback (b), recovering from dieback (c), and dead (d) (from 2011 to 2013). The box

xiii

descriptors in (a) are the disease severity classes from the 10- point disease rating scale developed for M. pigra (Table 3.2). ------45

Figure 3.7 An initially “dead” plot at Melaleuca Station, where all M. pigra adults and juveniles are dead, but in the background are healthy paperbarks (Melaleuca quinquenervia) (A), on the same site are healthy M. pigra seedlings (about 10 cm in height). ------46

Figure 3.8 Plant growth against initial plant volume in the first and second years (from healthy to dead plots). Only new plants are included in the dead site and were first tagged in year 2. ------47

Figure 3.9 Stem diameter averaged for the three biggest stems of individual plants ordered by the plants’ initial size. Open squares are dead plants. ------47

Figure 3.10 The relationships between adult plant health (line, right-hand axis) versus average seedling densities (bars, left-hand axis) through time for plot that were initially classified as a) healthy, b) dieback, c.) recovering dieback and d.) dead. ------49

Figure 3.11 Principal Coordinate Analysis of the different treatments and the abiotic factors (pH and salinity). ------51

Figure 3.12 Relationship between insect abundance and prior change in health class. Scatterplot diagram of three insect borers: Carmenta, blue circles, Maroga, orange circles and Platyomopsis, gray circles scatterplot showing no support for the change in plant health 6 months before (t-(t-1)). ------52

Figure 3.13 Relationship between insect abundance versus change in health class in the subsequent time step ((t+1)-t. Carmenta, blue circles; Maroga, orange circles; and Platyomopsis, gray circles; scatterplot shows no support for the change in plant health 6 months after ((t+1)-t. ------53

Figure 3.14 Scatterplot diagram of the total borer abundance versus change in health class in the prior time step (t-(t-1)). ------54

Figure 3.15. Scatterplot diagram of the total borer abundance versus change in health class (HC) in the subsequent time step ((t+1)-t. ------55

xiv

through time. The Botryosphaeriaceae were abundant and present in almost all plots but the pattern cannot explain the declining health change of the different dieback levels. ------55

Figure 3.17 Abundance of Pestalotiopsis spp. across all plots and through time (blue and red legend represent the replicate plots). The abundance of Pestalotiopsis spp. fluctuates through time except on the dead site. No strong and clear pattern was seen on the plant health change over time. ------56

Figure 3.18 Abundance of Diaporthe spp. across all plots and through time (blue and red legend represent the replicate plots). The abundance of Diaporthe spp. through time is not related to the declining health of plants ------56

Figure 4.1 Adult plant stem collection (a); Stem branches cut from adult dieback plots (b) -63

Figure 4.2 Fungal isolation methodology. Cutting into stem pieces (a); Debarked stem pieces; (b) Plating the stem pieces into different media (c); fungal culture growth (d) ------64

Figure 4.3 (a) Sporulation using pine needles; (b) sporulation on oatmeal agar. ------66

Figure 4.4 Conidia of Pestalotiopsis type (a) and Botryosphaeriaceae (b). Scale white bar = 10µm. ------67

Figure 4.5 Phylogenetic tree of M. pigra based on the bootstrap analysis of ITS sequences. The tree was rooted to Sclerotinia sclerotiorum. ------70

Figure 4.6 Maximum parsimony tree based on bootstrap analyses of EF1-α sequence of the Botryosphaeriaceae isolates in M. pigra. The tree is rooted to Diplodia seriata (CBS 112555). Isolates in bold were obtained from GenBANK. The bar represents 10 changes. Additional information on sample identification is found on Appendix H.73

Figure 4.7 One of the most parsimonious trees obtained from the combined analysis of ITS and EF1 – α sequences from the species of the genera Lasiodiplodia, Cophinforma and Neofusicoccum. Additional information on sample identification is found on Appendix H. ------74

Figure 4.8 The different Botryosphaeriaceae morphotypes of M. pigra (a - j are the order of groupings Botryo A, B, C, D, E, F, G, H, I & J, respectively, and k-l are the representative conidia of this taxa). Scale white bars = 10µm. ------75

Figure 4.9 The non-Botryosphaeriaceae isolates. Pestalotiopsis A & B (a-b) the spore of Pestalotiopsis (c), Diaporthe spp. A&B (d-e) beta-condia of Diaporthe spp. (f), C. cladospoiriooides (g) Colletotrichum spp. (h) Epicoccum spp. (i), Unidentified isolates (j-l). Scale white bar = 10µm. ------76

Figure 5.1 Steps in the isolation and pathogenicity steps of M. pigra. (A) stem sampling from 4 different dieback expression sites; (B) ~1cm stem pieces ready for isolation; (C) stem pieces plated on four types of media, initial isolation from stem pieces. ------83

Figure 5.2 M. pigra seeds: individual seeds with bristle cover (A); seeds in pods (B). ------84

Figure 5.3 Treatment/inoculum preparation using the gelatine capsules (A); the different treatments taken to the field; (B) sealing of the plant after inoculation;(C) harvesting inoculated stems using a chainsaw (D). ------89

Figure 5. 4 Results of the pathogenicity trials. Disease symptoms on seedling trials; necrosis on tissues, fungal colonisation by mycelia and black spots on cotyledons (Figure 5.4 A-C); glasshouse external and internal lesion development above and below the inoculation point (6 weeks after inoculation) (Figure 5.4 D-F); glasshouse negative control showing no internal dark staining (Figure 5.4 G); oval-shaped lesion observed 6 months after field inoculation trials at Melaleuca Station (Figure 5.4 H); internal staining observed an inoculated field main stem (Figure 5.4 I); field inoculated stem split lengthways showing lesion development into the pith (Figure 5.4 J-K). ------93

Figure 5.5 Average length (mm) of lesions produced on M. pigra field adult plants by 3 isolates belonging to Lasiodiplodia and two isolates from Diaporthe species (most isolates were sourced from healthy trees), a positive control (NT 039) and a negative control (sterile millet) 6 months after the inoculation trials. Isolates are arranged in order of decreasing virulence and the coloured bars indicate from which site the recovered isolates originated. (Error bars represent the standard deviations from the mean). --93

Figure 5.6 The proportion of dead seedlings at harvest two weeks after inoculation (10 replicate seedling trials), n=20 ------94

Figure 5.7 Average length (mm) of lesions produced on M. pigra juveniles inoculated with isolates belonging to Botryosphaeriaceae species (isolates came from healthy to dead tree sources in the field), a positive control (NT 039) and a negative control, after 6 weeks in a glasshouse. Isolates are arranged in decreasing virulence and the grid lines

xvi

indicate sites from which the recovered isolates originated. (Error bars represent the standard deviations from the mean). ------95

xvii

List of Tables

Table 2.1 Important wetlands infested with M. pigra in the Northern Territory (Source: Walden et.al 2000) ------8

Table 2.2 Summary of the negative impacts of M. pigra in the Top End of the Northern Territory. ------10

Table 2.3 Summary of control methods used to control M. pigra in the Top End of the Northern Territory. ------10

Table 2.4 Impact Status of pathogenic fungi following authorised introduction to Australia for the biological control of weeds ------13

Table 2.5 Status of agents released for biological control of Mimosa pigra (in chronological order of release in Australia) ------15

Table 2.6 Evaluation studies conducted on M. pigra in the Northern Territory (compiled from the different published evaluation studies on M. pigra). ------24

Table 2.7 Common and major species of woody fungal endophytes. ------26

Table 3.1 Dieback class descriptions of M. pigra used in this study. ------37

Table 3.2 Ten-point disease rating scale developed for M. pigra. ------41

Table 3.3 Average annual count of live adult plants (>1.5 m tall), juveniles (0.2 – 1.5 m tall) and seedlings (0.2 m tall) per plot over three years in a 5 m by 5 m plot. ------48

Table 3.4 Percent mortality of tagged adult plants per 5 m x 5 m plot over three years. ------49

Table 3.5 Mean pH and salinity of the dieback classes. ------50

Table 3.6 Abundance of Neurostrota gunniella across all plots and seasons. On average, N. gunniella abundance did not explain the changes in M. pigra health over time. However, N. gunniella is ubiquitous and abundant in all plots and seasons. ------53

Table 4.1 Fungal endophyte communities in Mimosa pigra over time. ------71

Table 6.1 Summary table collating all finding of this Thesis and its possible implication. ---98

xviii

CHAPTER 1

General Introduction

1.1 Research background

Plants which appear to be suffering from ill health and death of their growing points are characterised as experiencing “dieback”. The term ‘dieback’ has been used to define many perennial plant disease syndromes (Houston 1992; Jurskis 2005). Dieback involves crown thinning because of foliage loss, failure of normal new growth to replace foliage and death of branches beginning at the growing tips and spreading to the primary branches (Ellis 1964; Innes 1992). Dieback is an important phenomenon which occurs in many perennial crops and natural forests worldwide. Dieback is thought to be caused by many complex and interacting factors (Desprez-Loustau et al. 2006; Bansal et al. 2013). It has also been observed in many invasive trees and shrubs across Australia (Scarlett et al. 2013; Aghighi et al. 2014). An understanding of the primary causal mechanisms of dieback events is complicated by the complex interactions of biotic and abiotic factors, and many of the dieback symptoms are not unique to those caused by single agents (Elmer 2014). Thus, there is a need to consider the “multiple stress hypothesis” (Smith et al. 2012; Mitchell et al. 2013; Bergstrom et al. 2015) in evaluating and describing this phenomenon.

Mimosa pigra L. (Fabaceae) is a Weed of National Significance (WONS) in Australia where it invades extensive wetland areas in the Top End of the Northern Territory (Thorp & Lynch 2000). Widespread dieback of M. pigra has been reported in the Adelaide River floodplains in the Northern Territory since at least 1988 (Miller 1988). In 2005, the Northern Territory Weed Management Branch of the Department of Land Resource Management established a set of long-term monitoring plots to evaluate the health and spread of mimosa stands, including two plots at Melaleuca Station in the Mary River catchment. These plots recorded the progression of dieback leading to the death of almost all M. pigra plants in the monitored plots by 2009.

M. pigra has also been the subject of classical biological research control programs for more than 30 years and still remains a serious problem. A biocontrol program which commenced in 1979 released 15 insects and 2 fungi, resulted in partial success (Heard 2012). Although extensive research has been carried out on the monitoring and evaluation of classical biocontrol agents of M. pigra particularly insects and also a rust fungus (McIntyre & Menges 2004; Paynter 2005, 2006; Ostermeyer & Grace 2007), very little is known about the potential of

classical biocontrol and the other possible contributing factors at play in relation to the occurrence of the dieback phenomenon. Furthermore, most of the evaluation studies have been short term and were conducted in controlled conditions (Steinbauer 1998; Paynter & Hennecke 2001). The manipulative and post-release evaluation studies often concentrated on the effects of a single agent on only one plant stage at one sample time and tell us very little about the overall dieback phenomenon.

Quantitative long-term studies of insect herbivores, together with reliable plant records (e.g. growth parameters) provide ample evidence of the importance of plant insect enemies that influence plant populations (Maron & Crone 2006). One of the greatest challenges in evaluating the impact of biocontrol programs is that there are confounding abiotic and biotic factors that may be overlooked. Studies such as those of Sims-Chilton et al. (2010) have identified climate as a confounding factor in the growth reduction of groundsel bush, Baccharis hamifolia, which has been a subject of biological control programs in Australia since 1969. Quantitative field and laboratory studies of Biswas et al. (2015) identified landscape heterogeneity as a confounding factor in the patterns and demographic consequences in a Canadian invasive forb, Alliara petiolata.

The fact that population (stand) dieback in M. pigra occurs in multiple catchments in the Northern Territory and over large areas, makes the dieback phenomenon of considerable ecological importance. To date, there has been no reliable evidence on the links between M. pigra dieback and its causal factors. The dieback phenomenon has not previously been quantified. There is a need to conduct repetitive ground surveys and quantitative assessments of the plant growth dynamics in relation to potential contributing biotic factors (e.g. pathogens/endophytes and native borers) and abiotic factors (e.g. soil acidity and soil salinity) of dieback in M. pigra.

1.2 Thesis aim and objectives

Aim

To evaluate the role of two biotic factors, fungal endophytes and insect borers, and two abiotic factors, soil acidity and salinity, in relation to the development of dieback in M. pigra at Melaleuca Station in the Northern Territory. Field surveys were conducted between 2011 and 2013 to determine the prevalence of these potential causal factors in the occurrence of dieback. The identification of the fungal endophyte communities used

a combination of morphological and phylogenetic characters for final identification. All growth stages of M. pigra, from seedlings to adult plants, were tested for pathogenicity in the laboratory, glasshouse and field experimental conditions.

Objectives

a. To describe the “dieback phenomena” at a plant and population level using field assessment trials at Melaleuca Station, Northern Territory, Australia. b. To test whether endophytes, insect borers and abiotic conditions, play a role in the observed dieback phenomena. c. To determine the diversity and abundance of fungal endophytic communities associated with healthy and dieback-affected M. pigra populations. d. To identify the fungal endophytic communities recovered from the stems of M. pigra using morphological and phylogenetic characters, the nuclear ribosomal internal transcribed spacer (ITS) and partial translation elongation factor (EF1-α) gene regions. e. To test the pathogenicity of fungal endophytes against seedlings, juvenile and adult M. pigra plants by fulfilling Koch’s postulates.

1.3 Thesis outline

This thesis is structured into six chapters. Chapter 1 provides an introduction to the research topic, while Chapter 2 documents a review of relevant literature. This review highlights the ecology, biology and management of M. pigra. It also outlines the complexity in identifying the contributing causes of the dieback phenomenon in natural ecosystems.

Chapter 3 evaluates the role of endophytic fungi, phytophagous insects and abiotic factors, in the development of dieback in M. pigra, by conducting field studies and field specimen collection. This chapter also describes the “dieback phenomenon” at a plant and population level.

Chapter 4 focuses on the identification of the fungal endophytic communities of M. pigra, using a combination of morphological and phylogenetic characters. This chapter identifies and assesses abundance of fungal endophytes used in the analyses in Chapters 4 and 5. Selecting a subset of endophyte isolates, Chapter 5 examines the differences in their pathogenicity against seedlings and juvenile plants in a controlled environment, and adult plants in field conditions.

In Chapter 6, the main findings and key results of the thesis research are summarised, and suggestions are provided for future field ecology and laboratory studies relating to M. pigra dieback.

CHAPTER 2

Literature Review

2.1 The Host Plant Mimosa pigra L.

Mimosa pigra L. (Fabaceae), an introduced invasive weed in Australia, Southeast Asia and the Pacific Islands (Creager 1992), is an example of a plant species which has become widely distributed because of the unrestricted movement of living plant material around the world. Invasive plants can be deliberately or accidentally introduced into a new habitat by humans as a result of increasing trade and commerce (Mack et al. 2000), or sometimes as contaminants in imported agricultural products such as shipments of seed and grain (Huelma et al. 1996).

The genus Mimosa comprises approximately 530 species, with most of the diversity occurring at low and middle altitudes in the tropical regions of America (Barneby 1991; Du Puy 1997), the neotropics stretching from Mexico through Central America, the Antilles, Colombia, Peru and Brazil, and Northern Argentina (Heard et al. 1999). In its native range, M. pigra occurs on seasonally flooded lowlands and rarely grows more than two meters in height (Lonsdale & Abrecht 1989).

Mimosa pigra was first described by Linnaeus (1759). Its common names are: giant sensitive tree, catsclaw mimosa and bashful plant. It belongs to the family Mimosaceae, sub-family Mimosoideae (Fabaceae) (Heard 2004). In Australia, there are three Mimosa species that exist as exotic weeds. These species are M. pigra, M. invisa and M. pudica. M. pigra is commonly referred to as the “giant sensitive plant” because of the sensitivity of its leaves, which collapse when touched.

2.2 The Biology of Mimosa pigra L.

The M. pigra shrub is a tall, prickly, woody branched perennial that forms impenetrable thickets in paddy fields and along watercourses (Figure 2.1). The weed grows up to 6 meters in height. Miller (1991) reported that M. pigra can withstand the anaerobic conditions of inundation and flooded soils. It has low nutrient requirements and can consequently grow over a wide range of soil types, including nutrient poor sands, alluvial red and yellow earths, silty loams and heavy black cracking clays (Lonsdale 1992).

The shrub is multi-branched, with 5-10 mm long spines on the stems. Its branches are usually straight. The young stems are green but become woody and grey in colour as the plant matures. Each leaf comprises a prickly leaf stalk, 20-25 cm long. Attached to the leaf stalk are approximately 15 opposing primary segments about 5 cm long comprising 20-42 pairs of sessile, narrowly lanceolate leaflets per pinna that fold up when touched or injured. Flowers are round, fluffy, pink or mauve balls, about 1-2 cm in diameter. The heads are borne on stalks 2-3 cm long, with two in each leaf axil, while the corolla has four lobes with eight pink stamens. The seed pods are 3-8 cm long and contain one-seeded, bristled segments, which fall away from the pod leaving a skeletal outline. The major root system consists of a 1-2 m taproot and lateral roots that extend up to 3.5 m from the stem (Lonsdale 1992).

A . B

C D

Figure 2. 1 Mimosa pigra photographed at Melaleuca Station, Northern Territory: (A) main infestation, (B) adult plants, (C) mature pods, (D) flower

Under ideal conditions, plants begin flowering 6-8 months after germination. Lonsdale (1992) reported that in the Northern Territory, a typical production rate is about 9,000 seeds per plant per year, with an average of 21 seeds per pod. However, under optimum conditions, plants can produce up to 27,500 seeds per plant per year. The seed pod segments are covered with bristles that facilitate floating by surface tension, thus allowing the weed to spread rapidly along

river systems. The lifespan of the seed in the ground depends greatly on depth of burial and the soil type, and can be up to 23 years in sandy soils (Lonsdale & Abrecht 1989).

Another advantageous weedy characteristic of the plant is that, if chopped down, M. pigra will easily re-sprout from the stump. If M. pigra is burnt, the foliage may become desiccated and fall, but up to 90% of the mature plants and up to 50% of seedlings, may regrow (Wanichanantakul & Chinawong 1979).

2.3 Distribution of Mimosa pigra in Australia and its Impact on Agriculture

M. pigra was probably introduced into the Royal Darwin Botanic Gardens in the Northern Territory, Australia, prior to 1891. In the literature, there is no record to clarify whether the plant was imported purposefully as a specimen, or accidentally as a seed (Miller 1987). From the Royal Darwin Botanic Gardens it entered the Adelaide River System and has now spread extensively with major infestations also occurring along the Finniss, Mary and East Alligator Rivers. However, Cook and Dias (2006) documented records of deliberate introduction and comments on utility on the Australian invasive weeds introduction in Australia. They described that, although a tropical legume, M. pigra proved to be a robust forage plant in the Melbourne Botanic Gardens in 1881.

The land area affected by Mimosa pigra increased from 4,000 to 80,000 hectares between 1980 to 1989 (NT Government 1997 cited in (Walden et al. 2002). Reported causes of rapid M. pigra invasion after its initial introduction into the Northern Territory were a combination of heavy overgrazing of the native flora by feral water buffalo and the ability of the seed to spread rapidly during the wet season (Lonsdale 1988; Lonsdale & Abrecht 1989; Lonsdale 1992). M. pigra is now present in catchments and coastal floodplains of most of the major Top End river systems, and its range reaches from the Victoria River in the west (approximately 50 km from the Western Australia (WA) border), to the Phelp River in south-eastern Arnhem Land, and the Arafura swamp in the north-east (Figure 2.2 ). The size of infestations varies between the river systems, with the largest infestations being along the Adelaide, Mary and Finniss rivers, and in the Daly River/Port Keats Aboriginal Land Trust (Table 2.1). In February 2001, the first infestation outside of the Northern Territory was confirmed at Peter Faust Dam, approximately 25 km west of Proserpine, Queensland, latitude 20°24’4”S, longitude 148°34’50”E (Chopping 2004).

The dry season carrying capacity of the 80,000 ha of M. pigra infestation represents potential economic losses from $3 to 32 millions of dollars for the beef and buffalo industries. The profitability of a significant component of the Northern Territory’s $180 million annual pastoral industry has been affected by M. pigra. Also, the mud crab (Scylla serrata) industry, which is the most valuable wild-capture harvest with an estimated annual value of more than $10 million, and the barramundi (Lates calcarifer) fisheries (worth more than $5 million annually) may be threatened by the degradation of wetlands by M. pigra (Walden et al. 2002).

Table 2.1 Important wetlands infested with M. pigra in the Northern Territory (Source: Walden et.al 2000) Wetland system Total area (ha) Estimate of area infested with M. pigra (ha) Adelaide River floodplain system 134,800 30 000

Arafura Swamp 71,400 5

Blyth-Cadell floodplain & Boucaut 35,500 >1 Bay system

Daly-Reynolds floodplain estuary 159,300 17 system

Daly River middle reaches 1,650 300

Port Darwin 48,800 > 10

Finniss floodplain & Fogg Bay system 81,300 500

Mary River floodplain system 127,600 17,484

Moyle River floodplain 48,100 10,000

Murgenella-Cooper, East Alligator 81,500 1,000 floodplain system

Kakadu (South Alligator), West 190,000 4,000 Alligator and Wildman River systems, 10 Stray Ck, Phelp River, Scott Ck Total area 979,950 ~80,000

The dense monocultures of M. pigra not only adversely affect the native flora and fauna, but are also detrimental to public recreation, tourism and other land uses dependent on the natural features of the Top End environment (Walden et al. 2002). The Northern Territory’s $718

Figure 2.2 Map Showing the Northern Territory M. pigra Catchments (Source: Weed Management Branch, Northern Territory). million per year tourist industry has been affected to some degree, due to reduced access to fishing, hunting and scenic areas. A threat from M. pigra to conservation areas in Northern Australia, including the Kakadu National Park world heritage site (Northern Territory

Government 2003), has been acknowledged by the declaration of M. pigra as a “Weed of National Significance”. The invasion of M. pigra into wetland and conservation areas of the Northern Territory also causes a major reduction in biodiversity. On Aboriginal land, a serious consequence of this biological invasion is the damage being done to the wetlands. M. pigra invasion disrupts the traditional aboriginal way of life. In pastoral areas the weed competes with pasture plants, restricting the movement of livestock by blocking their access to water and interfering with mustering. Summary of the negative impacts brought about by M. pigra infestation and the type of management control system is shown in Table 2.2 and 2.3.

Table 2.2 Summary of the negative impacts of M. pigra in the Top End of the Northern Territory. Type of Impact Examples Environmental  Aggressive growth habit  Forms dense, impenetrable thickets, 3-6 m high  Establishes along river banks, encroaches into billabongs and out into drier floodplains

Economic Smothers pastures, reducing available grazing area, and affects stock mustering; reduces agricultural production. Social Affects accessibility to water for recreation purposes; disrupts aboriginal way of life Table 2.3 Summary of control methods used to control M. pigra in the Top End of the Northern Territory. Type of Biological Chemical Mechanical Physical infestation Small (few Not suitable Spot spraying Not suitable Hand grubbing (remove plants, small by hand with roots and burn plants) area) registered herbicide

Medium Release of Spot spraying Chaining, rolling, Follow-up control of (medium biological by hand with raking or back- seedlings – could density, control agents registered ploughing, then include physical medium total herbicide burning removal. area)

Large (many Step 1. Release Step 2: Aerial Step 3: Attack Follow-up control of plants, many of biological spraying with with chaining, seedlings- could include hectares) control agents registered rolling or raking. physical removal. herbicide Use fire to kill any regrowth and break seed dormancy

2.4 Mimosa pigra Risk and Control

In 2000, a risk assessment conducted by (Walden et al. 2000) identified the likelihood that the M. pigra invasion was not only restricted to the Northern Territory wetlands but was also a serious risk to other nationally and internationally important wetlands in Australia. Among the other wetlands identified as being at potential risk are those in the tropical and sub-tropical areas of Queensland and Western Australia. M. pigra is a difficult weed to control, and its seeds will survive in the soil for extended periods (Lonsdale & Abrecht 1989). Its control requires substantial resources over many years. Therefore, an integrated program using several weed management techniques is the most effective way to deal with the dense infestations of M. pigra (Northern Territory Government 2003). An integrated weed management approach includes prevention strategies, chemical, mechanical, ecological and biological control. Nevertheless, the assessment and determination of the characteristics of infestation (for example, age, density, location and time of the year) and the availability of funding and resources, will influence the most appropriate type of control method adopted. The M. pigra Strategic Plan published in 2000 reported that M. pigra control at that time was costing between $250- $350 per hectare. The methods used to control M. pigra in the Northern Territory are outlined below.

2.4.1 Physical and Mechanical Control

These methods are usually considered as temporary control options for large infestations of M. pigra, and are usually employed in combination with herbicide application and burning . Hand weeding, hand hoeing and bulldozing are the main strategies being used within this approach.

2.4.2 Chemical Control

Various herbicides have been tested on M. pigra in the Northern Territory, with different products being recommended, depending on the application situation, e.g. soil application, cut stump application, and stem injection or foliar application. Herbicides used include Tebuthurion, Fluroxypyr, Dicamba and Glyphosate. As part of the integrated weed management plan, there are three types of chemical application approaches for M. pigra control: (1) aerial application using helicopters or aeroplanes; (2) foliar spraying using vehicles, quad-bikes, or low ground pressure vehicles; and (3) soil applied herbicides. Access is difficult within bigger areas of infestation along river systems, such as the Finniss and Adelaide Rivers; the use of chemical herbicides can become environmentally undesirable

and/or logistically difficult. Weed control can become very expensive and limited (Hennecke 2002).

2.4.3 Biological Control

In their native habitats, plant species and natural enemies co-exist in a dynamic but stable balance (Egerton 1973; Hawksworth 1995). Introduced into a favourable, exotic environment, plants often increase their competitiveness and vigour, assisted by a lack of coevolved natural enemies, including insects, fungi, or nematodes (Keane & Crawley 2002). Biological control is defined as “the action of parasites, predators and pathogens in maintaining another organism’s density at a lower average than would occur in their absence” (DeBach 1964). According to McFadyen (1998), this definition can be interpreted to include three types of approaches for applied biological control agents:

(i) “Conservation” through the protection or maintenance of existing populations of biocontrol agents; (ii) “Augmentation” through regular action to increase the populations of biocontrol agents, either by periodic release or environmental manipulation; (iii) “Classical biological control” through the importation and release of exotic biological control agents, with the expectation that the agent will become established and further releases will not be necessary.

2.4.3.1 The Augmentative Approach The augmentative or inundative method is a technological approach when compared to the more ecological classical approaches to a weed problem (Watson 1991). Phytopathogenic microorganisms/microbial compounds or their natural products, are collectively known as bioherbicides (Hoagland et al. 2007). Bioherbicides as a biological control agent are applied in similar ways to chemical herbicides to control weeds. The active ingredient in a bioherbicide is a living microorganism and it is applied in inundative doses of propagules (Auld & Morin 1995). Most commonly, the microorganism used is a fungus and its propagules are spores or fragments of mycelium. In this case the bioherbicide is also referred to as a mycoherbicide. Mass production of potential biocontrol agents, formulation, shelf-life, optimum inoculum concentration, adequate conditions for fungus effectiveness against the weed and host range, require a series of studies for the successful adoption of this approach (Upadhyay et al. 2011; Barton 2012).

2.4.3.1.1 Role of Fungi in Integrated Weed Management

The development of science-based technologies for weed control using plant pathogens and other microorganisms has garnered interest and momentum since the 1970s when the principles of biological control were adapted for commercial purposes. DeVine and Collego were the first successful bioherbicides (Hoagland et al. 2007; Charudattan 2010). Broadly defined,

Table 2.4 Impact Status of pathogenic fungi following authorised introduction to Australia for the biological control of weeds Pathogen Weed Target Date of Introduction Impact Status Puccinia chondrillina Skeleton Weed Mid 1971 Low to high (Chondrilla juncea) (depending on weed forms) Phragmidium violaceum European blackberry(Rubus October1990 Low to high fruticocus agg.) Uromyces heliotropii Common heliotrope (Heliotrope January 1991 Nil europaeum) Puccinia abrupta Parthenium weed (Parthenium April 1991 Low Var: partheniicola hysterophorus) Puccinia cardui-pycnocephali Slender thistles (Cardus September 1993 Low pycnocephalus & C. tenuiflorus) Phloeospora mimosa-pigrae Mimosa (Giant sensitive plant) January 1995 Nil (Mimosa pigra) Diabole cubensis Mimosa (Giant sensitive plant) Early 1996 Nil (Mimosa pigra) Maravalia cryptostegiae Rubber-vine (Cryptostegia Late 1993 High grandiflora) Puccinia melampodii (= Puccinia Parthenium weed (Parthenium November 1999 Uncertain xanthii f.sp. parthenium- hysterophorus) hysterophorae Puccinia evadens Groundsel bush (Baccharis Late 1997 Uncertain halimifolia) Puccinia myrsiphylli Bridal creeper (Asparagus June 2000 High asparagoides) Prospodium tuberculatum Lantana (Lantana camara) October 2001 Uncertain Source: (Morin et al. 2006) Selection of pathogen agents in weed biological control: critical issues and peculiarities in relation to arthropod agents). bioherbicides are weed control products that are derived from living organisms, including any natural products they produce their growth that suppress weed populations (Bailey et al. 2010). The biological origins of most bioherbicides are microbial (bacteria, fungi, virus and nematodes) (Bailey 2004). Table 2.4 shows the impact status of pathogenic fungi following authorised introduction to Australia for biological control of weeds.

2.4.3.2 The Classical Biocontrol Approach The classical concept of biological control is defined as the intentional introduction of an exotic, usually co-evolved, biological control agent for permanent establishment and long-term

pest control (Coombs & Lisansky 1993). The goal of transferring natural enemies, which are usually insects or pathogens, from their native range to another area where the agent does not occur, is the establishment, self-perpetuation and natural dispersal of the natural enemy within the new target habitat without further manipulation. Classical biological control will not eradicate a weed but aims to achieve control of the weed population at a non-damaging or sub- economic level (Harley & Forno 1992).

2.5 Classical Biological Control of Mimosa pigra

In 1979, the Commonwealth Scientific International Research Organisation (CSIRO) collaborated with the Northern Territory Department of Primary Industry and Fisheries (DPIF) within the Northern Territory to study prospective biocontrol agents of M. pigra. The studies resulted in the release of 13 insects and two rust fungi over 30 years, presenting a partial success in release and establishment and the most obvious damage being to the stems of affected M. pigra (Heard 2012b). These biocontrol agents originated in tropical America, the native environment of M. pigra. Harley et al. (1995) found that M. pigra is attacked by more than 200 species of insect herbivores in its native range, while Evans et al. (1995) reported several obligate fungal pathogens which have the potential to act as biocontrol agents for M. pigra (Table 2.5).

2.5.1 Insect Biocontrol

A large collection of insect agents have been assessed but rejected as potential biocontrol agents for various reasons. There have been 15 agents released, of which 10 are known to have become established (Heard 2012a). The evaluation of the success of individual agents on M. pigra has been difficult to quantify although by 2003 it was clear that both Carmenta mimosa Eichlin & Passoa (Lepidoptera) and Neurostrota gunniella Busck (Lepidoptera) were having an impact on M. pigra (Heard & Paynter 2009). Both of these agents are stem borers; their larvae tunnel into the stems of infested M. pigra plants (Figure 2.3).

Table 2.5 Status of agents released for biological control of Mimosa pigra (in chronological

order of release in Australia) Agent Plant part attacked When Establishment and released impact Acanthoscelides Mature hard seeds 1983 Established, 0–10% seed puniceus (insect) kill

Acanthoscelides Mature hard seeds 1983 Initially established but quadridentatus (insect) did not persist

Chlamisus mimosa Leaves, stems 1985 Established, but causes (insect) no significant damage

Neurostrota gunniella Bores in pinnae and 1989 Established, very (insect) small stems abundant, impact uncertain

Carmenta mimosa Bores in large stems 1989 Established, very (insect) damaging

Coelocephalapion Flower buds 1992 Did not establish or did aculeatum(insect) not persist Coelocephalapion Leaves and flower buds 1994 Established, destroys pigrae (insect) 10% of flowers

Phloeospora mimosae- Leaves, stems and pods 1995 Did not establish or did pigrae (fungi) not persist Diabole cubensis Leaves 1996 Establishment confirmed (fungi) in 2011

Chalcodermus serripes Mature green seeds 1996 Establishment confirmed (insect) in 2008

Sibinia fastigiata Young green seeds 1997 No evidence of (insect) establishment

Malacorhinus Roots and leaves 2000 Established at several irregularis (insect) sites, impact unknown

Macaria pallidata Leaves 2002 Widespread (insect) establishment confirmed

Leuciris fimbriaria Leaves 2004 Limited establishment (insect) confirmed Nesaecrepida infuscate Roots and leaves 2007 Early signs of (insect) establishment Source: Heard, T.A. (2012) Mimosa pigra. In: Biological Control of Weeds in Australia (Eds. M Julien,R McFayden and J Cullen). Pp 378-397. CSIRO Publishing, Melbourne.

A B

Figure 2.3 Adult Carmenta mimosa (A); Adult Neurostrota gunniella (B) Photos from Weeds Branch, Northern Territory.

2.5.2 Fungal Biocontrol

Two fungal pathogens were released in 1995 and 1996 as classical biological control agents against M. pigra in the Northern Territory. These pathogens were the wet-season fungus Phloeospora mimosa-pigrae H.C Evans & Carrion (Coelomycete) and dry season rust Diabole cubensis (Arthur & J.R. Johnst.) (Table 2.2). These host specific pathogens are thought to have co-evolved with Mimosa pigra in the Neotropics (Evans et al. 1995) and, that in their native range, they are capable of causing leaf spots, leaf chlorosis and premature leaf fall.

Neither of the above agents has had an impact on M. pigra in Australia. A review of the post release studies (Hennecke 2006) showed that D. cubensis failed to establish in the Northern Territory due to: (1) premature leaf drop before the rust could complete its disease cycle, this leaf drop being attributed to inherent resistance to the rust and attack by insects: (2) the more rapid increase in the density of the newly developed canopy after inoculation, and a long lag period for disease development and disease spread and, (3) a possible change in the microclimate around infected leaves caused by the dense leaf canopy that prevents disease spread by reducing air (and therefore spore) movement. In addition, Hennecke (2006) identified a need for detailed ecological studies to fill the gaps in our understanding of the ecology of M. pigra in Australia. However, by 2011 D. cubensis had become established in Australia, having been seen for the first time on the Finniss River, Northern Territory (Burrows et al. 2012; Heard 2012b).

In the case of Phleoespora mimosa-pigrae, Hennecke (2002) reported that the pycnidiospores of P. mimosa-pigrae were not able to spread the short-distances to neighbouring plant tissue by rain splash. Run-off also has no significant impact on leaf growth occurring below the diseased area of plants. The teleomorph of P. mimosa-pigrae did not develop in Australia, thereby preventing the long distance spread of the fungal pathogen by ascospores, and reducing its ability to survive the non-active period of the dry season. This pathogen appears not to have established in Australia.

In classical weed biocontrol, limitations to the successful use of fungal pathogens and arthropods can arise due to factors such as, differences in the climatic and ecological conditions between native and introduced ranges of alien weeds. This can affect the establishment, survival and propagation of a particular classical agent (Harley & Forno 1992). The need to undertake and pass expensive and rigorous host-specificity tests is also a limitation that classical weed biocontrol programs face.

2.6 Dieback

2.6.1 Dieback Definition

Plants which appear to be suffering from ill health and death of their growing points are characterised as experiencing “dieback”. The term ‘dieback’ has been used to define many perennial plant disease syndromes (Houston 1992; Jurskis 2005). Australian use of the term usually refers to protracted malfunction in stands of trees due to the persistent action of damaging factors (Podger 1981). Another definition of ‘dieback; from the dictionary of forestry states that “it is a progressive dying from the extremity of any part of the plant which may or may not result in the death of the entire plant” (Ostry et al. 2011).

2.6.2 Dieback Causes

Dieback, an important phenomenon in many crops and natural forests worldwide (Brouwers et al. 2013; Lygis et al. 2014), is caused by multiple factors, often interacting, that can be complicated and difficult to unravel. The search for primary causal mechanisms is complicated by the fact that dieback events arise from a complex of interacting biotic and abiotic factors (Muellerdombois 1987; Manion & Lachance 1992). For example, from a pathological viewpoint, dieback can always be considered as having an association with disease symptoms (Arif et al. 2013; Gross et al. 2014). From an ecological viewpoint, however, dieback is

defined as a natural phenomenon (Jeltsch & Wissel 1994). Moreover, other studies discuss dieback as being a result of the interaction between biotic factors and stressful environmental factors (Desprez-Loustau et al. 2006; Raffa et al. 2008; Bansal et al. 2013).

An hypothesis relating to the causes of plant death was developed by Manion and Lachance (1992). To address the issue of potential complex causes involving a variety of etiological agents, including biotic and abiotic factors, (Manion & Lachance 1992) illustrated the concept using a decline spiral (Figure 2.4). The decline spiral starts with three tiered inward rings, corresponding to the predisposing, inciting and contributing stress factors, respectively. Decline is induced by the sequential occurrence of one stressing factor from each ring, the factors within each ring being interchangeable. The spiral model has been widely used to analyse a number of decline situations. For example, the blackberry decline of Rubus anglocandicans in Western Australia indicates flooding to be a predisposing factor, the lack of genetic potential as an inciting factor, and a leaf rust (Phragmidium violaceum) as a contributing factor (Aghighi et al. 2014). In oak decline, drought has very often been considered as an inciting factor, with a root pathogen, such as Armillaria, being a contributing factor. The role of pathogens as predisposing or triggering agents of decline diseases has also been recognised, especially powdery mildew and Phytophthora spp. (Wargo 1996).

2.6.3 Mimosa pigra Dieback

Widespread dieback of M. pigra has been reported in the Adelaide River floodplains of the Northern Territory since at least 1988 (Miller 1988). Dieback in M. pigra is characterised by stem death, top-kill, defoliation, re-shooting from the base, and partial mortality of individual plants (Figure 2.5). The disease has continued to be observed in the Adelaide, Mary, Finniss and Daly River catchments in the Northern Territory (Natasha Burrows, Weed Sciences Branch of the Northern Territory Department of Land and Resource Management pers comm). A survey of M. pigra stands conducted in 2012 provided evidence of population (stand) dieback at 52% of the sites assessed in the Daly River catchment, and 20% of sites in the Finniss River catchment (DLRM unpublished data). Large areas of dieback (>10 hectares) are readily visible in the Adelaide, Daly and Mary River catchments (L. Elliott, pers. comm.).

Genetic potential

Age Viruses

(Interacting with) Old age Urban Environment Defoliating

insects Air

Canker Pollution

Soil Air pollution fungi Nematodes compaction Excavation Virus DEAD Wood and bark Saprobic

Boring decay Frost Poor Climate insects Armillaria fertility change Root rot Drought

Excessive Poor soil salt Outside climatically drainage

Adapted range

Low soil

Moisture Salt Holding capacity

Figure 2.4 Decline Disease Spiral (Manion, 1992).

Figure 2.5 Mimosa pigra population dieback at Melaleuca Station, characterised by stem death but with the presence of green, young reshoots coming from the base and main stems

2.6.4 Dieback in other Weeds of Australia

Natural dieback phenomenon is also reported in other Weeds of National Significance (WONS) species across Australia. In the temperate environment of the south-west of Western Australia (SWWA), blackberry (Rubus anglocacandicans) dying patches have been observed on the banks of Donnelly and Warren Rivers (Yeoh et al. 2006). Prickly acacia (Vachiella nilotica) dieback in the tropical north-western Queensland has also reported since 1970 (March 2004). Parkinsonia aculeata in northern Australia exhibits naturally occurring dieback in large areas (Diplock et al. 2006). Dieback has also been reported on athel pine (Tamarix aphylla) in the Finke River in central Australia, and on bitou bush (Chrysanthemoides monilifera) in eastern Australia.

2.7 Impact Evaluation of Biological Control Programs

2.7.1 The Importance of Impact Evaluation

The impact evaluation approach of evaluating biological control as a method of controlling invasive weeds is very important. One of the advantages of impact evaluation is that it can reveal the effectiveness of a biological control program and thus can improve the program and its practical orientation (Thomas & Reid 2007). On the other hand, this approach can be challenging, given that many biocontrol programs are focused mainly on the establishment and prevalence of the biocontrol agents, with only limited studies being done on the impact of biocontrol at the individual or population plant level (van Klinken & Raghu 2006; Carson et al. 2008).

In Australia, more than $65 million has been spent on biological control programs (Page & Lacey 2006). Most of this budget has been used for programs relating to agent exploration, importation, rearing, release and establishment. Very little has been allotted to the assessment of the effectiveness of the agents released in the field (McClay 1995; McFadyen 1998; McEvoy & Coombs 1999). Quantitative evidence of the efficiency of the program is important for improving the effectiveness of future programs for other weeds.

2.7.2 Challenges of Impact Evaluation of Biocontrol Programs

The evaluation of the impact of biological control programs is sometimes poorly quantified. A challenge of impact evaluation for biocontrol agents is that there can be confounding factors (both biotic and abiotic) relating to the effects of these agents on plant abundance, distribution and population growth (Sims-Chilton et al. 2010).There has been a little follow-up evaluation of biocontrol programs in the past, while the impact of biological agents is rarely quantified (Sims-Chilton et al. 2009). There is a need to consider the “multiple stresses” hypothesis to gain a fuller picture of the effectiveness of implemented biocontrol programs.

2.7.3 Different Methods of Evaluating the Impact or Causality of Biocontrol Programs

There are different approach options for the evaluation of the impact of a biocontrol agent (Blossey & Skinner 2000; Morin et al. 2009). These approaches can involve manipulative experiments (in the laboratory, glasshouse and field) (Liu et al. 2002), demographic modelling (Buckley et al. 2004), and pre and post evaluation studies (van Klinken & Raghu 2006). These approaches each provide unique information but also have specific limitations. For example,

the advantage of working in a controlled environment enables the application of single factors in order to isolate their effects (Campbell et al. 2009; Meisner et al. 2014).

A simplified, controlled environment can reduce “unwanted noise” and allow the identification of the signal, test predictions, or allow the study of mechanisms that would not be possible in the field. In contrast, working in controlled environments is often unrealistic. Controlled studies should be followed by field studies and “natural experiments”, to learn whether the laboratory results are realistic and whether they also be related to larger spatial and temporal scales (Underwood 2009).

Another method of evaluation is through population modelling. It is a cost-effective tool for predicting the establishment and spread of ‘invasions’ before they actually do much harm. Climate based distribution tools, for example CLIMEX (Sutherst & Maywald 1985), help predict a species’ potential distribution; they rely on species distribution and abundance data to fit the models (Pearson & Dawson 2003; van Klinken et al. 2009). A disadvantage of demographic modelling is that it requires quantitative data on life history and population dynamics of the target invasive plant, as well as quantitative data on the impact of potential biological agents (Davis et al. 2006). Such data are not routinely available and past biological introductions have been largely based on informed expert opinions.

2.7.4 Evaluation of Mimosa pigra Biocontrol Programs

Extensive research has been carried out on the monitoring and evaluation of classical biocontrol agents, particularly on insect agents. This research is summarised in Table 2.6. Different research methods have been used. Manipulative experiments have been conducted in shade houses or in the field (Steinbauer 1998; Paynter & Hennecke 2001; Paynter 2006). A modelling study was conducted by (Buckley et al. 2004), while other studies have focused on the post- release abundance of the established agents (Steinbauer et al. 2000; Ostermeyer & Grace 2007).

These evaluation studies found a significant impact of the insects agents released. However, these studies have also had some drawbacks. The manipulative and post-release evaluation studies often concentrated on the effect of a single agent on only one plant stage at one sample time and tell us little about dieback. In addition, most of the evaluation studies on M. pigra have not taken into account other biotic and abiotic factors that may have contributed to changes in the plants vigour. There is a need to evaluate what role other biotic factors play (for

example, endophyes/pathogens and native borers), as well as environmental factors which can potentially contribute to plant decline.

2.8 Possible Role of Biotic Factors in the Development of Dieback in M. pigra

The fact that population (stand) dieback in M. pigra occurs in multiple catchments in the Northern Territory and over large areas, makes the dieback phenomenon of considerable ecological importance. Although extensive research has been conducted on the evaluation of the biological agents in M. pigra in relation to decreasing plant health, there is a need to consider what role of other factors play in the observed dieback phenomenon. Possible factors that may play a role in M. pigra dieback include the following:

2.8.1 Endophytes

The role of endophytes as predisposing or triggering agents in the decline of some diseases has been recognised (Manion & Lachance 1992). The fungus Botryotinia (Lasiodiplodia) theobromae (Pat.) Griff & Maubl. (Botryosphaeriaceae), a reported endophyte of many multiple fruits and woody plants, has previously been isolated from typical dieback symptoms: reddish- brown lesions originating from leaf axils, yellowing of the leaves and subsequent stem death in M. pigra on the floodplains of the Adelaide River catchment (Wilson & Pitkethley 1992). No further studies have so far been conducted to examine the role of endophytes in the development of M. pigra dieback.

Table 2.6 Evaluation studies conducted on M. pigra in the Northern Territory (compiled from the different published evaluation studies on M. pigra). Type of evaluation Method used for Type of Reference study evaluation specimen 1. Manipulative Shade house Seedlings Steinbauer, 1998 experiment studies 2. Post-release Field (malaise M. pigra adult Steinbauer et al, 2000 evaluation traps) 3. Post-release Surveys Insect Osternmeyer and Grace evaluation abundance on 2000 field site 4. Manipulative Controlled (pot seedlings Paynter, 2001 experiment experiment) 5. Population Analytical models Life history of Buckley, et al, 2004 modelling M. pigra 6. Manipulative Field Seed banks Barratt et al, 2004 7. Comparing Field assessment Agent Paynter, 2005 historical and abundance and contemporary litter fall data 8. Manipulative Structured field Litter fall on 3 Paynter, 2006 experiment survey different sites 9. Manipulative Controlled Seedlings and McIntyre, 2007 experiment experiment larvae

2.8.1.1 Endophyte Definition The term “endophyte” is derived from two Greek words, “endon” meaning within, and “phyton” meaning plant (Petrini et al. 1992). Translated literally, an endophyte is an organism that lives inside a plant. It has also been proposed that endophytes not only be fungi, but also bacteria that reside in plant tissues without causing disease. Endophytes have been isolated from the Ascomycotina, Basidiomycotina, and Deuteromycotina fungal groups, and also from the Oomycetes (Redlin & Carris 1996; Arnold et al. 2001). They are distinguished from “mycorrizal fungi” by virtue of the location where the fungus resides in the plant tissues (Wilson 1995). They live entirely inside the plant tissues but cause no apparent disease symptoms, whereas the hyphae of mychorrizal fungi emanate into the soil surrounding infected roots (Nicolson & Johnston 1979; Addy et al. 2005).

2.8.1.2 Role of Endophytes in their Host Endophytes have one or more of a variety of interactions with their host plant. Some fungi are widespread and found on many different plant species Davis & Shaw (2008), while others are highly specific to single hosts in a single environment. Further, a diverse range of interactions

between plants and fungi have been found. Evidence that endophytes have developed symbiotic relationships with plants is examined in the studies of (Wilkinson et al. 2000; Rodriguez et al. 2004). For example, food is the direct and prime benefit that endophytic fungi derive from host plants, since fungi are heterotrophs relying on other organisms for food such as carbohydrates. A number of endophytes produce compounds that are active as antagonists to plant pathogens and herbivores. For instance, alkaloids, which can be toxic to livestock as well as to insect pests, are produced by endophytic fungi of the family Clavicipitaceae (Strobel & Daisy 2003; Strobel et al. 2004). Taxol is a compound isolated from endophytes of a yew tree (Taxus sp.) and other plant species residing in damp environments. It is reported that taxol may protect those host plants from root pathogenic fungi, such as Pythium and Phytophthora (Strobel et al. 1996; Strobel 2003).

2.8.1.3 Opportunistic Endophytes Fungal endophytes associated with woody weeds and plants are especially diverse. The Botryosphaeriales is an important group of fungi due to the ecological and economic significance of many of its species. All species are plant associated, and many are classified as pathogens, known to cause disease on a wide range of ecologically and economically important plants (Mehl et al 2012). Some species are also known to cause opportunistic infections in woody weeds. For example, the fungus Botryotinia theobromae (Lasiodiplodia) theobromae caused dieback to M. pigra after the tunnelling of a phytophagous insect, Neurostrota gunniella on M. pigra stem tips (Wilson & Pitkethley 1992). Furthermore, endophytic species of Diaporthe and Pestalotiopsis and Colletotrichum have also been reported to cause endophytic infections in plants (Mejía et al. 2008; Maharachchikumbura et al. 2011; Gomes et al. 2013). These species are common phytopathogens that cause a variety of diseases when plant is under stressful conditions. Table 2.7 shows the common and major and common species of woody fungal endophyte.

Table 2.7 Common and major species of woody fungal endophytes. Family Genus Species Country of Study Host Reference Botryosphaeriaceae Lasiodiplodia L. theobromae North-west of Adansonia Slippers et al. Australia gregorii 2010 Pterocarpus Mehl et al. 2011 angolensis South Africa Acacia karroo Jami et al. 2014 Botryosphaeria B. dothidea Uruguay Eucalyptus Perez et al. 2010

Lasiodiplodia L. pseudotheobromae South Africa Terminalia spp. Begoude et al. L. thebromae 2011 L. parva Neofusicoccum N. australe Western Australia Agonis flexuosa Dakin et al. 2010 N. australe Western Australia Eucalyptus Slippers et al. 2009

N. luteum California grapevines Urbez-Torres et N. parvum al. 2009

Diaporthaceae Diaporthe Phomopsis sp. China Pistachio Chen et al. 2011 Amphisphaeriaceae Pestalotia P. hainanensis China Mangifera indica Liu et al. 2006 P. jester Wei et al. 2007 P. kunmingensis P. pallidotheae

2.8.1.4 Patterns of Endophyte Mutualism Transmission

Two patterns of endophytic mutualism are constitutive mutualism and inducible mutualism. Constitutive mutualism involves the transmission of endophytes from one generation to the next via the germ line. Constitutive endophytes infect host ovules and are carried through the seeds of the host plant (Tintjer et al. 2008). This type of transmission is sometimes called vertical transmission in the literature (Saikkonen et al. 2004; Afkhami & Rudgers 2008). These endophytes develop a systemic infection throughout the aerial parts of the plant, with considerable mycelium biomass. Fungal toxins, giving immediate and direct benefits to the endophytes’ host, are produced (Selosse & Schardl 2007). Grass endophytes are an example of constitutive mutualism. This is described as an inducible mutualism or horizontal transmission, and involves an infection from one host to another by spores transmitted through the air or in water (Rodriguez et al. 2009; Su et al. 2010; Oberhofer & Leuchtmann 2014).

2.8.1.5 Taxonomy and Identification of Endophytes

Fungi isolated from surface-sterilised plant tissues are primarily Ascomycetes or the asexual stage (deuteromycetes), with few Basidiomycetes (Arnold et al. 2000; Gazis et al. 2012). Some oomyctes have also been isolated as endophytes (Ploch & Thines 2011).

2.8.1.5.1 Classical Approach of Identification

For centuries, the classification and identification of fungi have been mainly based on observable characteristics and the fungal life cycle (Bary 1887), a method known as the phenotypic approach. The types of conidia (asexual spores) and the process involved in conidium formation are considered the most important characteristics for the identification and classification of fungi. Unfortunately, some fungi have a limited number of morphological characteristics that can be used for taxonomic purposes. For example, sterile fungi show only the vegetative phase; that is, they do not sporulate and only hyphal elements or other nonspecific structures, such as chlamydospores, sclerotia or hyphae, are observed (Alexopoulos 1996).

Morphological and physiological characteristics of fungi are often dependent on cultural conditions. Ascomycetes have a high degree of plasticity when grown on artificial media (Slepecky & Starmer 2009). Thus classifying fungi using morphological characters is problematic when a single very plastic characteristic is used. However, one of the advantages of using morphological characteristics is that these are the easiest traits to use for identification of fungi. Morphological identification is still a common procedure practiced worldwide. It is an inexpensive and quick method that requires only simple guidelines and laboratory equipment.

2.8.1.5.2 Molecular Approach of Identification

The increasing availability of nucleotide data is helping mycologists to review the evolutionary relationships among fungi. Informative sequences come from a single copy region of genes that perform the same function in all taxa, and evolve at a rate fast enough to separate the taxa into monophyletic groups. Ribosomal RNA (rRNA) genes from the nuclear and mitochondrial genomes, cytochrome oxidase genes and some protein synthesis elongation factors, are the genes that show evolutionary relationships among fungal species (Guarro et al. 1999). Although the internal transcribed sequences (ITS) of rDNA were recently thought to be suitable as “Barcodes” for fungal classification (Schoch et al. 2012) it is now apparent that multigenic sequencing is required to discriminate between species in many fungal families, for example, the Botryosphaeriaceae (Liu et al. 2012; Phillips et al. 2013).

The rDNA cluster is comprised of three structural regions, the 5.8S, 18S rRNA and 28S rRNA genes (Figure 2.6) (Mitchell et al. 1995). These three regions are transcribed into RNA molecules that form part of a 40S ribosome. Internal spacers (ITS) sitting between these three genes are transcribed and spliced out during post transcriptional processing. Each genomic region has an informative sequence for delineating taxa at a certain taxonomic level. The rDNA genes are frequently used for phylogenetic studies, because these genes which are highly conserved, are found in all organisms with a common function of evolutionary origin, and it does not encode proteins (Gray et al. 1984). Identification using a molecular approach is fast and more reliable. However, a potential disadvantage of using the ITS sequencing method may come from the interspecies variability in DNA extraction, efficiency, primer binding, PCR amplification kinetics and cloning efficiency (Buchan et al. 2002).

At present, many studies are focusing on unravelling species complexes by using a combination of the morphological species concept with the phylogenetic species concept provided by comparing multiple gene genealogies. For example, (Taylor et al. 2009; Phillips et al. 2013) used 5 gene regions to identify taxa members of the Botryosphaeriaceae. The use of molecular phylogeny is now solving many taxonomic problems. Molecular phylogenetic data may be particularly helpful in supporting or disproving fungal taxonomic arrangements.

2.8.2 Fungal Pathogens

The role of fungal pathogens as a contributing factor to plant decline and dieback is being recognised. For example, the family Botryosphaeriaceae are reported to cause dieback in woody trees and shrubs worldwide (Perez et al. 2010; Shah et al. 2010; Begoude et al. 2011). The genera Diaporthe and Pestalotia/Pestalotiopsis are classified as fungal pathogens globally causing canker and dieback in

legumes, multiple fruits and woody plants (Espinoza et al. 2008; Diogo et al. 2010; Urbez-Torres et al. 2013).

PFPRIM-R4 PFPRIM-F3 ITS86 ITS1 ITS4 ITS1 ITS2 18S rRNA 5.8S rRNA 28S rRNA

500-800bp 401bp

280bp Figure 2.6 The nuclear ribosomal RNA gene cluster is comprised of a Non-Transcribed Spacer (NTS), External Transcribed Spacer (ETS), 18S rRNA, Internal Transcribed Spacer 1 (ITS1), 5.8S, Internal Transcribed Spacer 2 (ITS2) and 28S rRNA genes. IGS stands for the intergenic spacer region (adapted from Mitchell, 1995).

2.8.3 Pathogenicity and Virulence Defined

The terms pathogenicity and virulence are used in many scientific disciplines including ecology, plant pathology and epidemiology. The definition of these terms vary among and within disciplines. The definition used by Tanada and Kaya (1993) was used in this study. Pathogenicity is defined as the disease-producing ability of a microorganism while virulence is the ability of the microorganism to invade and cause injury to the host. It is important to note that virulence and pathogenicity are considered properties of the pathogen, but we measure them by recording the response of the host in a particular environment. A pathogen that is highly virulent in one host may have low virulence in another, or may be less virulent in different abiotic conditions (Casadevall 2001).

2.8.4 Native Borers

Apart from fungal pathogens, native borers may also play a role in the dieback process. Surveys of three native species of Mesoclanies (Diptera:Tephritidae) flies in South Africa concluded that native insects can reduce seed production and reduce reproductive (ovule) development in Chrysanthemoides (Asteraceae) (Edwards & Brown 1997). Studies have also shown that native borers of Northern Australia, Platyompsis humeralis and Maroga sp., negatively affect the plant population dynamics of Acacia auriculiformis (Montagu 1999; Steinbauer et al. 2000).

2.9 Possible Role of Environmental Factors in the Development of Dieback in M. pigra

2.9.1 Floodplain Habitat

The floodplain landscape is a complex, dynamic mosaic shaped by interactions between disturbance and succession (Kalliola et al. 1991; Malanson 1993). The floodplains of the Northern Territory are the largest of their kind in Australia and cover an area of approximately 10,000 square kilometres, from the Moyle River in the west to eastern Arnhem Land. (Australian Government Department of Sustainability 2012). A tropical wet-dry (monsoonal) climate includes high temperatures and seasonally high precipitation (wet season) followed by seasonal drought and high evaporation (dry season). The climate and gentle topography of much of the floodplains creates seasonal rather than permanent wetlands. Most of the 1,709 mm average yearly rainfall occurs between late October and early April, even though the actual rainfall may vary considerably year-to-year (McDonald & McAlpine 1991). The average maximum temperature in the region is above 30°C, with only small annual fluctuations. During the wet season the relative humidity is between 70 to 100%, considerably higher than in the dry season, while higher evaporation occurs in the dry season.

The geomorphological and hydrographic features of the Northern Territory river system are detailed in Woodroffe (1995). Flooding of the river is strongly regulated by the monsoonal climate. During the wet season, groundwater storage accumulates with maximum runoff occurring late in the wet season when freshwater dominates the system. With the cessation of rainfall, seawater penetration with tidal behaviour develops, becomes most extensive in the late dry season (July through November). The river is subject to semi-diurnal tidal behaviour, with maximum spring tides ranging from 6 to 8 m. As the tidal length of the river is eight times greater than the flood and ebb excursion of water during a tidal cycle Ball (1998), the upstream propagation of salt water after the wet season is due to turbulent diffusion. By this means, saline water penetrates as far as 80 km upstream from the river mouth by the late dry season (Chappell & Woodroffe 1985).

2.9.2 Temperature

High temperature can predispose plants to other biotic stressors like fungal infection, leading to rapid and severe dieback (Nepstad et al. 2008). Temperature is one of the environmental factors that play a critical role in the germination of fungal spores. The knowledge of various spore reactions to temperature has been shown to significantly improve the management options in many crops (Michailides et al. 1992; Miller et al. 2003). Pathogens differ in their preference for higher or lower temperatures. Some fungi grow much faster than others at lower temperatures, and there may be 30

significant differences among ecotypes of the same fungus (Agrios 2005). Temperature affects the number of spores formed in a plant unit area. Pre-release evaluation of Phloeospora mimosa-pigrae, a holomorph of Sphaerulina mimosa-pigrae, a widespread pathogen specific to M. pigra in Central and South America, determined that the optimum conditions required involves temperatures of 20- 25°C combined with a relative humidity of 70 to 100%, for a period of 24 hours (Hennecke 2006). The obligate rust Diabole cubensis showed disease symptoms in the form of chlorotic leaf-spots developed after a symptomless period of approximately 2 weeks, within a temperature range of 21 to 23°C at 60 to 80% relative humidity.

2.9.3 Soil Acidity

Soil acidity (pH) restricts the availability of essential macro-nutrients, leading to root mortality and eventually dieback (Houle et al. 2007; Watmough 2010). Studies such as those of (Burke & Raynal 1998; Duchesne et al. 2002) have indicated that soil acidification restricts the availability of plant macronutrients, leading to root mortality, nutrient deficiency and eventually tree dieback.

2.9.4 Waterlogging and Salinity

Waterlogging is defined as a condition of the soil where excess water limits gas diffusion. An investigation of the causal agents in the decline and mortality of Arizona cypress revealed that water salinity leads to the speeding up and boosting of the dieback process (Khakdaman et al. 2008). A direct relationship between salinity and dieback was found in a study of sites of Acacia xanthophloea in Tanzania. The results of the study showed that electrical conductivity and water soluble Na+ , K+ - 2- , Cl , and SO4 , were significantly higher in dieback sites than the healthy sites (Mills 2006). Salt accumulation is the major cause of floodplain tree dieback, reducing the environmental, ecological and social values of the floodplain environment (Holland et al. 2000). Excessive flooding leads to anoxia and root death, and branch abscission maintains a constant ratio between the living and below ground and above ground biomass (Palik et al. 2011).

In summary, this literature review outlines the contrasting theories relating to the possible causes of dieback on floodplain habitats. Both biotic and abiotic factors are possible key players in the development dieback in singly or in combination. Moreover, as illustrated by the disease spiral hypothesis developed by Manion and Lachance (1992), dieback might be an interaction of the three factors: “predisposing, inciting and contributing”. Hence, this thesis evaluates the relative roles of biotic factors (insect borers and fungal endophytes) and abiotic factors (soil pH and salinity), on the

development of dieback in M. pigra, with more focus on fungal endophytes because very scarce information is available on this aspect. Having an accurate understanding of the underlying causes of the dieback phenomenon in M. pigra is important in order to provide an understanding of the ecology and biology of this noxious weed in the Northern Territory.

CHAPTER 3

Evaluating the relative roles of endophytic fungi, phytophagous insects, and the environment, in the development of dieback in Mimosa pigra.

3.1 Abstract

Invasive plants have a detrimental influence on ecosystems globally with the economic impact estimated at millions of dollars per invasive species each year. Biological control has long been used as a management tool for invasive plants, as it is considered a long- term and cost-effective strategy. However, the impact of biological control agents is not often quantified. There are very few studies which have examined the impact of the agents on the plant population dynamics of the invader, particularly once the agents have been established for a long time. Mimosa pigra L. (Fabaceae), a noxious weed which has invaded the wetlands in Northern Australia, is also a host of sporadic dieback phenomenon, first observed in 1988 on the Adelaide River floodplains in the Northern Territory. This chapter evaluates the role of endophytic fungi, phytophagous insects and abiotic factors, in the development of dieback in Mimosa pigra in Australia’s Top End. Ground surveys were conducted from 2011 to 2013 at Melaleuca Station in the Mary River region of the Northern Territory. The results of the survey indicate that dieback in M. pigra is highly dynamic but fluctuating in its severity, resulting in partial mortality, reduced plant vigour which is reflected by foliage loss, the death of terminal stems and death of main stems. Insect borers and fungal endophytes were ubiquitous across all plots covered by the survey, and over time, and showed no relationship with the M. pigra population dynamics in the area covered by the survey. Soil pH and salinity were also not observed to have a biologically significant influence on M. pigra population dynamics. It is not clear what the cause of the dieback is. On the basis of the results of this study, a single causal agent was not able to be identified. It is likely that dieback is the product of an interaction among biotic and abiotic factors, in the floodplain habitat.

3.2 Introduction

Mimosa pigra L, is a woody perennial shrub native to tropical America, which is listed in the Global Invasive Species Database as one of the One Hundred of the World’s Worst Invasive Alien Species (Global Invasive Species Database 2015). It was first noticed in Australia in 1952 in the Adelaide River System of the Northern Territory Miller and Lonsdale (1987) and is now widespread in the Top End of the Northern Territory. M. pigra has been a subject of biocontrol programs for over 30 years, but still continues to be a serious problem. It is also a host of dieback phenomenon, which was first observed in the Adelaide River floodplain in 1988 (Miller 1988). 33

Dieback is an important phenomenon, not only in many crops and natural forests worldwide, (Jeltsch & Wissel 1994; Marques et al. 2012; Matusick et al. 2013) but also seen in many invasive trees and shrubs across Australia (Diplock et al. 2006; Haque et al. 2013; Aghighi et al. 2014; Laurance et al. 2011). This phenomenon is caused by many interacting factors. An understanding of the primary causal mechanisms of dieback events is complicated by the complex interactions of biotic and abiotic factors (Desprez-Loustau et al. 2006) and many of the symptoms observed are not similar to those caused by single agents (Elmer 2014). There is therefore a need to consider the “multiple stress hypothesis” (Lotze & Worm 2002; Niinemets 2010) in evaluating and describing this phenomenon.

Dieback of M. pigra was first reported on the Adelaide River floodplains in 1988 (Miller 1988), and since then has been reported in other large infestations in the Top End. Descriptions of the phenomenon remain anecdotal, but it stem death, top-kill, defoliation, re-shooting from the base and mortality has been observed. Large areas of dieback (>10 hectares) are readily visible in the Adelaide, Daly and Mary River catchments (L. Elliott, pers. Comm.). A survey of M. pigra stands conducted in 2012, showed evidence of stand dieback at 52% of the sites measured in the Daly River catchment (DLRM unpublished data).

Classical biological control is defined as the introduction or manipulation of natural living organisms, such as herbivores and pathogens, to control pests. Its aim is to use the weed’s co-evolved natural enemies to achieve manageable weed control (Gassmann et al. 2006). Classical biological control (CBC) is a proven powerful management tool, which can provide great potential benefits.

In 1979, a collaborative initiative between CSIRO and the Northern Territory DPIF commenced a study of the biocontrol of M. pigra. This initiative resulted in the successful release and establishment of 13 insects and 2 rust fungi as biological control agents (Heard 2012). The most obvious damage occurs in the stem of the affected plants. While the impact of individual agents has been difficult to prove, a study in 2003 found that both Carmenta mimosa Eichlin and Passoa (Lepidoptera) and Neurostrota gunniella Busck (Lepidoptera) had a major impact on M. pigra (Heard & Paynter 2009). This resulted in reduced seed production, seedling regeneration and soil seed banks. These agents are stem borers, whose larvae tunnel into the stems of the infested plants. Two fungi, Diabole cubensis (Arthur & J.R. Johnston) Arthur and Phloeospora mimosa-pigrae H.C Evans & Carrion were also released. The microcyclic rust D. cubensis has established but remains rare Burrows et al. (2012), whereas P. mimosa-pigrae appears not to have established (Hennecke 2006).

In the case of observed dieback in M. pigra, three agents were released prior to the first observation of dieback symptoms. These were the seed feeders Acanthoscelides puniceus and Acanthoscelides

quadridentatus released in 1983 and the leaf and stem feeder Chlamisus mimosae released in 1985 (Wilson & Flanagan 1991). Two of these agents established but caused no significant damage on M. pigra by 1988 (Steinbauer et al. 2000). The establishment of the stem-boring moth Neurostrota gunniella, first released in 1988 was linked to a widespread dieback in the Adelaide river floodplains (Wilson & Pitkethley 1992). During the subsequent dry season, dieback was observed in M. pigra shrubs where in. N.gunniella was established and abundant.

The impact evaluation approach of evaluating biological control as a method for controlling invasive weeds is very important but challenging. One of the major challenges of this approach is that other factors (both biotic and abiotic) may be overlooked. As Smith and De Bach (1942) (page 845) state, “critics contend, quite justifiably, that a reduction in the population following the establishment of one or more of its insect enemies, is not necessarily proof that the enemy was the cause of the reduction…” Population modelling studies of an invasive weed, Baccharis hamifolia, which has been a subject of biocontrol programs since 1969, has been able to demonstrate that climate has had an influence on the reduction in population size of this weed (Sims-Chilton et al. 2010).

While classical biological control may be responsible for M. pigra dieback, there are other factors which can be potential drivers of M. pigra dieback. The role of pathogens as predisposing or triggering agents in the decline of diseases has been recognised (Muellerdombois 1988). The fungus Botryotinia (=Lasiodiplodia) theobromae (Pat.) Griff & Maubl. (Botryosphaeriaceae) has previously been isolated from the lesions of plants showing dieback symptoms: reddish-brown lesions originating from leaf axils, yellowing of the leaves and subsequent stem death, in M.pigra (Wilson & Pitkethley 1992).

Abiotic factors have also been implicated in dieback, either as directly or as a stressor (Fensham et al. 2009; Manion & Lachance 1992). Soil salinity has been shown to boost the dieback in process in the floodplain habitat in Florida, USA (Kaplan & Munoz-Carpena 2014). Drought conditions have caused increased soil salinity, thereby potentially contributing to massive die-off of the salt marshes in southern United States (Silliman, 2005). In addition, soil acidity (pH) restricts the availability of essential macro-nutrients, leading to root mortality and eventually dieback in sugar maple, Acer saccharum (Houle et al. 2007; Watmough 2010). In blackberry (Rubus anglocandicans) decline in Western Australia, flooding events is considered as a predisposing factor and this stress condition makes blackberries susceptible to root pathogen infection (Aghighi et al. 2014).

Saltwater intrusion into the floodplains is considered a potential stressor leading to dieback and may subsequently lead to plant death (Busch & Smith 1995; Volkmar et al. 1998). In the Northern

Territory, flooding of the river is strongly regulated by the monsoonal climate. The river is subject to semi-diurnal tidal behaviour, with maximum spring tides ranging from 6 to 8 m. As the tidal length of the river is eight times greater than the flood and ebb excursion of water during the water cycle, the upstream propagation of salt water after the wet season is due to turbulent diffusion. By this means, saline water penetrates as far as 80 km upstream from the river mouth by the late dry season (Woodroffe 1995).

This thesis chapter aims for the first time to quantitatively describe the effects of dieback on M. pigra plants and populations in the field, and to assess the role of biotic and abiotic factors in its occurrence. Field surveys were designed to make assessments of the prevalence and abundance of biocontrol agents, native insect borers, endophytic fungi and abiotic factors, to help explain the dieback phenomenon. Specifically, the objectives of this study were to: (1) to describe the “dieback phenomena” through time at a plant and population level; and (2) to test whether classical biological control, fungal endophytes and abiotic factors (soil salinity and pH) are the cause of the dieback phenomenon.

3.3 Method

3.3.1 Study System

3.3.1.1 Site

Fieldwork was conducted at Melaleuca Station, Mary River region, Northern Territory, approximately 200 kilometres east of Darwin (12°38.49’S, 131°46.59’E). The region has a tropical wet-dry monsoonal climate. Mean maximum temperatures are high throughout the year (32-37°C), with a mean annual rainfall of 1,584 mm. Rainfall is highly seasonal, with more than 90% of the rain falls during the wet season (November to April) (www.bom.gov.au/climate/data).

Melaleuca Station has a total area of 450 sq. km, of which 300 sq. km is on a floodplain. About 60% of the station is seasonally inundated (T. Searle, pers.comm.). The primary land use is as a pastoral cattle enterprise (Searle 2004). Initially mimosa established on the property in the 1980’s as a very small infestation, but by 1993 the area infested had increased to 10,000 ha.

3.3.1.2 Study Design

Permanent field sites were across levels of dieback observed during a reconnaissance visit in August 2011. Levels of dieback were categorised into four classes based on visual descriptions: healthy, dieback, recovering from dieback and dead (Table 3.1, Figure 3.1 to 3.2). There were four locations selected on the station, each with four plots (two treatments x 2 replicates), giving a total of 16 plots (Figure 3.3). Treatments were the “four health classes” observed in the area namely: healthy, dieback, recovering and dead” All the plots were located where there was a moderate to very dense M. pigra population (approximately 5 plants/m2) (Figure 3.1 to 3.2). Treatments were grouped as best as possible given the limitations imposed by distribution of dieback symptoms across the study location. A major constraint was that healthy plants were difficult to find and were largely restricted to the edge of the study location. As a result, healthy replicates were paired with dead replicates at two locations, and dieback with recovering at another two locations (Figure 3.3). Sixteen permanent plots were identified from healthy populations that showed no symptoms, to populations where all juveniles and adults plants were dead, but where young seedlings were present.

Table 3.1 Dieback class descriptions of M. pigra used in this study.

Dieback classes Visual definitions

Healthy Little plant stress evident, all main stems alive, 50 – 70% of the terminal stems with foliage.

Dieback Plants stressed, some main stems recently dead or affected by die-back. Typically < 30% terminal stems with foliage.

Recovering from dieback All main stems dead but fresh, green, young stems reshooting from the base.

Dead Plant death confirmed; no living material observed.

A B

Figure 3.1 An example of a healthy class of M. pigra population (usually located on the edge of a stand) (A); an example of M. pigra population showing a partial dieback class (B).

A B

Figure 3.2 An example of a recovering class of M. pigra population (A); an example of M. pigra population showing a “dead” class (B).

3.3.1.4 Plot Description

The plot was the primary sampling unit. The size of individual plots was 5 m x 5 m (Figure 3.4). In each plot, 5 adult plants were individually tagged using cattle tags. Each tagged tree had its health 38

status assessed. The purpose of individual plots was to conduct a census of the number of adult plants, juveniles and seedlings over time, in an identified space. Each plot was assessed every 6 months to record demographic changes, abundance/prevalence of major insect stem-borers, extent of pathogen symptoms, and plant health. Plant demography was assessed annually over three years, from November 2011 to November 2013.

Figure 3.3 The study site: Melaleuca Station, Mary River, Northern Australia (inset) A, and map of Australia (B)

3.3.2 Describing M. pigra dieback phenomenon

This plot-level assessment was used to assess the overall picture of the dieback progression over time. Plant population count and growth measurements were conducted. Inside each plot, five plants were selected and tagged for their health rating, using a ten-point scale developed for multi-branched M. pigra shrubs (Table 3.2).

Plant Population Count

Demographic changes in each plot was recorded, by counting the number of seedlings (< 0.2 m tall), juvenile (0.2 – 1.5 m tall) and adult plants (> 1.5 m tall). It was also recorded whether the juvenile or adult plants were dead or alive. Plant population was assessed annually for two years.

Plant Growth

Plant size and stem diameters were measured on each tagged plant on each visit to track plant growth against health status. Plant height and width were measured for all five tagged plants in each plot,

using a five-metre ruler. Where the stem canopy was not approximately circular in cross section, two measurements were taken for the living and dead canopy widths (Fig 3.5 A). Using this data, the volume of each plant was calculated using the formula of a cone which is: 1 푉 = 휋푟2ℎ 3 The formula of the cone was appropriate, as M. pigra architecture has a cone- shaped structure (Fig 3.5 B). Three stem diameter measurements were made on the three biggest stems, 10 cm from the ground surface, the measurement being made using digital Vernier calipers. For approximately circular stems in cross section, two measurements were taken and averaged.

Table 3.2 Ten-point disease rating scale developed for M. pigra. Scale Definition

1 Definitely dead.

2 All/most stems dead, no reshooting evident yet but may still be alive.

3 Top-kill, post wet season re-shooting from base only.

4 Died back to main stems with post wet re-shooting from base and main stems only.

5 Partial dieback, some main stems recently dead or died-back, some stems still alive and/or healthy.

6 Early stage dieback, all or nearly all main stems alive, significant signs of pathogen damage; typically <30% terminal stems with foliage.

7 Stressed, all main stems alive, minor or no signs of pathogen damage; typically 30- 50% of terminal stems with foliage.

8 Moderate stress, all main stems alive, minor or no signs of pathogen damage; typically 30 – 50% of terminal stems with foliage.

9 Minor stress, all main stems alive, only minor insect and/or pathogen damage; typically 50-70% of terminal stems with foliage.

10 Very healthy, fully foliated, 70-90% terminal stems with foliage, minor no signs of current insect or pathogen damage. Significantly healthy stand. Not typically found in NT.

A B

Figure 3.5 Measuring of canopy width using a height rule (A). Volume calculation used the formula of a cone (B).

3.3.3 Insect Activity Assessment

Insect activity counts were conducted on all tagged trees in each plot, by assessing all the stems and branches, from ground level to the top of the plants. The incidence of four insects regarded as causing most damage to M. pigra was recorded. These insects were: the biocontrol agent, Carmenta mimosa (a stem borer), and two native insect borers, Platyomopsis humeralis (a stem girdler) and Maroga setiotricha (a stem-borer). In the field, each insect was recorded as being present or absent; then, if present, a count was recorded on a data sheet (Appendix B) following the methodology of Smith and Wilson (1995). The assessment was supported by an expert in insect field identification, Bert Lukitsch (Biological Control, Weed Sciences, DLRM, Northern Territory). For the leaf and tip-boring moth, Neurostrota gunniella, ten pieces of 50- cm long mimosa stem tips were sampled randomly from the M. pigra plants outside of the plot (Smith & Wilson 1995). This procedure allowed random checking to validate the presence of these insects inside the stems without destroying the plants in the experimental plots. Counting of Neurostrota holes was done in the Weeds Branch Laboratory, due to the need for adequate light and time to identify very small holes in the stems. The presence of other insects aside from the stem-borers was also noted.

3.3.4 Endophyte Community Assessment

For each survey plot, destructive stem sampling was done of three M. pigra plants found just outside the permanent plots. Selection of plant specimen was done randomly as the plants located outside the plot represents its health class. In each survey endophytic communities were sampled in three adult

M. pigra plants located just outside of each plot. Using secateurs, a 20 cm-long stem piece was cut from each plant. Stem pieces were put in separate paper bags, labelled and stored in a cool room (~4- 5oC) until isolation, which was done within two weeks of collection. The blades of the secateurs were wiped and cleaned between treatment plots, using commercial wet wipes containing isopropyl alcohol. External and internal disease symptoms were recorded for each stem piece. Isolations were done in batches commencing with the healthy stems. Stems with corky bark were washed thoroughly with running water before further treatment. Details of isolation, identification and abundance methodology of the fungal endophytes are outlined in Chapter 4, Sections 4.2.4 to 4.2.9.

3.3.5 Abiotic Conditions

3.3.5.1 Soil Sampling

Soil pH and electrical conductivity were assessed in each plot and in each survey period. Within each plot, soil samples were collected at a depth of 20-30 centimetres using a soil corer. The soil sampling was done after first dividing the plot into four parts. The sample sites within each plot were defined based on a diamond-shaped sampling path. The samples from within individual plots were then combined in a paper bag. Quartering of the samples was done in the Weeds Branch Laboratory, Berrimah, Northern Territory. The samples were then air-dried before then being transported to the Central Analytical Laboratory, Gatton campus, of the University of Queensland, for the analyses of pH and electrical conductivity.

The methodology used for analysing pH and electrical conductivity of the soil samples, was from the method outlined in the Australian Laboratory Handbook of Soil and Water Chemical Methods. All analyses were done in triplicate (Rayment & Lyons 2011) (Appendix C).

3.3.8 Data Analysis a. Plant Health. Plant health was determined by plotting the average health rating value of each dieback class against time. The pattern was then examined visually whether there was an increase or decrease in health over time across all dieback classes. b. Plant Growth. Two types of data were examined for plant growth: these were plant volume and plant stem diameter. Plant volume was assessed by comparing the change of the individual plant volume growth from year one and year two. Patterns were then visually assessed whether individual plant has grown or decreased in volume over time. For plant diameter, the three measurements taken were averaged and plotted against time and across all dieback classes. Patterns were visually assessed whether stem diameters continued to increase over time.

c. Plant population count. Each plant category (adult, juveniles and seedlings) count per dieback class was averaged. Qualitative description was done on the pattern observed in terms of recruitment, abundance of juveniles and seedlings over time. Mortality of tagged plants was calculated as percentage of plants that died in the duration of study. d. Abiotic factors. To test the effect of salinity and pH on the changing M. pigra population dynamics over time, we examined health change patterns through time. In particular, we focus on the association of pH and salinity compared to the change in health class. Using PRIMER and PERMANOVA E Software v7, Principal Coordinate Analysis (PCA) was used to analyse whether the two abiotic factors (pH and salinity) had an influence on the abundance of fungal communities over time. The distance matrix used was Euclidean distance, while the similarity measure that was used was Bray-Curtis, which were both default in the software. PERMANOVA was used to calculate permutations and see whether there was a statistical difference between the two distance matrixes plotted. e. Biotic factors. For both biotic factors, fungi and insect borers, the relationships between the causal factors and health classes at the time of sampling were assessed. The causality of changes that occurred prior to the survey, and the change in health classes during the survey, were also examined. The relationship change was also determined after the plant samples had died.

3.4 Results

3.4.1 Plant Health

The assessment of tagged trees showed that tree health declined or remained static in most plots throughout the three year period of study. Plots that were initially healthy showed a rapid decline in health within the first 6-12 months, with stem death and top kill which then did not deteriorate further for the remainder of the study period (Figure 3.6a). Plots that were initially categorised as dieback- affected declined in health throughout the study to the point where stem death and top kill was predominant in all plots (Figure 3.6b).

Figure 3.6 Plant health through time for four replicate plots initially assessed as healthy (a), dieback (b), recovering from dieback (c), and dead (d) (from 2011 to 2013). The box descriptors in (a) are the disease severity classes from the 10- point disease rating scale developed for M. pigra (Table 3.2).

The health of plots initially assigned as recovering from dieback began as being primarily top-kill or stem death, with healthy re-shoots from the base, and then remained in this category for the duration of the study (Figure 3.6c). The four “dead” plots were set up after the mortality of all existing plants (this means that this site was characterised as M. pigra populations with no living adult or juveniles present), but new seedlings about 10 cm high were already present (Fig 3.7). New seedlings in previously categorised dead plots were still healthy in the dry season following the dieback event (Figure 3.6d). However, plant health subsequently declined very rapidly in two plots, and more gradually in the remaining two plots, with stem death being common in all plots by the end of the trial period.

A B

3.4.2 Plant Growth

Overall, no positive growth was recorded for plants above 12 m3 and was only rarely observed in smaller plants (Figure 3.8). The highest growth rates were observed in initially healthy plots, (Figure 3.8a). By the second year most trees had negative growth rates, trending with initial plant size. In initially recovering plots, growth rates for all but the smallest plants (< 4m3) was negative, increasing with plant volume [year 2 – relationship reflects high mortality] (Figure 3.8c). Positive growth rates in initially dieback plots was modest for smaller plants (< 12 m3) in the first year, and negative in year 2, trending with plant size (Figure 3.8b). Growth of new plants in plots where plants had died the previous year were close to zero or declining, despite almost all plants being small in size (< 8m3) (Figure 3.8d).

50 a) Healthy b) Dieback 40 growth at year1 50 30 growth at year2

) 40

) 3 20 30 20 10 10

0 0 0 10 20 30 40

Changein volume(m 0 10 20 30 40

-10 -10 Changein volume (m -20 -20 -30 Initial plant volume (m3) -30 Initial plant volume (m3) c) Recovering d) Dead growth at year2 50 50 40

) 3 ) 3 30 30 20 20 10 10

0 0 0 10 20 30 40 0 10 20 30 40

-10 -10

Changein volume (m Change in volume (m volume in Change -20 -20 -30 Initial Plant volime (m3) -30 Initial plant volume (m3)

Figure 3.8 Plant growth against initial plant volume in the first and second years (from healthy to dead plots). Only new plants are included in the dead site and were first tagged in year 2.

100 Diameter Year 1 100 90 a.) Healthy Diameter Year 2 90 b.) Dieback 80 Diameter Year 3 80 70 70 60 60 50 50 40 40 30

30 Stemdiameter (mm) 20 20 10 Stem diameter (mm) diameter Stem 10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Individual Plants Individual Plants

100 c.) Recovering Diameter year 2 100 90 d.) Dead 80 Diameter year 3 80 70 60 60 50 40 40 30 StemDiameter (mm) 20 20 (mm) Diameter Stem 10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Individual Plants Individual Plants

Figure 3.9 Stem diameter averaged for the three biggest stems of individual plants ordered by the plants’ initial size. Open squares are dead plants.

Plant diameter growth at 10 cm height above the ground was observed for all sites except for the initially recovering plots where plant diameter decreased for most plants (Figure 3.9c). Although experiencing top-kill and canopy loss, plants continued to increase in stem diameter for the duration of the study. Plants in the initially dieback-affected site, on average, had the thickest stem diameters, but by the third year most of the stems measured were already dead (Fig 3.9b).

3.4.3 Plant Demography

All seedlings inside the plots were counted after each wet season and during the dry season when the full population census was conducted (Table 3.2). New recruits (seedlings) were more common in the first year, excluding the dead sites, where the reverse was true. Overall, recruitment declined with decreasing stand health. In initially healthy plots, seedling recruitment1 continued through to at least the juvenile stage throughout the study, although adult mortality was high in the final year. Recruitment in initially healthy plots gradually declined as the health of M. pigra decreased (Figure 3.10a).

Table 3.3 Average annual count of live adult plants (>1.5 m tall), juveniles (0.2 – 1.5 m tall) and seedlings (0.2 m tall) per plot over three years in a 5 m by 5 m plot.

Initial Plot Description Average Count (living) Level 2011 2012 2013 Healthy Seedling 28 186 90 Juvenile 1 42 52

Adult 7 16 5 Dieback Seedling 81 257 14 Juvenile 2 35 92 Adult 7 8 7 Recovering Seedling 40 17 11 Dieback Juvenile 0 19 10 Adult 5 33 12 Dead Seedling 22 8 114 Juvenile 2 7 11 Adult 0 23 13

Seedlings were relatively abundant in the dieback plots, (Figure 3.10 b) and recruitment through to juveniles was high; however adult numbers remained largely unaltered (Table 3.2). The seedling

1 Recruitment is the term used to classify the movement of plants from one category to another. 48

count in dieback sites was high in the first year but dropped markedly from the start of the second year.

Recruitment through to adults was high after the first year in plots assigned as recovering, but juvenile and adult mortality was high after the second year and seedling densities declined. Recovering sites did not have much recruitment, given a very low and steady health status over time (health disease rating all throughout the trial period was between rating 1-4) (Figure 3.10).

In initially dead plots, post-dieback recruits recorded at the start of the trial became adults after the first year, which may, in turn, have contributed to increased seedlings in the second year. However, adult mortality was high by the end of the second year. Dead sites showed a different trend to other plots, with recruitment increasing with decreasing health (Figure 3.10 d).

Table 3.4 Percent mortality of tagged adult plants per 5 m x 5 m plot over three years. Initial Plot Level 2011 2012 2013 Healthy na 20% 45% Dieback na 10% 45%` Recovering dieback na 15% 35% Dead na na 5%

350 10 300 10 a)healthy 9 9 300 250 8 b) dieback 8 250 7 200 7 6 200 6 5 150 5 150 4 4

100 3 100 3 Plant Health 2 2 50 1 50 1 Mean Seedling MeanSeedling count perPlot 0 0 0 0 Wet 2011 Dry 2012 Wet 2012 Dry 2013 Wet 2013 Wet 2011 Dry 2012 Wet 2012 Dry 2013 Wet 2013 Season

300 10 300 10 9 9 250 c) recovering 8 250 d) dead 8 7 200 200 7 6 6 150 5 150 5 4 4

100 Plant health 3 100 3 2 50 50 2 Mean Seedling MeanSeedling Count per Plot 1 1 0 0 0 0 Wet 2011 Dry 2012 Wet 2012 Dry 2013 Wet 2013 Wet 2011 Dry 2012 Wet 2012 Dry 2013 Wet 2013 Season Season Figure 3.10 The relationships between adult plant health (line, right-hand axis) versus average seedling densities (bars, left-hand axis) through time for plot that were initially classified as a) healthy, b) dieback, c.) recovering dieback and d.) dead.

Adult mortality among tagged plants was moderate after the first year (2012) and high in the second (Table 3.3). The exception was in plots where tagged plants were all new recruits following a dieback event (“dead” plots), where mortality was only 5% in the first year following tagging (Table 3.3). The dead plots did not show much additional mortality while there was also not much regrowth; there was a modest increase in stem diameter resulting in bigger plants.

3.4.4 Abiotic Factors

Soils were black cracking clays. pH and electrical conductivity measured at a depth of 20-30 cm showed that, on average, all study plots are on slightly acidic soil and have either non-saline or slightly saline electrical conductivity values (Table 3.5).

Table 3.5 Mean pH and salinity of the dieback classes. Dieback classes pH Description Electrical Description conductivity #(compared to standard scale) Units pH* mS/m Healthy Plot Wet 2011 4.28±0.01 Slightly acidic 19.5±2.0 Slightly saline Dry 2012 4.75±0.02 Slightly acidic 16.4±10.0 Non-saline Wet 2012 4.69±0.08 Slightly acidic 22.1±6.0 Slightly-saline Dry 2013 4.78±0.19 Slightly acidic 18.9±11.0 Slightly saline Wet 2013 4.60±0.09 Slightly acidic 21.1±15.0 Slightly saline Dieback Plot Wet 2011 4.88±0.05 Slightly acidic 25.5±19.0 Non-saline Dry 2012 4.74±0.03 Slightly acidic 27.3±14.0 Slightly-saline Wet 2012 4.78±0.04 Slightly acidic 26.4±8.0 Slightly saline Dry 2013 4.68±0.08 Slightly acidic 18.8±19.0 Non-saline Wet 2013 4.78±0.10 Slightly acidic 22.1±4.0 Non-saline Recovering Dieback Plot Wet 2011 5.00±0.06 Slightly acidic 20.3±8.0 Non-saline Dry 2012 5.16±0.15 Slightly acidic 27.3±11.0 Slightly-saline Wet 2012 5.29±0.21 Slightly acidic 22.7±5.0 Non-saline Dry 2013 5.48±0.13 Slightly acidic 30.8±16.0 Slightly-saline Wet 2013 5.49±0.21 Slightly acidic 29.1±17.0 Non-saline Dead Plot Wet 2011 4.55±0.05 Slightly acidic 30.2±10.0 Slightly-saline Dry 2012 4.74±0.03 Slightly acidic 13.5±12.0 Non-saline Wet 2012 4.63±0.06 Slightly acidic 12.3±2.0 Non-saline Dry 2013 4.70±0.10 Slightly acidic 19.0±18.0 Non-saline Wet 2013 4.93±0.11 Slightly acidic 14.5±10.0 Non-saline *(sample analyses in triplicate and averaged of 4 replicate plots) # (As compared to standard scale for salinity from Department of Agriculture and Food (www.agric.wa.gov.au); (pH standard is from the Standard Methods for the Examination of Water and Wastewater book eds Greenberg,A.E et.al. 1992).

There was no significant pH pattern observed in relation to the change in health class. Although pH fluctuated at some sites and between seasons, there was no relationship to the changing M. pigra population dynamics over time. Salinity also fluctuated over time but it did not explain the relationship pattern of the increase and decrease of the M. pigra population dynamics observed. The results in the representation of the Principal Component Analysis (PCA) showed that the two abiotic factors, pH and salinity, had very little influence on the fungal communities through time (Figure 3.11).

Figure 3.11 Principal Coordinate Analysis of the different treatments and the abiotic factors (pH and salinity).

3.4.5 Biotic Factors: Stem Borers

Analysis suggests that there was no support for the possibility that the three borers, Carmenta, Maroga and Platyomopsis were the causal agents in the observed changes in health class before and after time (t) (Figures 3.12 and 3.13). There was no positive relationship observed, as the R value was close to zero. Big changes at ((t+1)-t) in health were observed even at low insect abundance (Figure 3.12). A positive change was when plant health improved.

A weak negative relationship was observed between insect abundance and prior change in health class (Figure 3.12). Insects were more abundant following a decline in health. A positive change is when health was classified as having improved in the preceding 6 months. All three insect borers 51

showed the same relationship patterns and this relationship was ubiquitous in all plots, with a greater abundance of insects being associated after the observed decline in plant health.

There was no positive relationship between the changes in health class against the total borer count over time (t). Even when total borer abundance was combined, there was only a slight support (Figure 3.14 and 3.15) for a conclusion that the feeding of the borers of plants before time t ((t-(t-1)) was positively correlated with the declining health class change recorded.

Change in Health Class at (t-(t-1))

0 0 5 10 15 20 25 30 35

-2 R² = 0.0168 Changein Health R² = 0.0363 R² = 0.0082 -4

-6

-8

-10 Insect Abundance

carmenta Maroga Platyomopsis Linear (carmenta) Linear (Maroga) Linear (Platyomopsis)

Change in Health Class at ((t+1)-t)

8 6 4 2

0 R² = 0.0075 R² = 0.0009 R² = 0.0007 0 5 10 15 20 25 30 35

-2 Change in Health in Change -4

-6

-8

-10 Insect Abundance Carmenta Maroga Platyomopsis

Linear (Carmenta) Linear (Maroga) Linear (Platyomopsis)

Table 3.6 Abundance of Neurostrota gunniella across all plots and seasons. On average, N. gunniella abundance did not explain the changes in M. pigra health over time. However, N. gunniuella is ubiquitous and abundant in all plots and seasons.

Seasons Healthy Plot Dieback Plot Recovering Plot Dead Plot Wet 2011 7 6 7 7 Dry 2012 4 4 5 6 Wet 2012 5 5 4 4 Dry 2013 4 5 4 4 Wet 2013 3 3 4 4

Total Borers @ HC (t-(t-1)) 60

Insect Insect Abundance 20

R² = 0.0382

0 -10 -8 -6 -4 -2 0 2 4 6 8 10 Change in Health Class

Figure 3.14 Scatterplot diagram of the total borer abundance versus change in health class in the prior time step (t-(t-1)).

There was no positive relationship between the change in health class against the total borer count before and after time (t). Even when the abundance of all borers were combined, there was only a slight support (Figure 3.14) that feeding of the borers to plants before time t ((t-(t-1)) was positively correlated to the declining health class change recorded.

3.4.6 Biotic Factors: Fungal endophytes/pathogens

A total of 24 endophyte taxa were isolated during the study (Chapter 4, Section 4.4.2). Of those only three major taxa were sufficiently common for analysis (recorded from at least 169 M. pigra samples). These taxa included multiple species that could not be reliably differentiated morphologically (see Chapter 4).

Analysis suggests that the three taxa: Botryosphaeriaceae, Pestalotiopsis spp. and Diaporthe spp. were ubiquitous and abundant across all plots and seasons (Figures 3.16 to 3.18). The abundance of borers did not show association with the health change across all plots and over time.

Total Borers @ HC ((t+1)-1) 60

Insect Insect Abundance 20

10 R² = 0.001

0 -10 -8 -6 -4 -2 0 2 4 6 8 10 Change in Health Class

Figure 3.15. Scatterplot diagram of the total borer abundance versus change in health class (HC) in the subsequent time step ((t+1)-t.

a) healthy 9&10 b) dieback 13&14 5&6 15&16 20 14 18 12 16 14 10 12 8 10 6 8 4 6 4 2 2 Mean count of count of Mean 0 0

Botryosphaeriaceae Jun-12 Nov-12 Jun-13 Nov-13 Jun-12 Nov-12 Jun-13 Nov-13

1&2 c) recovering d) dead 3&4

11&12 7&8 16 14 14 12 12 10 10 8 8 6 6 4 4 Mean Mean count of 2 2

Botryosphaeriaceae 0 0 Jun-12 Nov-12 Jun-13 Nov-13 Jun-12 Nov-12 Jun-13 Nov-13 Seasons Seasons

Figure 3.16 Abundance of the Botryosphaeriaceae across all plots (1-16) and through time (blue and red legend represent the replicate plots). No clear pattern was seen with the change in health class through time. The number of Botryosphaeriaceae fluctuates through time. The Botryosphaeriaceae were abundant and present in almost all plots but the pattern cannot explain the declining health change of the different dieback levels.

9&10 5&6 12 a) dieback 13&14 15&16 c) healthy . 12

spp

10 10

8 8 6 6 4 Pestalotiopsis 4 2 0 2

0 Jun-12 Nov-12 Jun-13 Nov-13

Mean Mean count of Jun-12 Nov-12 Jun-13 Nov-13

e) recovering g) dead . 12 1&2 3&4

spp

11&12 7&8 9 12 8 10 7 6 8 5 Pestalotiopsis 6 4 3 4 2 1 2 0 0

Mean Mean count of Jun-12 Nov-12 Jun-13 Nov-13 Jun-12 Nov-12 Jun-13 Nov-13 Season Season

9&10 a) healthy 13&14 b) dieback 15&16 5&6

spp. 8 10 9 7 8 6 7 5

6 Diaporthe 5 4 4 3 3 2 2 1 1 0 0 Jun-12 Nov-12 Jun-13 Nov-13 Jun-12 Nov-12 Jun-13 Nov-13

of count Mean

c) recovering 1&2 11&12 d) dead 3&4 7&8

spp. 10

9 6 8 5 7 6 4

Diaporthe 5 3 4 3 2 2 1 1 0 0 Jun-12 Nov-12 Jun-13 Nov-13 Jun-12 Nov-12 Jun-13 Nov-13

Mean count of count Mean Season Season

3.5 Discussion

The fact that population (stand) dieback in M. pigra is in multiple catchments in the Northern Territory and over a large areas, makes the dieback phenomenon of considerable ecological importance. Understanding the underlying causes of dieback helps us design better management strategies in the control of this weed. This study shows that M. pigra dieback is ubiquitous across the study period. Its effect on plant health and survival means that M. pigra populations were highly dynamic throughout the three-year study site. No evidence was found that abiotic factors (pH and electrical conductivity) was implicated in dieback. Four species of insect borers, two native species and two released as biocontrol agents, were consistently abundant throughout the study but no relationship was found between the dieback cycles and borer densities. This is despite previous studies showing that the two biocontrol agents were the most damaging of the 13 species released to date (Heard 2012). All sites had similar fungal communities and no relationship was established between the three most common endophytic fungal taxa and declining plant health. Endophytic fungi included Botryodiplodia theobromae = Lasiodiplodia theobromae which was previously implicated in M. pigra dieback (Wilson & Pitkethley 1992).

In this study, all sites across the study populations were affected by dieback, but the effects were highly variable and asynchronous. Some sites showed a gradual decline in health while other plots remained static. All size classes were affected by dieback. Dieback resulted in high mortality (up to 45%) in the third year of the trial period. This result was similar to that reported by Paynter (2005) who reported that M. pigra survival significantly declined over time. In the present study only 45% of the tagged trees survived beyond two years, which suggests a half-life of c. 24 months.

In Australia, it is not clear how long M. pigra plants survive in the absence of a “decline syndrome”. Mortality rates differ by soil type. Lonsdale (1992) showed that the half- life of tagged plants in Australia was 28 and 22 months on heavy and sandy soil types, respectively (Lonsdale 1992), with Miller (1988) reporting that mature plants generally died within a period of 5 years. However, the size of some individual plants and the persistence of some populations has suggested that plants can live for up to 15 years or longer in Australia (Rea 1998). Recruitment among plant dieback categories was partially reduced as the health of the plants declined. Overall, average seedling densities recorded in this study was lower than that recorded by (Paynter 2004). This present study demonstrated that recruitment in dead sites was from the seed bank not from the adult plants.

Over the three year duration of the study, there was no strong supporting relationship with the dynamics of the dieback phenomena that was observed on the property when two biocontrol insect

agents were evaluated for causality. Borers were ubiquitous at low density except during the 2012 wet season sampling period, although the change occurred across all plots and it was not related to the decline in plant survival. The presence of the borers everywhere but at low densities contradicts the results of Paynter (2006) who reported that two borers, Neurostrota gunniella and Carmenta mimosa were abundant at the edges of stands. Recruitment in this present study was correlated with declining stands, except for the dead plots, with recruitments coming from seed banks and not the seed fall. Modelling studies by Buckley et al. (2004) have indicated that if the number of recruits produced from a single M. pigra plant is high, then the likelihood of eliminating M. pigra using seed- feeding insects is low. Given the dynamics of the plant population at Melaleuca station where this study showed a cyclic increase and decrease of plant health, Buckley et al. (2004), in M. pigra modelling studies, demonstrated that biocontrol alone is only successful at low levels of disease incidence. Current biocontrol agents may take decades to reduce a stand to < 5% of current site populations. Paynter (2005) found that M. pigra seed rain was negatively correlated with Carmenta mimosa damage. Caution needs to be taken in attributing causality to one agent like insect herbivores. There may be other causes of dieback, including root pathogens and root boring insects.

There was no biological relevance of the soil pH and salinity measurements that were made over time and across different dieback levels in this study. The results of the impact of soil pH recorded at Melaleuca Station, is lower than that reported by Edmeades (2011) at Mary River, Northern Territory, for native grazing vegetation. No direct evidence was found in this study of effects of either pH or salinity on M. pigra dieback. A possible explanation for this result might be related to the fact that, in Australia, M. pigra can tolerate higher levels of salinity than reported elsewhere (e.g ̴ 18 ppt) (Miller 1983). M. pigra favours and is adapted to a wet-dry tropical climate and seasonal inundation, and dieback has not been observed in similar habitats in its native range (Harley 1985). Observational studies by Mulrennan and Woodroffe (1998) in the coastal plains of the Lower Mary River in the Northern Territory have showed that the penetration of salt water into the floodplains stresses the natural vegetation and leads to extensive dieback of paperbark (Melaleuca spp.) stands. However, (Lonsdale 1992) reported that in its native range, M. pigra occupies similar habitats and conditions to paperbark but has not shown dieback events.

The limitations of this study were as follows:

1. For fungal endophyte communities, samples were only taken from stems. Other microorganisms like oomycetes, bacteria or viruses, which may also be a causal agent for a decline in plant populations, were not examined.

2. There may be possible cryptic damage that has not been recorded, for the biocontrol agent Carmenta mimosa and other native insects like Grapthostethus cardinalis (swarming bug). In addition, other four agents, the looper moth Macaria pallidata, the tip and green seed feeder Chalcodermis serripes, the tip and flower weevil Coelocephalapion pigrae, and the root feeder Malacorhinus irregularis may be capable of causing dieback, but because their populations are mobile and their damage is less persistent, it is more difficult to survey and record their populations and effects (Ostermeyer & Grace 2007; Heard et al. 2010).

3. Measurement of the effects of soil salinity might also need to be more detailed to evaluate the impact of salinity on individual ion content, together with related nutrient analysis in relation to the potential physiological effects on the plant.

4. Finally, there was difficulty in making an accurate comparison between dieback affected and healthy sites in the study area, as the initially healthy sites at the commencement of this study did not remain so as the dieback began to manifest itself over time.

3.6 Conclusion

This study is the first to describe M. pigra dieback at the level of individual plants and whole plant populations. The results reveal that M. pigra dieback has potential dramatic plant population-level effects. Dieback in M. pigra fluctuates, and the different sites in this study reflected in an inconsistent increases and decreases in plant health over time. In such circumstances extensive classical biocontrol evaluation was not able to consider other potential effects of other factors in the declining population dynamics of M. pigra. Dieback is an important phenomenon in M. pigra, but this study has showed no association between dieback and the biotic factors (insect borers and endophytic stem pathogens) and the abiotic factors (pH and salinity). On the basis of this study, a single cause of dieback in M. pigra was not able to be determined. It is likely that dieback is the result of the interaction several biotic and abiotic factors in the M. pigra floodplain habitat.

CHAPTER 4

Identification of endophytic fungi from Mimosa pigra based on cultural morphology, ITS and partial translation elongation factor (EF1-α) sequences.

4.1 Abstract

A survey of the endophytic community in stems of Mimosa pigra was carried out on plants growing on Melaleuca Station in the Mary River Region of the Northern Territory. Stems were collected during the early wet and dry season in 2012 and 2013. The endophyte assemblages identified using cultural characters consisted 24 “morphospecies”, being grouped based on cultural morphology and sporulation. The objective of this study was to identify species of endophytes associated with healthy and dieback-affected Mimosa pigra. A total of 67 isolates were used and identified through a combination of morphology and phylogenetic analyses, based on partial translation elongation factor 1- α sequence (EF1-α) and internal transcribed spacers (ITS) regions. The following species were identified: Lasiodiplodia iraniensis, L. pseudotheobromae, Pestalotiopsis clavispora, P. ellipsospora, Diaporthe phaseolorum, and a Diaporthe sp. These species represented 61% of all isolates sampled over time Only Lasiodiplodia theobromae has been reported to have been previously isolated and identified in earlier studies. For the remaining species, this was the first time that they have been reported as being associated with healthy and dieback-affected Mimosa pigra in the Northern Territory.

4.2 Introduction

Most plants are inhabited by microorganisms called endophytes. While a variety of definitions of the term endophyte have been suggested (Hyde & Soytong 2008), the definition suggested by Rollinger (Rollinger & Langenheim 1993) was used in this study. In brief, endophytes are fungi isolated from surface-sterilised plant material, and are colonisers of the living internal tissues of their host plant.

Endophyte colonisation influences the form and function of plant populations (Sinclair & Cerkauskas 1996; Rudgers et al. 2010). Fungal endophyte asssociations with their host plant have been described as a continuum. They establish dynamic relationships (mutualism, commensalism and parasitism) (Schulz & Boyle 2005; Aguilar-Trigueros et al. 2014). They may also become saprobes on dead trees, de Errasti et al. (2014) and possible antagonists against host pathogens (Rocha et al. 2011).

The species concepts used for the identification of fungi have changed continually as modern techniques of identification have emerged. Traditional approaches for the isolation and identification of fungal endophytes have been employed in many endophyte studies in grasses and woody plants (Guo et al. 2000). However, over the past two decades, much progress has been made in relation to the ability to define fungal species, through the use of molecular data (Denman et al. 2003; Damm et al. 2012). More recently, Multi Locus Sequencing Analysis (MLSA) has been increasingly used to accurately identify the isolated endophytes. For example, Marsberg et al. (2014) identified that, combined analysis of datasets from five different gene regions of fungal leaf endophytes of Eucalyptus grandis and Syzygium cordatum in South Africa, resolved the taxonomic identification of their isolated endophyte.

Mimosa pigra L. (Fabaceae) is a Weed of National Significance in Australia where it invades extensive wetland areas in the Top End of the Northern Territory (Thorp & Lynch 2000). Widespread dieback of M. pigra has been reported in the Adelaide River floodplains in the Northern Territory since at least 1988 (Miller 1988). Dieback is a condition often found in woody plants, and is associated with progressive death of stems, shoots, branches or roots, often starting at the tips (Ostry et al. 2011). The fungus Lasiodiplodia theobromae (Pat.) Griff & Maubl. (Botryosphaeriaceae) has previously been isolated from typical dieback symptoms of M. pigra in the Northern Territory (Wilson & Pitkethley 1992). Significant patches of dieback in M. pigra continue to be observed in other parts of the Northern Territory, including sites in the Mary River and Daly River regions.

Despite the increasing knowledge about the presence of fungal endophytes in plants (Arnold & Lutzoni 2007), very little is known about the association between endophytic fungi and plant decline (Giordano et al. 2009; Dakin et al. 2010; Elmer & Marra 2011). To our knowledge, no information 61

is presently available on the aetiology of the dieback phenomenon in relation to the fungal endophytes in M. pigra. In addition, it is believed that other fungi may also play a role in the dieback phenomenon. Investigations are lacking that have directly compared the fungal endophytic communities of healthy and dieback-affected M. pigra in the same sampling sites and through time.

This chapter focuses on the occurrence and identification of endophytic fungal communities within M. pigra in the Northern Territory. The nature of the interaction between the host, M. pigra, and its isolated endophytes, is examined in Chapter 5.

The aim was to test the use of a culture-based approach (4 types of culture media) to isolate and survey fungal endophytic communities in stems from 16 sites representing different levels of dieback expression (from healthy to recovering) observed in the Mary River catchment in the Northern Territory. The study used a combination of morphological and molecular characters for species identification of a subset of fungal endophytes. The partial translation elongation factor 1- α sequence (EF1-α) and internal transcribed spacers (ITS) regions were the gene regions used for sequencing.

4.3 Methods

4.3.1 Study Area and Field Sampling Design

Stem sampling was conducted as part of the ecological studies aimed at selecting four locations to represent the four dieback classes observed during a reconnaissance visit to the target area in August 2011. The dieback classes used were: healthy, dieback, recovering from dieback, and dead. Details of the field sampling design and site selection are outlined in Chapter 3 (Section 3.3.1). Four locations were selected on Melaleuca Station, each comprising four plots. The concept of the plots was to represent the stratification of dieback expression observed on the sites. Stem samples from adult plant collected represented the four dieback classes. The stems collected were used for fungal isolation to compare the frequency of recovered isolates across time and space.

4.3.2 Specimen Collection

For each plot on Melaleuca Station, destructive stem sampling was done on three adult mimosa plants (> 1.5m tall) found just outside the permanent plots representing each of the four dieback classes. Using secateurs, a 20 cm-long stem piece was cut from each plant (Figure 4.1). Stem pieces sampled were 30 mm ± 5mm thick and were part of a living branch which was free from frass. Samples were not taken from dead branches or from stems with evidence of insect damage except for the “dead” plots during the first survey. In the subsequent sampling where the initially labelled “dead” plots recovered and became healthy, stems sampled were taken from living branches and were free from 62

the evidence of insect damage. Individual stem pieces were put into separate paper bags, labelled and stored in a cool room (~4-5oC) until isolation which was done within two weeks of collection. The blades of the secateurs were wiped clean between sampling individual treatment plots, using commercial wet wipes containing isopropyl alcohol.

a b

Figure 4.1 Adult plant stem collection (a); Stem branches cut from adult dieback plots (b)

4.3.3 Fungal Isolations

A total of 169 stem pieces from different plants expressing dieback were processed within two weeks of storage. External and internal disease symptoms were recorded for each stem piece. Isolations were done in batches commencing with the healthy stems. Stems with corky bark were washed thoroughly with running water before further treatment.

A manual mitre saw was used to cut six 1cm thick mimosa sections from each of the 20 cm long stem pieces (Figure 4.2). The sections were moved to a laminar flow hood where they were surface sterilised with 2% NaOCl solution for two minutes, rinsed twice by stirring in sterile water, and then dried between sheets of sterile paper towels. Using sterile forceps, two stem pieces were then plated onto ½ strength potato dextrose agar (Difco), acidified malt extract agar, tap water agar and V8 agar media. These were amended with penicillin G (200µg/ml) (Sigma PEN-NA) and streptomycin sulphate (100µg/ml) (Sigma S6501). The plates were sealed with Parafilm® and incubated at 25ºC in the dark and observed regularly for fungal growth. Subcultures were prepared from the fungi growing out from the edge of the stem pieces. Each fungal colony was transferred to ½ PDA medium for further identification.

a b

c d

Figure 4.2 Fungal isolation methodology. Cutting into stem pieces (a); Debarked stem pieces; (b) Plating the stem pieces into different media (c); fungal culture growth (d)

For stem sections yielding multiple colonies, the slower growing fungi were selectively sub-cultured before they were over-grown by faster growing colonies. All fungal isolates were numbered from 001 and given the code Mp(X). For the first sample time the code was Mp, followed by Mp1 then Mp2, respectively. The effectiveness of the surface-sterilisation protocol was confirmed by imprinting randomly selected surfaced-sterilised stem pieces in separate petri dishes (Schulz et al. 1993).

4.3.4 Identification of Recovered Isolates

Identification of the fungal isolates recovered from stem samples was based on the use of a combination of morphology and phylogenetic characters. Three steps were followed for final identification and final designation of isolates. First, all the 1,352 recovered isolates were morphologically assigned and identified to the nearest taxon possible (usually a genus). Then, a subset of 90 isolates to be used for Internal Transcribed Spacer (ITS) and 67 isolates for EF1-α gene regions, were chosen for further phylogenetic identification. Finally, a “rationalisation table” was used to assign each isolate a final identification. The rationalisation table included a combination of two characters and a decision on whether they could be confidently assigned a species level identification.

4.3.4.1 Morphological Identification

Sub-cultures were incubated in the dark at 25ºC for seven days for (fast- growing isolates) and more than two weeks for (slow- growing isolates) before being assigned to morphological groups based on colony surface texture, colony colour (above and below), elevation, radial diameter, margin shape, growth rates and sporulation structure. For radial diameter, two measurements at 90o to each other were taken if the culture had an irregular margin every twenty-four hour interval (Bold et al. 1980). The “Illustrated Guide of Imperfect Fungi”, (Barnett & Hunter 1998) taxonomic keys and mycological colour chart were used to aid in the identification (Rayner 1970). A Vernier caliper was used to measure the diameter of cultures every 24 hours during growth. Colour changes on the front and back of the plate was also noted during the 2 weeks incubation in the dark. When cultures were beginning to sporulate, a dissecting microscope was used to check the structures on the plate. A microscope slide was prepared by removing the sporulating structure using a sterile scalpel and placing this in a drop of lactophenol cotton blue; the structures were then teased apart before a cover slip was added. The scotch tape method (Ramaling & Rati 1974) for slide preparation was used for fungal isolates that produced free spores. Finally, a representative of culturally grouped isolates was confirmed and identified to the nearest taxa by Dr Ken Goulter2, by simultaneously comparing the microscopic structure observed with the culture on PDA plate. The Illustrated Guide of Fungi Imperfecti (Barnett & Hunter 1998) was used to compare the spores seen under the light microscope. Non-sporulating cultures were grouped separately and separate attempts were used to induce sporulation.

Induction of Sporulation

Some groups, like Pestalotiopsis type spp., Fusarium- type spp. and Colletotrichum- type spp., readily sporulated when incubated on ½ strength PDA at 25ºC in the dark for several weeks. Many of the other fungi did not sporulate after extensive incubation under these conditions. To induce sporulation, isolates thought to belong to the family Botryosphaeriaceae were grown on pine needles, as described in the literature (Pavlic et al. 2008; Abdollahzadeh et al. 2010; Begoude et al. 2010). A representative of each Botryosphaeriaceae morphospecies was subcultured on WA (water agar), then six segments of autoclaved pine needles were laid in parallel on a plate. Each plate was sealed with Parafilm®™, put in a ziplock bag and then incubated under black light for thirty days. Morphospecies

2 Dr Ken Goulter is a plant pathologist and molecular biologist. He is the Head of Product Development and Operations of BioHerbicides Australia. He is also a thesis advisor. 65

thought to belong to the genera Phomopsis/Diaporthe were grown on oatmeal agar and incubated at 25ºC in the dark until sporulation (Figure 4.3).

b a

Figure 4.3 (a) Sporulation using pine needles; (b) sporulation on oatmeal agar.

4.3.5 Microscopy

Conidia and mycelium of a representative on each group of fungal cultures were photographed using a Canon Digital Camera under 400x magnification. Conidia (n=20) length and width of two representative mature and conidiogenous cell of the representative morphotypes were measured and recorded. Conidia color, septation and striation, were also recorded. Examples are shown in Figure 4.4. Multiple attempts were made to promote sporulation of the different Botryosphaeriaceae morphotype groupings. Of the 10 groups, only 5 sporulated on pine needles, showing mature spores under a microscope. The other 5 groups either did not sporulate after multiple attempts or only produced immature spores after a period of more than one month (Appendix D).

4.3.5.1 Molecular Identification

Molecular identification included the sequencing of two gene regions of chosen isolates. A total of 40 Botryosphaeriaceae isolates were sequenced in both gene regions. A total of 90 isolates were sequenced using the ITS and 67 isolates were sequenced using the EF1-α gene regions.

Selection of isolates used in molecular techniques

Given the frequency data of the temporarily assigned isolates through morphology, the selection of isolates used for sequencing was based on the following weighted selection scheme: differences in cultural morphology, replication plots and sampling time. Non-sporulating and unidentified isolates were included. Two isolates were randomly selected from within each of these criteria.

a b

Figure 4.4 Conidia of Pestalotiopsis type (a) and Botryosphaeriaceae (b). Scale white bar = 10µm.

4.3.5.2 DNA Extraction, Amplification and Sequencing

Mimosa pigra isolates were grown on half strength potato dextrose agar PDA (Difco TM Becton, Dickinson and Company, Sparks, USA, 19.5 g PDA, 7.5 g agar and 1 L of distilled water) for 4-5 days in the dark at 25°C. The mycelia were harvested by scraping an ethanol and flame-sterilised, cooled glass slide across the plate, taking care not to scrape up the agar medium. Harvested mycelia were transferred to labelled 1.5 ml sterile microtubes which were stored at -80°C until DNA extraction. Mycelia were ground using a sterile micropestle, and fungal genomic DNA was extracted according to the protocol given in the Rapid Fungal Genomic Isolation Kit (Bio Basic Inc). The region spanning the internal transcribed spacer (ITS1-5.8S- ITS2) region of the ribosomal DNA was amplified using the primers ITS1 and ITS4, and EF1-α gene was amplified using the primers EF1- 688F and EF1-1251R. Each 50 µl PCR reaction mixture included 10 µl of 5x My Taq buffer, 33.5 µl of distilled water, 2 µl of DNA template, 0.5 µl My Taq, and 2µl of each primer. Gel electrophoresis was done to check if the amplification worked for all samples. The bands were viewed under a UV Tran illuminator and then photographed. Successfully amplified samples were purified using Wizard® SV Gel and PCR Clean-Up System (Promega), and then quantified using a BioSpec-nano

spectrophotometer (Shimadzu). Samples were adjusted to a concentration of 50ng/µL before being sent to MACROGEN Inc (Seoul, South Korea) for Sanger sequencing.

4.3.6 Phylogenetic Analysis

Sequences (ITS and EF’s) were cleaned and edited using Sequencher 5.2.4 (registered) (Gene Codes Corporation). Since the elongation factor (EF1-α) region amplicons were sequenced in reverse direction, each of the sequences were first reverse complemented using an online tool called The Bio- WEB (http://cellbiol.com/). Multiple sequence alignment was done using Clustal Omega version 1.2.1 at the EBI site (http://www.ebi.ac.uk/tools/msa/clustalw2). The alignments included FASTA file of the holotypes, or ex-type and outgroup sequence isolates downloaded from GenBANK. Phylogenetic analyses were performed in PAUP (Phylogenetic Analysis Using Parsimony)(Swofford 2003). An alignment file was viewed on TreeBASE (http://www.treebase.org/).

Maximum parsimony analyses were performed using 1000 random taxa addition with branches swapped using the tree bisection and reconnection (TBR) algorithm. The heuristic search option was used with unordered characters of equal weight, and gaps were treated as missing data. Maxtrees was set to 100 with automatic increase if needed. All multiple equally parsimonious trees were saved. The following indices of characters were determined: CI (consistency index) calculated by dividing m = minimum amount of change that the character may show on any conceivable tree, by s = length (number of steps) required by the characters on the tree being evaluated; and RI (retention index) calculated from (g-s)/ (g-m) where g = maximum possible amount of change that a character could possibly require on any conceivable tree. The robustness of the most parsimonious trees was evaluated from a bootstrap analysis using 1000 replications.

4.3.7 Final Designation of the Isolates ID Using Morphology and Genetic Data.

Using an excel spreadsheet where the details of the 1,352 isolate data entries (both morphological and genetic characters) were outlined, data were first assessed in relation to whether the type of media, the sampling time, location and replicates, were making a difference on the frequency and the groups of morphotypes and morphospecies identified. To differentiate between the morphotypes, the characters used to discriminate between isolates were: colour, colour change with age, growth rate and the type of spore and spore dimensions (when mature conidia were seen). Given that only a subset of the 1,352 isolates were sequenced in both gene regions, a “rationalisation table” was used to propose the final designation of the isolated M. pigra endophytes, using combined morphology and phylogenetic characters identified. The “rationalisation table” included the lumping of species, especially in cases of difficulties encountered when deciding which species would be used for 68

identification, given one morphotype grouping yielded two or three different species in GenBANK searches (See Appendix E for the rationalisation table of identification).

In performing the Basic Local Alignment Search Tool (BLAST) searches (Madden 2002), each of the sequences were carefully checked for accuracy by downloading the published papers and double- checking whether individual references compared their method of identification to a holotype or to an ex-type specimen. Where an isolate matched both the morphology and the 2 gene sequences, then the species ID was used in the subsequent analysis. In a case where both ITS and EF alpha gave 2 different species, the species name under the EF-1α identification was given more weight. However, the phylogram and the similarity tree were given the highest priority in making the final decision of identification used for subsequent analysis.

In the Internal Transcribed Spacer (ITS) phylogenetic tree, nine M. pigra sequence data were analysed by Dr Andrew Bissett, a research scientist at the Commonwealth Scientific Research Organisation (CSIRO) in Canberra, Australia. The M. pigra pure cultures were supplied by our team from the University of Queensland.

4.3.8 Storage and Maintenance

Selected representative isolates were maintained in sterile water storage (Smith & Onions 1983) until needed. Four or five pieces from the margin of actively growing colonies were transferred aseptically to 20mL of sterile RO water in 30mL plastic tubes. General cultures that were being actively used in seedling and glasshouse pathogenicity trials were subcultured on PDA plates and kept in a cold room until needed.

4.4 Results

4.4.1 Fungal Endophytes of M. pigra

Out of 169 plants screened in this study, a total of 1, 352 isolates (total count for the four sample times) were purified into individual axenic cultures. All plants contained fungal endophytes (incidence=100%). Out of 1,352 isolates, 24 morphotypes were identified, based on differences in cultural morphology on ½ PDA and sporulation. Within the 24 assigned morphotypes, a total of 67 representative isolates were sequenced using the partial translation elongation factor 1- α sequence (EF1-α) and 108 internal transcribed spacers (ITS) regions. Frequently recovered taxa were from Botryosphaeriales, Xylariales and Diaporthales. No sexual (teleomorph) structures were observed during this study.

Sclerotinia . MP490 (L.theobromae) MP115 (L.theobromae) MP046 (L.theobromae) MP689 (L.theobromae) MP1831 (L.theobromae) MP80 (L.theobromae) MP1321 (L.theobromae) MP869B (L.theobromae) MP682 (L.theobromae) MP19 (L.theobromae) MP220 (L.theobromae) MP84 (L.theobromae) MP669 (L.theobromae) MP853 (L.theobromae) MP1319 (L.theobromae) MP18 (L.theobromae) MP1320 (L.theobromae) MP312b (L.theobromae) MP117 (L.theobromae) MP036 (L.theobromae) MP113 (L.theobromae) MP15 (L.theobromae) MP343 (L.theobromae) Botryosphaeriaceae MP1202 (L.theobromae) MP1318 (L.theobromae) MP112 (L.theobromae) MP304 (L.theobromae) MP039 (L.theobromae) MP10 (L.theobromae) MP107 (L.theobromae) MP212 (L.theobromae) MP64 (L.theobromae) MP296b (L.theobromae) MP202 (L.theobromae) MP1143 (L.theobromae) MP68 (L.theobromae) MP353 (L.theobromae) MP188 (L.theobromae) MP285 (L.theobromae) MP284 (L.theobromae) MP384 (L.theobromae) MP586 (L.theobromae) MP1200 (L.theobromae) MP1171 (L.theobromae) MP1201 (L.theobromae) MP291 (L. pseudotheobromae) MP129 (L. pseudotheobromae) MP315 (L. pseudotheobromae) MP272 (L. pseudotheobromae) MP333 (L. pseudotheobromae) MP298 (L. pseudotheobromae) MP1260 (L. pseudotheobromae) MP088 (L. pseudotheobromae) MP432 (L. pseudotheobromae) MP681 (N.parvum) MP1209 (C.eucalypti) MP1790 (C.atrovirens) MP290 (C.eucalypti) MP1464 (C.eucalypti) MP596 (Uncultured) Pleosporaceae MP864 (Uncultured) MP296 (A.alternata) MP299 (A.alternata) MP294 (C.cladosporoides) MP305 (C.cladosporoides) Davidiellaceae MP293 (C.cladosporoides) MP311 (C.cladosporoides) MP1614 (P.clavispora) MP289 (P.clavispora) MP287 (P.clavispora) Amphisphaeriaceae MP1748 (P.clavispora) MP1678 (P.clavispora) MP1716 (P.clavispora) MP1770b (P.clavispora) MP1682 (P.clavispora) MP301 (Fusarium sp.) MP1402 (Fusarium sp.) MP286 (Fusarium sp.) Nectriaceae MP1126 (Fusarium sp.) MP544B (Fusarium sp.) MP292 (Fusarium sp.) MP307 (Fusarium sp.) MP1611 (Phomopsis sp.) MP295 (Phomopsis sp.) MP288 (D.phaseolorum) MP312 (D.phaseolorum) MP308 (D.phaseolorum) MP1263 (D.phaseolorum) MP1115 (D.phaseolorum) MP300 (D.phaseolorum) Valsaceae MP206 (D.phaseolorum) MP303 (D.phaseolorum) MP01 (D.phaseolorum) MP020 (D.phaseolorum) MP078 (D.phaseolorum) MP297 (D.phaseolorum) MP213 (D.phaseolorum) MP306 (D.phaseolorum) Figure 4.5 Phylogenetic tree of M. pigra based on the bootstrap analysis of ITS sequences. The tree was rooted to Sclerotinia sclerotiorum.

4.4.2 Agreement between Morphological and Phylogenetic Characters

All morphotype groupings were validated by genetic sequencing. Morphology groupings were lined up with the molecular sequencing ID (Figure 4.5). All ten assigned Botryospaeriaceae groupings were confirmed to be within the family Botryosphaeriaceae. The Pestalotia groupings were under the correct taxa = Pestalotiopsis spp. The morphologically grouped Diaporthe/Phomopsis were confirmed to be two species of Diaporthe using the ITS and EF-1α sequence data (Table 4.1). The most prominent taxa recovered from M. pigra are the species under Botryosphaeriaceae, followed by the taxa Pestalotiopsis spp. and Diaporthe/Phomopsis spp. All the other taxa listed in Table 4.1 were rarely isolated over time.

Table 4.1 Fungal endophyte communities in Mimosa pigra over time.

Taxa ID (using combined morphological and Assigned Morphotype Sample times/ Isolate count phylogenetic characters) Botryosphaeriales: Botryosphaeriaceae Dry 2012 Wet 2012 Dry 2013 Wet 2013 Total Count Lasidioplodia complex (includes L. Botryo Groups: theobromae, L. pseudotheobromae, L. A,B,C,D,F,H,I; 164 192 115 131 602 hormozganensis, L. iranensis ) Unidentified #2,#4 Neofusicoccum parvum N. parvum 1 0 0 0 1 Cophinforma spp. (includes C. atrovirens, Botryo E 14 13 9 21 57 C. eucalypti ) Botryosphaeriaceae sp. Botryo G,J 6 0 8 5 19 Diaporthales: Valsaceae

Diaporthe spp. (includes D. phaseolorum, Diaporthe/Phomopsis 102 86 51 23 262 D. longicola, D. arengae. Diaporthe sp.) A,B

Xylariales: Amphisphaeriaceae Pestalotiopsis spp. (P. clavispora, P. Pestalotia A,B 165 76 80 37 358 ellipsospora, Pestalotiopsis sp. ) Hypocreales: Nectriaceae Fusarium spp. ( F. pseudocircinatum , Fusarium spp. 11 2 7 2 22 Fusarium sp.) Pleosporales: Pleosporaceae Alternaria spp. Aternaria sp. 9 0 1 2 12 Sordariomycetidae: Glomerallaceae Colletotrichum sp. Colletotrichum sp. 3 6 3 0 12

Unidentified Unidentified #5 0 0 0 7 7

Total Isolate count 475 375 274 228 1352

4.4.3 Taxa of Mimosa pigra Endophytes

4.4.3.1 Botryosphaeriaceae

There were at least seven confirmed species of Botryosphaeriaceae, based on the phylogenetic characters used (Table 4.1) these species were: Lasiodiplodia theobromae, L. pseudotheobromae, L. hormozganensis, L. iraniensis, Neofusicoccum parvum, Cophinforma eucalypti, and C. atrovirens. The genera Neofusicoccum and Cophinforma were reliably identified as the assigned morphotypes by validation using the DNA sequence data (ITS and EF1-α). Maximum parsimony tree, based on the elongation factor data sequence, also supports this result (Figure 4.6). The taxa ID “Lasiodiplodia complex” was both morphologically (Figure 4.8) and genetically diverse (Figure 4.6 and 4.7). It is evident that there was over-splitting of the ten (10) Botryosphaeriaceae groupings based on morphological characters and that the EF1-α was necessary to discriminate between species. The final species designations, for cryptic and sibling species, cannot be based on the use of the morphological characters used in this study nor ITS sequence data of the different morphotypes of this taxa. ITS sequencing classifies morphotypes mostly as Lasiodiplodia theobromae (Figure 4.5).

Diplodia seriata CBS 112555

L. pseudotheobromae

L.pseudotheobromae

L. hormozganensis

L. iraniensis

MP689 L. theobromae L. theobromae C. eucalypti

(N.parvum) N.parvum

L. hormozganensis

L. pseudotheobromae

L. iraniensis

- N.parvum - C.atrovirens

C.eucalypti

a b c

d e f

g h i

j k l

a b c

d e f

g h i

j k l

The characters used in the study to distinguish between the 10 Botryosphaeriaceae morphotypes were - colour, pigmentation, growth rate differences, and the consistency of the aerial mycelium. These characters were not reliable and were considered a rather crude basis for assigning M. pigra fungal endophytes into taxa. The mature conidia measured were also not useful for the morphotype groupings (Figure 4.8 k to l). There were twenty-six isolates (Table 4.1) that could not be assigned to any family using morphological characters (Figure 4.9 j to l).

4.4.3.2 Pestalotiopsis spp.

There were two different morphotypes of Pestalotia (A and B) grouped. However, genetic sequencing showed that each grouping (A and B) has 2-3 different species. The morphological differences between the two groupings is that Pestalotia Group A was fast-growing with irregular colony margins, while Pestalotia Group B was slow-growing and has a circular margin. (Figure 4.9 a & b). Both have the same colour and colour change with age. EF sequencing showed that for each morphotype grouping, either identifies as P. clavispora or P. ellipsospora when analysed (Appendix F).

4.4.3.3 Diaporthe/Phomopsis

Two morphotypes were grouped for Diaporthe. Using morphological characters, the difference between the two is related to colour change with age and the colony margin. Both sporulated and have beta conidia (Figure 4.9 d & e). Sequencing of each morphotype resulted in two species being identified, D. phaseolorum and Phomopsis sp.

4.5 Discussion

This study was conducted to identify and determine the occurrence of fungal endophytic communities in M. pigra populations, using a combination of morphology and phylogenetic characters for species identification. The results show that M. pigra represents a diverse niche of fungal communities. The most commonly isolated taxa throughout the year were from the members of the family Botryosphaeriaceae and from the genera Pestalotiopsis and Phomopsis/Diaporthe. These findings are consistent with the results of an earlier study by Wilson and Pitkethley (1992), who isolated Botryodiplodia theobromae = Lasiodiplodia theobromae from diseased shrubs of M. pigra in the Adelaide River catchments in the Northern Territory. Changes in the rules of fungal nomenclature have affected the identification of fungal endophytes when compared to Wilson’s work (Hawksworth 2012). It is possible, therefore, that B. theobromae is a species complex which can be accurately discriminated within, depending on the methods employed in isolation and identification (Alves et al. 2008).

Apart from the three dominant taxa, the results of this study confirmed that fungi from the genera Alternaria, Colletotrichum, Cladosporium, Fusarium and Epicoccum were also present as endophytes, but were rarely isolated. These results are supported by those of (Strobel et al. 1996; Arnold et al. 2000; Lacap et al. 2003; Abdollahzadeh et al. 2010; Suciatmih & Rahmansyah 2013). Members of these taxa are important endophytes associated with different woody plants.

Perhaps a disadvantage of the classical method used in assigning the morphotype groupings in this study, the culture based approach for endophyte identification, is that it is a method dependent on a selective process, because endophytic fungi obtained from plant tissues are subject to surface sterilisation techniques, incubation conditions, the ability of the fungus to sporulate in culture, and the competence of the person doing the identification (Guo et al. 2000). In the present study, there was a high frequency (15%) of unidentified and sterile forms. However, to resolve the issue of identification of the fungal isolates, the ITS region of the ribosomal DNA and the partial translation elongation factor were sequenced. Using the combination of morphology and genetic characters allowed the confident identification of the M. pigra endophytes into higher ranks (family and genus). However, most morphology groupings under Botryosphaeriaceae were not defined to the species level using ITS sequencing. Phillips et al. (2013) discussed this, indicating that although ITS alone is sufficient to separate species within a genus of Botryosphaeriaceae, the inclusion of EF1-α results in a more robust separation, and is considered essential in some genera such as Diplodia, Lasiodiplodia and Neofusicoccum. The phylogeny of members of the Botryosphaeriaceae is confusing (Phillips et al. 2013) and the use of morphological characters is not sufficiently reliable to allow discrimination between species (Alves et al. 2008; Urbez-Torres & Gubler 2009).

One major disadvantage of the molecular approach, as identified by Ko et al. (2011), is that caution is needed in naming species, especially when using BLAST searches on public databases like GenBANK. Errors in identification using this tool arise from the fact that there are only a small number of fungal sequences available on these databases, and that the reference sequences deposited have not been compared with holotype and ex-type cultures.

The unidentified groupings (Fig 4.9 j-l) may indicate that these ungrouped isolates, which were phenotypically different from Botryosphaeriaceae but which were under the “Lasiodiplodia complex” when sequenced, can be due to the presence of mycoviruses (Zhang et al. 2009) of fungal endophytes. Mycoviruses (fungal viruses) which are recognised as important components of plant biosystems, are very common in all major fungal groups (Ghabrial & Suzuki 2009).

In future investigations, it is suggested that when performing stem isolation for M. pigra, the most appropriate selective medium yielding the highest number of “different” species is acidified malt extract agar (aMEA). The other three media yielded fewer endophytes when compared to aMEA. Whipps (1987) showed that the choice of media in isolation work is important.

In adopting the use of culture-based techniques, it is possible that the number of endophytes may be significantly higher than was documented in this study. This approach may overlook the presence of unculturable or slow-growing species (Hyde & Soytong 2008). To resolve this potential uncertainty, investigations based on culture-independent, direct molecular analyses, and the use of the micro-array technique, need to be adopted to provide a better picture of the fungal endophytes associated with M. pigra in the Northern Territory.

4.6 Conclusion

This results of this study show that M. pigra harbours diverse endophyte communities. The method used for identification gives confident taxonomic placement of M. pigra endophytes at the genus level. The use of morphological characters was crude and unreliable. However, species identification was not discriminated against when using the ITS sequence data. The results of the study show that, despite the limitations, taxonomic placement based on EF1-α allowed the discrimination between sibling species, especially in the confusing taxa, Botryosphaeriaceae.

CHAPTER 5

Fulfilling Koch’s postulate confirms the pathogenicity of fungal species recovered from healthy and dieback-affected Mimosa pigra populations in the Northern Territory, Australia.

5.1 Abstract

Members of the Botryosphaeriales were associated with dieback in Mimosa pigra in the Adelaide and Mary Rivers in the Northern Territory. Typical dieback symptoms associated with M. pigra dieback are: reddish- brown lesions originating from leaf axils, yellowing of the leaves and subsequent stem death. This chapter aimed to investigate the pathogenicity of isolated fungal endophytes against seedlings and juvenile M. pigra plants in a controlled conditions and adult plants in field conditions. Among the subset of isolates tested, Lasiodiplodia species can cause seedling mortality and aggressiveness to juvenile and mature plants in controlled and field conditions, resulting in the development of external and internal stem lesions during the trial period. In contrast, isolates belonging to Pestalotiopsis type species, and most of the Diaporthe species and Fusarium species, recovered from the dieback sites did not produce disease symptoms or cause seedling mortality. This study is the first to identify that fungal endophytes from both healthy and dieback-affected M. pigra showed pathogenicity to healthy plants confirming Koch’s postulates. Further studies are needed to elucidate the biological cycle of these fungi, determine the appropriate inoculum concentration, define techniques for field application and the prolongation of the shelf-life of the inoculum, to develop a fuller picture on how these isolates can be utilised as mycoherbicide treatments for M. pigra.

5.2 Introduction

Mimosa pigra L. is a woody perennial shrub native to tropical America and is listed in the Global Invasive Species Database as one of the One Hundred of the World’s Worst Invasive Alien Species (Heard & Paynter 2009). It was first noticed in Australia in 1952 in the Adelaide River System of the Northern Territory (Miller & Lonsdale 1987) and is now widespread in the Top End of the Northern Territory and has been declared a Weed of National Significance (WONS) (Thorp & Lynch 2000).

Dieback of M. pigra was first reported on the Adelaide River floodplains in 1988 (Miller 1988) and has since been reported in other large infestations of M. pigra in the Top End. The fungus Botryotinia (=Lasiodiplodia) theobromae (Pat.) Griff & Maubl. (Botryosphaeriaceae) was isolated from typical dieback symptoms: reddish- brown lesions originating from leaf axils, yellowing of the leaves and

subsequent stem death (Wilson & Pitkethley 1992). Dieback has continued to be observed in the Adelaide, Mary, Finniss and Daly River catchments in the Northern Territory (Natasha Burrows, Weed Sciences Branch of the Northern Territory Department of Land and Resource Management pers comm). A survey of M. pigra stands conducted in 2012 discovered evidence of population (stand) dieback at 52% of sites assessed in the Daly River catchment, and 20% of sites in the Finniss River catchment (DLRM unpublished data). Large areas of dieback (>10 hectares) are readily visible in the Adelaide, Daly and Mary River catchments (L. Elliott, pers. comm.).

Globally, fungal endophytes have been found in almost every plant and plant organ that has been studied. A variety of associations exist between endophytes and their host plants, spanning the continuum from biotrophic mutualists and benign commensals, to necrotrophic, antagonistic pathogens (Bacon & White 2000; Arnold et al. 2003; Mejia et al. 2008). In Chapter 4 of this thesis, it is reported that dieback-affected and healthy M. pigra stands in the Northern Territory have similar endophytic communities, with the species most frequently isolated belonging to the genera Lasiodiplodia, Diaporthe and Pestalotiopsis. Field surveys of endophytes/pathogens have concentrated only on the diversity and sources of the isolates over time. However, there is a need to clearly determine the relationships of the recovered isolates from both healthy (asymptomatic) and dieback-affected (symptomatic) M. pigra plants, by fulfilling Koch’s postulates to determine if there are differences in the frequency of pathogenic species.

Latent infection of plants by pathogens is the state where a host is infected by a potential pathogen and the signs of infection are absent. Latent infection caused by Lasiodiplodia species, Diaporthe species and Pestalotiopsis-type endophytes of grapevines, multiple fruit/nut and woody plants, are well documented (Photita et al. 2004; Pavlic et al. 2007; Jami et al. 2013; Pitt et al. 2013), but the evidence for latent infection of noxious woody weeds in Australia is still scarce. What is not clear is the nature of the interaction of endophytes and pathogens obtained from healthy and dieback tree samples when inoculated back to plants in controlled and field conditions. Determining this interaction is an important prerequisite to identifying a potential fungal biocontrol agent for this weed.

The term “endophyte” simply describes a lifestyle. Observational studies, such as those of (Photita et al. 2004; Moricca & Ragazzi 2008), found that pathogenic fungi can survive as symptomless endophytes in plants, but with the right timing and stressors, they can change from symptomless colonisers to pathogens that cause disease symptoms. Phylogenetic evaluation studies have suggested that some endophytic species change their ecological strategies and adopt a saprotrophic lifestyle (Promputtha et al. 2007). Considering the fact that endophytes are continuously interacting with their hosts, these associations play a key role in understanding the lifestyle of these fungi. This

thesis chapter therefore aims to determine the pathogenicity of a subset of fungal isolates against seedlings, juvenile and adult M. pigra plants, under controlled and field conditions.

Research Questions

Experiment 1: Pathogenicity of selected fungal isolates against Mimosa pigra seedlings.

Test 1: Do fungal isolates recovered from dieback sites cause disease when used to inoculate healthy M. pigra seedlings?

Experiment 2: Pathogenicity of selected fungal isolates against juvenile Mimosa pigra plants under glasshouse conditions.

Test 2: Do selected isolates recovered from M. pigra stems collected from the field have the same degree of pathogenicity when used to inoculate healthy juvenile plants in the glasshouse?

Experiment 3: Pathogenicity of selected fungal isolates against M. pigra plants growing in the field.

Test 3: Do selected isolates recovered from M. pigra stems have the same degree of pathogenicity against each other when used to inoculate healthy adult plants growing under field conditions?

The hypotheses were tested by fulfilling Koch’s postulates in seedlings and juveniles in controlled conditions and in adult M. pigra plants in field condition. External and internal lesion development was measured to determine the differences in the pathogenicity of the subset of isolates. Plant health rating, stem diameter and height difference before and after the trials, were also used to examine the differences in pathogenicity.

5.3 Materials and Methods

5.3.1 Collection of Samples

Samples were collected from 186 M. pigra plants at four different sites on Melaleuca Station in the Mary River Region of the Northern Territory, in 2012 and 2013. The sites are described in detail in Chapter 3 Section 3.3.1. Plant health at the sites ranged from plants showing no symptoms of dieback to sites where plants had branches showing dieback symptoms. The destructive stem sampling of plants was done on M. pigra plants found just outside the permanent plots. Using secateurs, a 20 cm- long stem piece was cut from each of three adult plants on the outside each plot.

5.3.2 Fungal Isolations

A total of 186 fungal isolates were obtained from 186 M. pigra stems during sample time 1(June 2012) and sample time 2 (November 2012). The fungal isolation procedure is outlined in Section 4.3.3 of Chapter 4 and illustrated in Figure 5.1 below.

Healthy Dieback Recovering Dieback Dead

B 1cm ± 0.2cm Stem Cuttings

2.5cm ± 0.5cm

C. Initial Isolations

1/2PDA WA V8 aMEA

Two-Replicate Seedling Trials (Rapid Preliminary Screening)

Ten-Replicate Seedling Trials (+NT 039/Positive Control and Negative Control)

Selected Top 25 Pathogenic Isolates Inoculated to Glasshouse Juvenile mimosa Plants

Field Infection Trial on Selected Mimosa Isolates

The number of isolates was reduced to 255 after grouping them on the basis of morphological features, replication and level of dieback expression. For each group, representative isolates were selected and used for further pathogenicity studies. All 255 isolates were tested in a preliminary pathogenicity screening trial involving two replicates per isolate.

Experiment 1: Pathogenicity Testing of M. pigra Seedlings

5.3.3.1 Seedling Germination

M. pigra seeds (Figure 5.2) were acquired from Natasha Burrows (Weed Sciences Branch of the Northern Territory Department of Land and Resource Management). Seeds were pre-germinated to ensure that only viable seeds were used for seedling pathogenicity testing. The seeds were first surface sterilised by soaking in 2% NaOCl solution for five minutes, followed by rinsing with sterile water. They were then transferred to sterile water at a temperature of 80°C for 4 hours, and then allowed to cool to room temperature. They were then incubated at 25°C for 2 days in Petri dishes lined with moist paper towels. Water was added daily to maintain sufficient moisture. After 48 hours, germinated seeds with root radicles extending at least ½ the length of the seed were chosen for the pathogenicity assay.

A B

Figure 5.2 M. pigra seeds: individual seeds with bristle cover (A); seeds in pods (B).

5.3.3.2 Inoculum Preparation

French white millet (Panicum milacium) seeds were rinsed twice and soaked in deionised water in a 1000 ml beaker for 24 hours at room temperature before being drained and rinsed again. Then approximately 20g of soaked millet was transferred to 30mL McCartney tubes (80 x 27mm) (www.technoplas.com.au) and autoclaved twice over 24 hours at 121°C for 25 minutes.

Using a scalpel, a 10mm x 10mm agar block was cut from the growing margin of each subcultured fungal isolate, the blocks being then transferred into the McCartney tubes. These tubes were incubated at 25°C in the dark. The tubes were shaken every 2 days to redistribute the grains and obtain even colonisation. The inoculum was dried when the fungus was observed to have fully colonised the millet. The caps of the tubes were replaced with sterile tissue paper held in place with a rubber band. Drying was done by placing the tubes under a vacuum in a desiccator containing indicator silica gel for at least 48 hours. The caps were replaced and the samples were stored at 4oC until used.

5.3.4 Preliminary Screening of Isolates and Trial Assessment

Preliminary screening was done to rapidly assess which, among the fungal isolates, displayed pathogenicity to M. pigra seedlings within a 14 day period. The experiment comprised units of two seedlings inoculated with a fungal isolate and one negative control of sterile millet. The screening was done in batches with 25 isolates per batch. This was only done once. The purpose of this was to reduce the number of isolates for pathogenicity testing.

For inoculation, approximately 20 ml of moist vermiculite (Grade#3) was placed in 30mL plastic screw-capped tubes (www.technoplas.com.au), in which a 3mm hole had been drilled in the base to facilitate drainage. The tubes were autoclaved. Using sterile forceps, one germinated seed was planted into a hole dug into the vermiculite; three grains of colonised millet were placed around the planted seed and then the seeds were covered with additional sterile moist vermiculite. The same was done with the negative control, except that sterile uncolonised grains were used. The tubes were then transferred onto plastic mesh over a 1.5 cm spacer in the base of plastic boxes (Décor). This allowed the tubes to freely drain. The plastic boxes were sealed and then incubated in a constant temperature chamber at 25°C. The tubes were watered with sterile RO water every 2 days and a general assessment of seedling health was done every 4 days. Disease assessment was qualitative. Visual observations of physical symptoms were made after 4, 8, and 12 days of incubation. A final assessment of seedling health was done after 14 days. A qualitative rating scale of 1-4 was used (0 = Pre-emergent death; 1= Post-emergent death; 2 = Emergence but showing significant disease symptoms; 3= Emergence but showing few symptoms; 4= Healthy and no symptoms). At the end of the trial, the numbers of dead and living seedlings were recorded.

Of the 255 isolates screened in the two-replicate seedling assay (a trial used for rapid preliminary screening to reduce a very large numbers of isolates for pathogenicity trials), 71 isolates caused some seedling mortality. These isolates were retested using 10 replicates and a positive control Lasiodiplodia iraniensis NT039 (recovered from dieback-affected Parkinsonia aculeata) which had

previously been demonstrated to be pathogenic to M. pigra seedlings (Dr Ken Goulter; pers comm), and a negative control (comprising sterile uncolonised millet grains). The trial was undertaken exactly as described above. Two duplicate trials were undertaken for this 10-replicate assay as repeat trial was undertaken to assess pathogenicity of 71 isolates. Daily observations made to check and record whether there were seedlings that did not grow at all and, occasionally, whether the radicle had emerged from the surface of the vermiculite. Such seedlings were removed from the experiment. If the incidence of root emergence was >40% then the isolate screening was redone. Visual observations were made on days 4, 8, and 12 of incubation. On the 14th day a final assessment was made of the seedlings. Isolates showing negative pathogenicity were excluded from the presentation in the results section. List of isolates used in the trials were shown in Appendix I.

5.3.5 Confirmation of Infection in Seedling Trials

Lesions from infected seedlings were plated on ½ PDA plates amended with penicillin G and streptomycin sulphate and the resultant colonies were compared with the original cultures to confirm pathogenicity and fulfil Koch’s postulates. Approximately, 5mm sections were excised from diseased shoots using a sterile scalpel. The procedure for surface sterilisation of tissue pieces was the same as described in Section 5.3.3.1 in this Chapter. Shoot pieces were plated out on ½ PDA and incubated at 25°C.

5.3.6 Identification of Isolates

Isolates were initially grouped based of culture morphology. DNA was extracted from the mycelium of 5-day old cultures grown on ½ PDA plates. Details of sequencing studies are outlined in Section 4.3.6 of Chapter 4. Not all the pathogenic isolates in seedling trials were sequenced. Rather, only those representative of different morphological types (morphotypes) were chosen for sequencing. Similarly, 10 out of the 23 isolates used in the glasshouse trials were sequenced, while the others were identified by morphology. All fungi used in the field inoculations were identified using ITS sequencing.

Experiment 2: Pathogenicity of fungal isolates against juvenile M. pigra plants under glasshouse condition.

5.3.7 Growing of M. pigra Plants at the UQ Gatton Plant Nursery Unit

Seeds were surface sterilised following the procedure outlined in Section 4.3.1 and then planted using one seed per cell in a 100 cell seedling tray (height = 4.0 cm and width = 3.0cm) containing potting mix. After two weeks growth in a propagation chamber at 30±2 °C, 70-80% relative humidity (RH)

and under natural light, the seedlings were transplanted into propagation tubes (50mm square native tube) and grown for one month before being finally transplanted into pots (140mm/1.5 litres). The potting mix used comprised of composted bark and nutritional additives (Appendix G). After 5 months growth the plants were staked with bamboo sticks. An overhead sprinkler was used to water the plants in the nursery. Before inoculation, the plants were transferred to another glasshouse where they were grown for the duration of the experiment. Drip irrigation was applied twice daily for a period of 15 minutes.

5.3.8 Study Design and Treatments

Since winter temperatures in Gatton are too low to be conducive for good growth of M. pigra, the glasshouse trials were conducted during the summers of 2012 to 2014. The trials were done in batches, comprising ten treatments with 6 replicates in each batch. A total of 32 isolates were screened, including the reference isolate NT039 and a negative control of sterile millet. Each treatment (isolate) was screened twice in the different batches. Isolates screened were chosen on the basis of being among the 25 most pathogenic isolates identified in the seedling trials, while another 5 were included because they showed low/no pathogenicity in the seedling trials. These latter isolates were included to help determine the validity of the seedling assay; i.e. did they also lack pathogenicity to the juvenile plants. A randomised block design was used for this experiment. The trials were conducted over a period of six weeks. The plants were then harvested for laboratory analysis. In the presentation of results, isolates showing no pathogenicity were excluded from the graphs except the negative control.

5.3.9 Inoculum Preparation and Plant Inoculation

The process of inoculum preparation is outlined in Section 5.3.2.2 in this Chapter. A cordless drill and a 6.0 mm drill bit were used to drill a single hole of about 3-4mm depth, about 10 cm from the base of the plant stem. Five colonised millet seeds were inserted into the hole which was then sealed using Selleys TM Roof and Gutter Silicone Sealant. Negative control plants were treated with sterile millet seeds.

5.3.10 Plant Assessment

Stem lesion length was measured every 2 weeks for six weeks. Symptoms of plant disease, such as sunken depressions, falling leaves, the presence of sooty mold, were recorded. After 6 weeks, when the plants were harvested, internal lesion length of each harvested stem was recorded after splitting the stem lengthwise. Plant height and stem diameter at the point of inoculation were also recorded at the commencement and conclusion of the trials.

5.3.11 Confirmation of Infection

Confirmation of infection was done by comparing cultural morphology and sporulation of the re- isolations with the original culture isolates used for the glasshouse inoculation trials. Only two replicates out of the six were chosen 2 weeks after inoculation for attempted re-isolation of the putatively pathogenic fungi. Stem shavings from the stem parts with and without lesions were plated on ½ PDA for re-isolation. The procedure for surface sterilisation of tissue pieces before plating to ½ PDA, was as described in Section 5.3.5.

5.3.12 Data Analysis.

To compare and analyse the differences in pathogenicity between treatments, data (lesion length, replication and trial) were analysed using a general linear model of Analysis of Variance (ANOVA); Tukey’s test was then used for means comparisons (Minitab version 16). Data were assessed at 95% confidence level.

Experiment 3: Pathogenicity of selected fungal isolates against M. pigra plants growing in the field.

5.3.13 Plot Selection, Design and Treatments

The field inoculation trial was carried out on Melaleuca Station, Mary River, Northern Territory, in November 2012. Two plots were selected where plants had only minor stress and all main stems were still alive and had at least 80% of terminal stems with foliage. Any unhealthy trees in the study area were ignored and were not inoculated. The aim was to inoculate trees with the same health rating. There were 7 treatments, including a negative and putatively positive control (Table 5.1). The choice of the five isolates to use in the field was based on the most pathogenic isolates identified from the first seedling and glasshouse trials performed in 2011 and early 2012. These 5 isolates had the highest seedling mortality results. Isolates were randomly assigned in the two plots selected. Each plot had 32 trees. The adult M. pigra plants used were all multi-stemmed with a main stem with an average diameter of 50 ±5mm at 30cm above ground level.

Table 5.1 List of isolates used in the field pathogenicity trials and its corresponding codes.

MP Code Isolate ID Plot Initial Health Class (Source) MP 117 L. theobromae Healthy MP 020 Diaporthe sp. Dieback MP 107 L. theobromae Healthy MP 213 Diaporthe sp. Healthy MP 115 L. theobromae Healthy NT 039 L. pseudotheobromae Positive control Negative control sterile millet only Negative control

5.3.14 Inoculum and Capsule Preparation

The inoculum preparation followed the procedure outlined in Section 5.3.9, except that the colonised millet was enclosed in transparent Size 0 gelatine capsules (Figure 5.3B). Capsules had a sealed

A B

C D

volume of 0.95 ml and contained on average 46-50 colonised millet grains. The preparation of the inoculum in a capsule was done using a Fast-Cap® Capsule Filling Machine (Capsugel, Greenwood, South Carolina) kindly provided by Bioherbicides Australia Pty Ltd (Figure 5.3A). Prepared capsules were stored in Ziplock bags containing silica gel dessicant at 4oC until taken to the field.

5.3.15 Plant Inoculation

Among the multi-stems the stem with the largest diameter at 30cm height was chosen for inoculation. A cordlless drill and a 10.0mm drill bit were used to prepare a hole at about 30cm above ground level. A capsule was inserted into the hole and the hole sealed using Selleys TM Roof and Gutter Silicone Sealant (Figure 5.3C). Orange cattle tags were labelled and attached to the plants using wire for the purpose of identifying treated plants.

5.3.16 Trial Assessment

Trial assessment was done twice, at 6 months and 12 months after inoculation. The health of each tree was assessed using the M. pigra rating scale in Chapter 3 Table 3.2. Insect and pathogen activity, stem lesion height and diameter at point of inoculation were recorded.

5.3.17 Confirmation of Infection for Field Samples

A chainsaw was used to harvest the main stems of the inoculated stems (Figure 5.3D). Two stems per treatment were randomly chosen to confirm the pathogenicity of the chosen fungal isolates in the field. On returning the stems to the laboratory they were stored at 4oC for no more than 2 weeks before being split lengthwise and internal lesions measured. For re-isolation, 3 stem shavings from each of three areas of the split stems: (a) white and clean area, (b) red area in the middle and (c) the grey area near the pith, were plated on ½ PDA after surface sterilisation.

5.3.18 Data Analysis

Data were analysed using a general linear model of Analysis of Variance (ANOVA) and then Tukey’s test for means comparisons (Minitab version 16). Data were assessed at a 95% confidence level.

5.4 Results

Experiment 1: Seedling Trials

Stem pieces from healthy sites were clean and free from insect damage, except for one stem piece which had some internal discoloration. Stem pieces from dieback plants showed internal staining and some evidence of insect damage when split into stem pieces. 90

Two replicate trials:

Of the 255 isolates screened in the two-replicate seedling assay, 71 isolates caused seedling mortality. Most (87%) of the isolates pathogenic to M. pigra seedlings belonged to the genus Lasiodiplodia, 6% to Cophinforma sp., 4% to Neofusicoccum parvum and 2% to Diaporthe. These isolates caused visible disease symptoms 7 days after the start of the trials, the symptoms including the presence of black necrotic spots and yellowing of the cotyledons, extensive necrosis, colonisation of the shoots and cotyledons by fungal mycelium and eventual death of seedlings (Figure 5.4). All the controls in the trials were healthy and symptomless. None of the isolates of Pestalotiopsis, Fusarium or Sterilia Mycelia showed pathogenicity to young seedlings.

Ten replicate trials:

Repetition of the screening using a greater number (10) of replicate seedlings found that of the 71 putatively pathogenic isolates, only 48 (67%) caused some degree of seedling mortality (Figure 5.6) while the remaining 23 (33%) showed no mortality and could be considered false positives.

Isolates from the four different field sites showed various levels of pathogenicity to M. pigra seedlings (ranging from 85% mortality, which is as equally aggressive as the positive control NT039, to as little as 10% mortality). Fifty-two percent of these were from putatively healthy stems (Figure 5.6). Species of Botryosphaeriaceae with isolates pathogenic to M. pigra seedlings included: L. theobromae, L pseudotheobromae, L. hormozganensis, L. iraniensis, and Neofusicoccum parvum. One isolate of Diaporthe taken from a healthy stem caused 70% mortality.

Experiment 2: Glasshouse Trials

All 22 isolates of Lasiodiplodia evaluated were pathogenic to juvenile M. pigra plants, resulting in visible stem lesions within six weeks after inoculation. Observed symptoms on the stems, particularly above and below the point of inoculation included, sunken lesions and external stem discoloration (Figures 4.5 D and E). Observable lesion development was first seen 3 days after inoculation. No lesions developed on the stems of the inoculated the negative control (Figure 5.4G). The positive control, L. iraniensis NT039, consistently produced the longest lesion length in all trials. The original Botryosphaeriaceae species were re-isolated from stems chosen for re-isolation. No Botryosphaeriaceae were re-isolated from the controls.

Statistical analyses showed that the differences in lesion length between isolates were significant (P<0.05). The longest lesions among the isolates was produced by L. iraniensis (average 78.8 mm) sourced from dead and dieback sites and L. theobromae (65 mm) sourced from a healthy site. None 91

of the non-Botryosphaeriaceous isolates that were not pathogenic in the seedling assay when tested in glasshouse trials, showed pathogenicity toward juvenile M. pigra (data not shown).

A B C

D D E G E

HH I I

J K K A Figure 5. 4 Results of the pathogenicity trials. Disease symptoms on seedling trials; necrosis on tissues, fungal colonisation by mycelia and black spots on cotyledons (Figure 5.4 A-C);

glasshouse external and internal lesion development above and below the inoculation point (6 weeks after inoculation) (Figure 5.4 D-F); glasshouse negative control showing no internal dark staining (Figure 5.4 G); oval-shaped lesion observed 6 months after field inoculation trials at Melaleuca Station (Figure 5.4 H); internal staining observed an inoculated field main stem (Figure 5.4 I); field inoculated stem split lengthways showing lesion development into the pith (Figure 5.4 J-K).

Isolates from healthy tissues 160 Isolates from dieback tissues 140

120 Isolates from recovering tissues 100 Isolates from dead tissues

80 60

40 20 0

(mm) length lesion external Mean

Fungal species

Positive control (NT 039) Negative control Lasiodiplodia sp. (MP 597) L. pseudotheobromae (MP 813) Lasiodiplodia sp. (MP 862) L. iraniensis (MP 808) L. theobromae (MP 194) B.mamane (MP 442) B. mamane (MP 480) L. theobromae (MP 048) Diaporthe sp. (MP 213) L. theobromae (MP 036) Lasiodiplodia sp. (MP 559) Lasiodiplodia sp. (MP 692) Lasiodiplodia sp. (MP 361) L. theobromae (MP690) Lasiodiplodia sp. (MP 373) Lasiodiplodia sp. (MP 577) N. parvum (MP 157) L. iraniensis (MP 698) L.pseudotheobromae (MP 684) L. iraniensis (MP 647) L. iraniensis (MP 372) B. mamane (MP 290) Lasiodiplodia sp. (MP 020) L. theobromae (MP 188) Lasiodiplodia sp. (MP 342) Lasiodiplodia sp. (MP 612) Lasiodiplodia sp. (MP 483) Lasiodiplodia sp. (MP 610) L. hormozganensis (MP 310) L. theobromae (MP 315) L. iraniensis (MP 596) L. iraniensis (MP 490) N. parvum (MP 341) L. iraniensis (MP 539) Lasiodiplodia sp. (MP 760) Lasiodiplodia sp. (MP 378) L. theobromae (MP 117) Lasiodiplodia sp. (MP 723) L. iraniensis (MP 864) L. pseudotheobromae (MP 431) L. iraniensis (MP 582) L. hormozganensis (MP 330) L. theobromae (MP 115) L. pseudotheobromae (MP 129) L. theobromae (MP 107) 0 10 20 30 40 50 60 70 80 90 Proportion of dead seedlings (%) Legend:

Isolates from healthy tissues

Isolates from dieback tissues Isolates from recovering tissues

Isolates from dead tissues

Figure 5.6 The proportion of dead seedlings at harvest two weeks after inoculation (10 replicate seedling trials), n=20

Experiment 3: Field Inoculation Trials All isolates six inoculated on adult M. pigra in the field produced visible surface lesions after 6 months of inoculation (Figure 5.4 H). However, the mean lesion lengths produced by the majority of the isolates were not significantly different from those of the control. Some trees inoculated as negative controls also developed lesions. No Lasiodiplodia species were re-isolated from the red and grey areas of the split stems, but Epicoccum spp. and Fusarium sp. were recovered. Lasiodiplodia spp., Penicillium spp. and Pestalotiopsis type sp. were re-isolated from the clean white area. Nothing was recovered from the clean, apparently healthy white area of the control plants, while Diaporthe sp. and Lasiodiplodia sp. were recovered from the lesions on negative control plants.

5.5 Discussion

Lasiodiplodia species can cause seedling mortality and aggressiveness to juvenile and mature plants in controlled and field conditions, resulting in the development of external and internal stem lesions during the trial period. In contrast, isolates belonging to Pestalotiopsis type species, and most of the Diaporthe species and Fusarium species, recovered from the dieback sites did not exhibit disease symptoms or cause seedling mortality.

Most isolates pathogenic to M. pigra were sourced from putatively healthy trees (52% in seedling trials, 41% in glasshouse trials, and 80% in field trials). There were more isolates pathogenic to field than in glasshouse and seedling trials. A possible explanation for this might be that plants in the field were more stressed and are predisposed by a potential biotic or abiotic factor. These results agree those in other reports, in which species that produced the longest lesions in woody plants were obtained from healthy tissues (Sakalidis et al. 2011; Jami et al. 2013). These results support the idea that members of the Botryosphaeriaceae exist as endophytes and latent pathogens in many woody plants (Slippers et al. 2009; Perez et al. 2010; Sakalidis et al. 2011) and are associated with dieback disorders when conditions become conducive for these fungi to become pathogenic. (Wilson & Pitkethley 1992) reported that Botryodiplodia (=Lasiodiplodia) theobromae) was associated with M. pigra dieback in the Adelaide River region of the Northern Territory. In recent years, with the advent of molecular analyses, the species L. theobromae has undergone many systematic revisions and been split into many previously cryptic species (Liu et al. 2012; Phillips et al. 2013; Slippers et al. 2013). Many of the Lasiodiplodia species recovered in this study (for example, L. pseudotheobromae, L. iraniensis, L. hormozganensis) have been previously identified as L. theobromae (Alves et al. 2008; Begoude et al. 2012).

The findings observed in Experiment 1 are consistent with those of studies of other woody weeds of Australia (Diplock et al. 2006; Toh et al. 2008; Haque et al. 2013). Lasiodiplodia species have been frequently isolated from field samples, and seedling trials performed identified these species to be highly pathogenic to these weeds. The present study shows that, within a particular species, isolates vary in pathogenicity, and that an isolate of L. iranensis (NT039), not isolated from M. pigra, regularly showed the greatest pathogenicity. Other reports also show that different isolates of the same Botryosphaeriacae species isolated from the same host, can differ markedly in pathogenicity to that host following inoculation (Marques et al. 2013; Yan et al. 2013; Rodríguez-Gálvez et al. 2015). A possible explanation for this is that some isolates may lose their aggressiveness when maintained for extensive periods on artificial media, or after successive subculturing. Attenuation of virulence has been reported in several taxa of fungi (Butt et al. 2006). Care was taken to avoid these practices 96

in the present study. Isolates, once purified, were stored under water until required. (Boesewinkel 1976) showed that water storage maintains the viability and pathogenicity of a wide range of fungi for extensive periods.

Many isolates of Diaporthe and Pestalotiopsis were recovered in the present study from apparently healthy plants, demonstrating that they are true endophytes of M. pigra. Members of these taxa are important endophytes associated with many different woody plants. For example, it is reported that various species of Pestalotiopsis produce taxol, while others produce newly discovered compounds with medicinal potential (Maharachchikumbura et al. 2011; Singh et al. 2011; Gomes et al. 2013; Urbez-Torres et al. 2013). The genus Diaporthe also contains many plant pathogens, including some with legume hosts (Rupe et al. 1999; Santos et al. 2011). It is notable that the Diaporthe isolates were pathogenic to young seedlings in the seedling assay and caused stem lesions on mature trees. This is another case of an endophyte also being a facultative pathogen.

The results of the field inoculation trials showed that the treatments were not significantly different from the control in terms of lesion length. It is possible that the native borers had an effect on this observation. Carmenta mimosa and Maroga setiotricha mostly feed and oviposit on M. pigra main stems (Heard 2012). This may facilitate secondary infections at the point of inoculation of the negative control thereby resulting in lesions as long as those of treatment isolates. Drilling a hole and inserting a gelatine capsule containing sterile millet could provide a wound and nutrient source for endophytes already in the stem to proliferate and attack the host. The external and internal lesions that developed on control plants after inoculation reveal the capacity of Lasiodiplodia species complex (which were re-isolated from the lesions of control plants and also from “healthy” field plants) to cause disease and spread through the vascular tissues of their host. This field inoculation study, while preliminary, suggests that L. theobromae isolates recovered from healthy stems can exhibit disease symptoms when inoculated back to a healthy tree in field conditions.

In summary, many fungal isolates recovered as endophytes are capable of attacking M. pigra under controlled conditions. This chapter provides further evidence of the association of these isolated fungal species in M. pigra. Further studies are needed to elucidate the biological cycle of the fungus, determine the appropriate inoculum concentration, define techniques for field application and the prolongation of the shelf-life of the inoculum, to develop a fuller picture on how these isolates can be utilised as a mycoherbicide for M. pigra.

CHAPTER 6

Conclusions and Suggestions for Future Research

This chapter summarises the key findings in relation to the research aims and underscores the implications for further field, glasshouse and laboratory research on M. pigra dieback. Table 6.1 collates all the key findings of this present study and its implications.

Table 6.1 Summary table collating all finding of this Thesis and its possible implication. Key findings Possible implications Ecology Chapter 1. The results of the survey indicate that Dieback in M. pigra is ecologically dieback in M. pigra is highly dynamic important. Understanding the underlying but fluctuating in its severity, resulting causes of dieback phenomenon helps us in partial mortality, reduced plant design better management strategies. vigour, the death of terminal stems and death of main stems.

2. Insect borers and fungal endophytes Although ubiquitous across all plots, insect were ubiquitous across all plots biocontrol agents, native borers and fungal covered by the survey, and over time endophytes are contributing factors to the and showed no direct relationship with decline process in M. pigra. Future work is the M. pigra population dynamics. needed to further understand the role of other biotic factors such as “damping off pathogens, oomycetes, bacteria and viruses in the dieback phenomenon.

3. Soil pH and salinity were also not There might be other predisposing factors observed to have a biological such as temperature, humidity, poor soil significant influence on M. pigra drainage, plant age and climate change population dynamics. that work in unison to explain the dieback phenomenon in M. pigra.

Morphology Chapter

1. A total of 1,352 isolates were purified An implication of this is that M. pigra into axenic cultures. There were 24 harbours diverse endophyte communities. morphotypes identified and species These isolates may be further studied for under Botryosphaeriaceae, biological control and natural products Pestalotiopsis spp., and Diaporthe spp studies. were the most prominent taxa.

2. Most morphology groupings under The use of more gene regions is needed to Botryosphaeriaceae were not identify species of these genera. identified to the species level using Morphological characters are used for ITS sequencing. The inclusion of EF1- initial identification as this has some α results in a more robust separation. disadvantages (i.e. sporulation in culture, incubation conditions and media selection).

Pathogenicity Chapter

3. Fungal isolates recovered as Fungal isolates may be further studied for endophytes are capable of attacking biological control of this weed in a M. pigra under controlled conditions. controlled condition (i.e the mechanism of infection and the life cycle of the fungus in relation to the effect in the plant physiologically).

4. Field inoculation trials showed that Mycoherbicidal activity of the most “L. theobromae complex” exhibit pathogenic isolates be further understood disease symptoms when inoculated (i.e. mass production of potential back into a healthy tree under field biocontrol agents, formulation, shelf-life, conditions. optimum inoculum concentration, and the adequate conditions for fungus effectiveness against the weeds host range).

Evaluating the relative role of endophytic fungi, phytophagous insects, and the environment in the development of dieback in Mimosa pigra.

Chapter 3 highlights the challenges of attributing single factors in evaluating the dieback phenomenon. Dieback is an important phenomenon in M. pigra and it has been possible to rule out some potential contributing factors relating to dieback. Despite their ubiquity, classical biological control agents and fungal endophyte communities have shown no direct relationship with the changing population health dynamics in M. pigra. However, as discussed above, care needs to be taken in evaluating the impact of single factors in relation to dieback (Sims-Chilton et al. 2010).

All sites across the study populations were affected by dieback, but the effects of dieback were highly variable and asynchronous. Dieback affects all age classes (from seedlings, juvenile to adult plants). The dieback dynamics observed in M. pigra which is cyclical and results in fluctuating plant health over time, has also been observed in other species (Simberloff & Gibbons 2004; Tang et al. 2012). The difficulty in explaining the cause of the dieback phenomenon seems to be consistent with other reported research. Experimental and observational studies, such as by Elmer et al. (2013), have failed to show clear evidence of causality, while there has been no clear pattern relating to abiotic factors (rainfall and temperature) associated with the (SVD) sudden vegetation dieback in the Atlantic and Gulf Coast Marshes. In unravelling the cause of crown dieback in aspen, no single factors have been able to identified to explain the differences between healthy and stressed aspen in Alberta, Canada (Hogg et al. 2002).

In Chapter 3, it was shown that insect borers and fungal endophytes were abundant and ubiquitous across all plots and seasons. Further studies on the interactions between insect and fungal endophytes are important in order to provide an understanding as to whether the insect borers act as vectors of the pathogenic fungal endophytes. Lee et al. (2011) has shown that ambrosia beetles are the vectors of the Korean oak wilt pathogen, causing crown dieback.

Previous studies have suggested that abiotic factors are the “predisposing factors” in the dieback of weeds (Aghighi et al. 2014) and in other systems (Manion & Lachance 1992; Mitchell et al. 2013). Contrary to general belief, the results of this study provided no direct evidence of the abiotic factors examined being the causal agents of M. pigra dieback. A possible explanation for this result might be related to the fact that in Australia M. pigra can tolerate higher levels of salinity (e.g ̴ 18 ppt) (Miller 1983). M. pigra favours and is adapted to a wet-dry tropical climate and seasonal inundation, and dieback has not been observed in similar habitats in its native range (Harley 1985). Observational studies of Mulrennan and Woodroffe (1998) in the coastal plains of the Lower Mary River in the Northern Territory have showed that the penetration of salt water into the floodplains stresses the natural vegetation and leads to extensive dieback of paperbark (Melaleuca spp.) stands. However, (Lonsdale 1992) reported that in its native range, M. pigra occupies similar habitats and conditions to paperbark but has not shown dieback events.

Identification of endophytic fungi from Mimosa pigra based on cultural morphology, ITS and partial translation elongation factor (EF1-α) sequences.

The key outcome of Chapter 4 was that M. pigra is a host to diverse endophytic communities and that there were numerous fungal species across all plots, seasons and years, in particular, three taxa were most frequently recovered, Botryosphaeriales, Diaporthaes and Xylariales. There were 15 confirmed fungal species identified, based on the phylogenetic characters used. In addition to the methodology used for identification of fungal isolates used in this study, future identification of endophytes in M. pigra, inclusion of more gene regions such as β – tubulin, small sub-unit (SSU) and large sub-unit (LSU) ribosomal sequences (Maharachchikumbura et al. 2012; Phillips et al. 2013; Thompson et al. 2014) would be more informative for species identification. To further elucidate the diversity, composition and ecological importance of endophyte communities, future investigations should include use of multi-species screening to determine their functional roles (Ren et al. 2014).

In Chapter 4, it was demonstrated that using the culture-based approach in isolation from M. pigra stems, fungal abundance was relatively high and diverse. This outcome is consistent with the data contained in the studies of Angelini et al. (2012) and Oono et al. (2014) . It would be interesting to

100

compare the fungal communities not only in stems but also in other plant parts such as roots, seeds, and leaves (Cook et al. 2009) to have a fuller picture of the endophyte communities in M. pigra. Furthermore, root pathogens, like Phytophthora, need to be investigated for their association with dieback (Aghighi et al. 2012), as this study focused more on quantitative evidence of the above- ground agents.

In Chapter 4 it is also reported there were “unidentified isolates” which cannot be grouped into any assigned taxa. An implication of this is the possibility that these could potentially be novel mycoviruses, which can be transmitted to conidia and be spread between compatible hyphae (Ghabrial & Suzuki 2009; Rodriguez-Garcia et al. 2014). Further research might include into the hypovirulence of these isolates, and testing of whether the virulence of the fungal pathogens is decreased due to the infection of mycoviruses (Zhang et al. 2009).

Fulfilling Koch’s postulates confirms the pathogenicity of fungal species recovered from healthy and dieback-affected Mimosa pigra populations in the Northern Territory, Australia.

Overall, the key finding of Chapter 5 was that the diverse group of isolates/endophytes tested for pathogenicity, species from the Lasiodiplodia and Cophinforma complex, both of which are Botryosphaeriaceae, showed seedling mortality and aggressiveness against M. pigra seedlings and juvenile plants. Although, there were no plant deaths in either the glasshouse or field inoculation trials, a number of isolates produced lesions, indicating their ability to be pathogens and hence confirming Koch’s postulates. It was also established which of the subset of isolates are “true endophytes”, and those that are likely to be contributing factors in M. pigra dieback, and that within the latter group that pathogenicity differs between and within the species of Lasiodiplodia complex. To further examine the association of the biotic and abiotic contributing factors and M. pigra dieback under controlled conditions, experiments on the synergistic or antagonistic interactive effect between the two factors are important to provide an understanding of the mechanism of the combined effects, or competition, on the compensatory growth of invasive plants (Ferrero-Serrano et al. 2008; Li et al. 2013). Bansal et al. (2013) demonstrated that the combined impacts from two stressors (herbivory and drought) had a lesser effect on the establishment of Pinus sylvestris seedlings when compared to the individual impact from a single stressor. On the other hand, Hogg et al. (2002) found the opposite relationship. In addition, Davison (2014) has pointed out that, to gain an accurate understanding of the complex dieback phenomenon, we should not forget the plant itself. An examination of plant resistance and its physiology in relation to stressors is important to help further understand dieback (Ramos et al. 1991; Pascoe & Cottral 2000).

101

6.1 Recommendations for future research work:

1. It is recommended that other possible predisposing factors like temperature, drought, humidity, soil fertility, soil moisture and plant age be further studied to determine what role these factors play in M. pigra dieback phenomenon. 2. It would be interesting to assess the role of other possible contributing “biotic” factors in the decline process. Further research might explore on “damping-off” and soil-borne pathogens, oomycetes, bacteria and viruses to establish a better understanding of these possible contributing factors in M. pigra dieback phenomenon. 3. Another possible area of future research is to do modelling studies of the dynamics of seedlings, juveniles and adults over time and across different plots to see what the outcome of the decline process may be in the long term. 4. Moreover, looking at the endophytes in the seedlings across the different plots and over time may help establish the role of below- ground agents in the decline process. 5. A number of possible studies in the pathogenicity and mycoherbicidal ability of the fungal isolates be determined. These include mass production of potential biocontrol agents, formulation, shelf-life, optimum inoculum concentration, and the adequate conditions for fungus effectiveness against the weeds host range. 6. Considerably more work will need to be done to identify the fungal isolates morphologically and phylogenetically. One of which is to include more gene regions aside from ITS and EF1- α. Finally, the use of metagenomics is recommended to further understand the dieback occurrence in M. pigra.

6.2 Conclusions

The fact that dieback in M. pigra occurs in multiple catchments in the Northern Territory and over large areas, makes the dieback phenomenon to be of considerable ecological importance. Although extensive research has been conducted on the evaluation of the biological agents in M. pigra in relation to decreasing plant health, there is a need to consider what role of other factors might play in the observed dieback phenomenon. Based on the results of this study, no single reason/factor can be directly identified among both the biotic and abiotic factors evaluated, as being responsible for the recent phenomenon of M. pigra dieback in the Northern Territory. However a number of contributing factors have been ruled out. This thesis highlights the challenges in attributing single factor as being responsible for the dieback phenomenon. Dieback might be associated with numerous interacting factors, including some below-ground agents.

102

The results of the laboratory and glasshouse studies undertaken shows the ubiquity and diversity of the fungal endophytes through time. The pathogenicity trials also demonstrated that members of the Botryosphaeriaceae are the most aggressive taxa. It is therefore suggested that these fungal endophytes might be further studied for their potential incorporation into strategies for the integrated weed management of M. pigra.

Given the complexity of these interacting factors in evaluating M. pigra dieback, future research might focus on the potential interactions of the contributing factors (both biotic and abiotic), and to undertake long-term research and monitoring of the plant’s demography in relation to dieback.

103

7.0 References

Abdollahzadeh, J, Javadi, A, Goltapeh, EM, Zare, R & Phillips, A 2010, 'Phylogeny and morphology of four new species of Lasiodiplodia from Iran', Persoonia, vol. 25, pp. 1-10.

Addy, HD, Piercey, MM & Currah, RS 2005, 'Microfungal endophytes in roots', Canadian Journal of Botany-Revue Canadienne De Botanique, vol. 83, no. 1, pp. 1-13.

Afkhami, ME & Rudgers, JA 2008, 'Symbiosis lost: Imperfect vertical transmission of fungal endophytes in grasses', American Naturalist, vol. 172, no. 3, pp. 405-16.

Aghighi, S, Hardy, GEJ, Scott, JK & Burgess, TI 2012, 'Phytophthora bilorbang sp nov., a new species associated with the decline of Rubus anglocandicans (European blackberry) in Western Australia', European Journal of Plant Pathology, vol. 133, no. 4, pp. 841-55.

Aghighi, S, Fontanini, L, Yeoh, PB, Hardy, GES, Burgess, TI & Scott, JK 2014, 'A Conceptual Model to Describe the Decline of European Blackberry (Rubus anglocandicans), a Weed of National Significance in Australia', Plant Disease, vol. 98, no. 5, pp. 580-9.

Agrios, GN 2005, Plant Pathology, United States of America.

Aguilar-Trigueros, CA, Powell, JR, Anderson, IC, Antonovics, J & Rillig, MC 2014, 'Ecological understanding of root-infecting fungi using trait-based approaches', Trends in Plant Science, vol. 19, no. 7, pp. 432-8.

Alexopoulos, CJ 1996, 'Characteristics of fungi and Fungal Systematics.', in Introductory Mycology, John Wiley & Sons, Inc, New York, vol. 26, pp. 61-85.

Alves, A, Crous, PW, Correia, A & Phillips, AJL 2008, 'Morphological and molecular data reveal cryptic speciation in Lasiodiplodia theobromae', Fungal Diversity, vol. 28, pp. 1-13.

Angelini, P, Rubini, A, Gigante, D, Reale, L, Pagiotti, R & Venanzoni, R 2012, 'The endophytic fungal communities associated with the leaves and roots of the common reed (Phragmites australis) in Lake Trasimeno (Perugia, Italy) in declining and healthy stands', Fungal Ecology, vol. 5, no. 6, pp. 683-93.

Arif, M, Zaidi, NW, Haq, QMR, Singh, YP, Khan, S & Singh, US 2013, 'Molecular phylogeny and pathotyping of Fusarium solani: a causal agent of Dalbergia sissoo decline', Forest Pathology, vol. 43, no. 6, pp. 478-87.

104

Arnold, A, Maynard, Z, Gilbert, GS, Coley, PD & T, K 2000, 'Are tropical fungal endophytes hyperdiverse?', Ecology Letters, vol. 3, pp. 267-74.

Arnold, AE, Maynard, Z & Gilbert, GS 2001, 'Fungal endophytes in dicotyledonous neotropical trees: patterns of abundance and diversity', Mycological Research, vol. 105, pp. 1502-7.

Arnold, AE, Mejia, LC, Kyllo, D, Rojas, EI, Maynard, Z, Robbins, N & Herre, EA 2003, 'Fungal endophytes limit pathogen damage in a tropical tree', Proceedings of the National Academy of Sciences of the United States of America, vol. 100, no. 26, pp. 15649-54.

Arnold, AE & Lutzoni, F 2007, 'Diversity and host range of foliar fungal endophytes: Are tropical leaves biodiversity hotspots?', Ecology, vol. 88, no. 3, pp. 541-9.

Auld, BA & Morin, L 1995, 'Constraints in the Development of Bioherbicides', Weed Technology, vol. 9, no. 3, pp. 638-52.

Australian Government Department of Sustainability, E, Water, Population and Communities 2012, Parks and Reserves Kakadu National Park.

Bacon, C & White, J (eds) 2000, Microbial Endophytes, An Overview of Endophytic Microbes: Endophytism Defined, Marcel Dekker, New Jersey.

Bailey, KL 2004, 'Microbial weed control: an off-beat application plant pathology', Canadian Journal of Plant Pathology-Revue Canadienne De Phytopathologie, vol. 26, no. 3, pp. 239-44.

Bailey, KL, Boyetchko, SM & Langle, T 2010, 'Social and economic drivers shaping the future of biological control: A Canadian perspective on the factors affecting the development and use of microbial biopesticides', Biological Control, vol. 52, no. 3, pp. 221-9.

Ball, MC 1998, 'Mangrove species richness in relation to salinity and waterlogging: a case study along the Adelaide River floodplain, northern Australia', Global Ecology and Biogeography Letters, vol. 7, no. 1, pp. 73-82.

Bansal, S, Hallsby, G, Lofvenius, MO & Nilsson, M-C 2013, 'Synergistic, additive and antagonistic impacts of drought and herbivory on Pinus sylvestris: leaf, tissue and whole-plant responses and recovery', Tree Physiology, vol. 33, no. 5, pp. 451-63.

Barneby, RC 1991, 'Sensitivae Censitae: A Description of the genus Mimosa Linnaeus Mimosaceae in the New World', Memoirs of the New York Botanical Garden, vol. 65, pp. 1-835.

105

Barnett, HL & Hunter, B 1998, Illustrated Genera of Imperfect Fungi, APS Press.

Barton, J 2012, 'Predictability of pathogen host range in classical biological control of weeds: an update', Biocontrol, vol. 57, no. 2, pp. 289-305.

Bary, D 1887, 'Comparative morphology and biology of the fungi, Mycetozoa and Bacteria', Clarendon, Oxford.

Begoude, BAD, Slippers, B, Wingfield, MJ & Roux, J 2010, 'Botryosphaeriaceae associated with Terminalia catappa in Cameroon, South Africa and Madagascar', Mycological Progress, vol. 9, no. 1, pp. 101-23.

Begoude, BAD, Slippers, B, Wingfield, MJ & Roux, J 2011, 'The pathogenic potential of endophytic Botryosphaeriaceous fungi on Terminalia species in Cameroon', Forest Pathology, vol. 41, no. 4, pp. 281-92.

Begoude Boyogueno, AD, Slippers, B, Perez, G, Wingfield, MJ & Roux, J 2012, 'High gene flow and outcrossing within populations of two cryptic fungal pathogens on a native and non- native host in Cameroon', Fungal Biology, vol. 116, no. 3, pp. 343-53.

Bergstrom, DM, Bricher, PK, Raymond, B, Terauds, A, Doley, D, McGeoch, MA, Whinam, J, Glen, M, Yuan, ZQ, Kiefer, K, Shaw, JD, Bramely-Alves, J, Rudman, T, Mohammed, C, Lucieer, A, Visoiu, M, van Vuuren, BJ & Ball, MC 2015, 'Rapid collapse of a sub- Antarctic alpine ecosystem: the role of climate and pathogens', Journal of Applied Ecology, vol. 52, no. 3, pp. 774-83.

Biswas, SR, Kotanen, PM, Kambo, D & Wagner, HH 2015, 'Context-dependent patterns, determinants and demographic consequences of herbivory in an invasive species', Biological Invasions, vol. 17, no. 1, pp. 165-78.

Blossey, B & Skinner, L 2000, 'Design and importance of post release and monitoring.', paper presented to X International Symposium of Biological Control of Weeds, Montana USA.

Bold, HC, Alexopoulos, CJ & Delevoryas, T 1980, Morphology of plants and fungi, Harper & Row, New York.

Brouwers, NC, Mercer, J, Lyons, T, Poot, P, Veneklaas, E & Hardy, G 2013, 'Climate and landscape drivers of tree decline in a Mediterranean ecoregion', Ecology and Evolution, vol. 3, no. 1, pp. 67-79.

106

Buchan, A, Newell, S, Moreta, JL & Moran, M 2002, 'Analysis of internal transcribed spacer (ITS) regions of rRNA genes in fungal communities in a southeastern US salt marsh', Microbial Ecology, vol. 43, no. 3, pp. 329-40.

Buckley, YM, Rees, M, Paynter, Q & Lonsdale, M 2004, 'Modelling integrated weed management of an invasive shrub in tropical Australia', Journal of Applied Ecology, vol. 41, no. 3, pp. 547-60.

Burke, MK & Raynal, DJ 1998, 'Limiting influences growth and nutrient balances in sugar maple (Acer saccharum) seedlings on an acidic forest soil', Environmental and Experimental Botany, vol. 39, no. 2, pp. 105-16.

Burrows, NJ, Lukitsch, BV & Liberato, JR 2012, 'Rediscovery of the rust Diabole cubensis, released as a classical biological control agent against the invasive weed Mimosa pigra in Australia', Australasian Plant Disease Notes, vol. 7, no. 1, pp. 171-5.

Busch, DE & Smith, SD 1995, 'Mechanisms associated with decline of woody species in riparian ecosystems of the Southwestern US', Ecological Monographs, vol. 65, no. 3, pp. 347- 70.

Campbell, DLM, Weiner, SA, Starks, PT & Hauber, ME 2009, 'Context and control: behavioural ecology experiments in the laboratory', Annales Zoologici Fennici, vol. 46, no. 2, pp. 112-23.

Carson, WP, Hovick, SM, Baumert, AJ, Bunker, DE & Pendergast, TH 2008, 'Evaluating the post- release efficacy of invasive plant biocontrol by insects: a comprehensive approach', Arthropod-Plant Interactions, vol. 2, no. 2, pp. 77-86.

Casadevall, A., Pirofski, L., 1999. Host-pathogen interactions: redefining the basic concepts of virulence and pathogenicity. Infect. Immun. 67, 3703-3713.

Chappell, J & Woodroffe, CD 1985, 'Morphodynamics of Northern Territory tidal rivers and floodplains', in KN Bardsley & CD Woodroffe (eds), Coasts and tidal wetlands of the Australian monsoon region, pp. 85-96.

Charudattan, R 2010, 'A Reflection on My Research in Weed Biological Control: Using What We Have Learned for Future Applications', Weed Technology, vol. 24, no. 2, pp. 208-17.

107

Chopping, C 2004, 'Mimosa pigra at Peter Faust Dam, Proserpine, Queensland, Australia', in Research and management of Mimosa pigra: papers presented at the 3rd International Symposium on the Management of Mimosa pigra, Darwin, Australia, 22-25 September 2002., pp. 102-5.

Cook, GD & Dias, L 2006, 'It was no accident: deliberate plant introductions by Australian government agencies during the 20th century', Australian Journal of Botany, vol. 54, no. 7, pp. 601-25.

Cook, D, Gardner, DR, Ralphs, MH, Pfister, JA, Welch, KD & Green, BT 2009, 'Swainsoninine concentrations and endophyte amounts of Undifilum oxytropis in different plant parts of Oxytropis sericea', Journal of chemical ecology, vol. 35, no. 10, pp. 1272-8.

Coombs, RF & Lisansky, SG 1993, Dictionary of biological control and integrated pest management, CPL Press, Newbury, England.

Creager, RA 1992, 'Seed Germination, Physical and Chemical Control of Catclaw Mimosa (Mimosa pigra var. pigra) ', Weed Science Society of America, vol. 6, pp. 884-91.

Dakin, N, White, D, Hardy, G & Burgess, TI 2010, 'The opportunistic pathogen, Neofusicoccum australe, is responsible for crown dieback of peppermint (Agonis flexuosa) in Western Australia', Australasian Plant Pathology, vol. 39, no. 2, pp. 202-6.

Damm, U, Cannon, PF, Woudenberg, JHC, Johnston, PR, Weir, BS, Tan, YP, Shivas, RG & Crous, PW 2012, 'The Colletotrichum boninense species complex', Studies in Mycology, no. 73, pp. 1-36.

Davis, AS, Landis, DA, Nuzzo, V, Blossey, B, Gerber, E & Hinz, HL 2006, 'Demographic models inform selection of biocontrol agents for garlic mustard (Alliaria petiolata)', Ecological Applications, vol. 16, no. 6, pp. 2399-410.

Davis, EC & Shaw, AJ 2008, 'Biogeographic and phylogenetic patterns in diversity of liverwort- associated endophytes', American Journal of Botany, vol. 95, no. 8, pp. 914-24.

Davison, EM 2014, 'Resolving confusions about jarrah dieback - don't forget the plants', Australasian Plant Pathology, vol. 43, no. 6, pp. 691-701.

DeBach 1964, Biological Control of Insect Pests and Weeds, Reinhold Publishing Corporation, New York.

108

de Errasti, A, Victoria Novas, M & Carmaran, CC 2014, 'Plant-fungal association in trees: Insights into changes in ecological strategies of Peroneutypa scoparia (Diatrypaceae)', Flora, vol. 209, no. 12, pp. 704-10.

Denman, S, Crous, PW, Groenwald, JZ, Slippers, B, Wingfield, BD & Wingfield, MJ 2003, 'Circumscription of Botryosphaeria species associated with Proteaceae based on morphology and DNA sequence data', Mycologia, vol. 95, no. 2, pp. 294-307.

Desprez-Loustau, ML, Marcais, B, Nageleisen, LM, Piou, D & Vannini, A 2006, 'Interactive effects of drought and pathogens in forest trees', Annals of Forest Science, vol. 63, no. 6, pp. 597-612.

Diogo, ELF, Santos, JM & Phillips, AJL 2010, 'Phylogeny, morphology and pathogenicity of Diaporthe and Phomopsis species on almond in Portugal', Fungal Diversity, vol. 44, no. 1, pp. 107-15.

Diplock, N, Galea, V, Van Klinken, RD & Wearing, R 2006, 'A preliminary investigation of dieback in Parkinsonia aculeata ', paper presented to 15th Australasian Weeds Conference, Adelaide, Australia.

Du Puy, D 1997, 'The Leguminosae of Madagascar', Curtis's Botanical Magazine, vol. 14, no. 4, pp. 231-41.

Duchesne, L, Ouimet, R & Houle, D 2002, 'Basal area growth of sugar maple in relation to acid deposition, stand health, and soil nutrients', Journal of Environmental Quality, vol. 31, no. 5, pp. 1676-83.

Edmeades, BF 2011, Increasing the Productivity of Improved pastures by the sustainable management of native vegetation regrowth, Northern Territory Agricultural Association Incorporated, Northern Territory.

Edwards, PB & Brown, EM 1997, 'Mesoclanis seed flies (Diptera: Tephritidae) on Chrysanthemoides (Asteraceae) in South Africa: distribution, attack strategy and impact on seed production', Bulletin of Entomological Research, vol. 87, pp. 127-35.

Egerton, FN 1973, 'Changing concepts of balance of Nature', Quarterly Review of Biology, vol. 48, no. 2, pp. 322-50.

109

Elmer, WH 2014, 'A Tripartite Interaction Between Spartina alterniflora, Fusarium palustre, and the Purple Marsh Crab (Sesarma reticulatum) Contributes to Sudden Vegetation Dieback of Salt Marshes in New England', Phytopathology, vol. 104, no. 10, pp. 1070-7.

Elmer, WH & Marra, RE 2011, 'New species of Fusarium associated with dieback of Spartina alterniflora in Atlantic salt marshes', Mycologia, vol. 103, no. 4, pp. 806-19.

Elmer, WH, LaMondia, JA, Useman, S, Mendelssohn, IA, Schneider, RW, Jimenez-Gasco, MM, Marra, RE & Caruso, FL 2013, 'Sudden Vegetation Dieback in Atlantic and Gulf Coast Salt Marshes', Plant Disease, vol. 97, no. 4, pp. 436-45.

Ellis R. C. (1964) Dieback of alpine ash in north eastern Tasmania. Aust. For. 28, 75 -90.

Espinoza, JG, Briceno, EX, Keith, LM & Latorre, BA 2008, 'Canker and twig dieback of blueberry caused by Pestalotiopsis spp. and a Truncatella sp. in Chile', Plant Disease, vol. 92, no. 10, pp. 1407-14.

Evans, HC, Carrion, G & Ruizbelin, F 1995, 'Mycobiota of the giant sensitive plant, Mimosa pigra Sensu-lato in the Neotropics', Mycological Research, vol. 99, pp. 420-8.

Ferrero-Serrano, A, Collier, TR, Hild, AL, Mealor, BA & Smith, T 2008, 'Combined impacts of native grass competition and introduced weevil herbivory on Canada thistle (Cirsium arvense)', Rangeland Ecology & Management, vol. 61, no. 5, pp. 529-34.

Gassmann, A, Cock, MJW, Shaw, R & Evans, HC 2006, 'The potential for biological control of invasive alien aquatic weeds in Europe: a review', Hydrobiologia, vol. 570, pp. 217-22.

Ghabrial, SA & Suzuki, N 2009, 'Viruses of Plant Pathogenic Fungi', in Annual Review of Phytopathology, vol. 47, pp. 353-84.

Giordano, L, Gonthier, P, Varese, GC, Miserere, L & Nicolotti, G 2009, 'Mycobiota inhabiting sapwood of healthy and declining Scots pine (Pinus sylvestris L.) trees in the Alps', Fungal Diversity, vol. 38, pp. 69-83.

Gazis, R, Miadlikowska, J, Lutzoni, F, Arnold, AE & Chaverri, P 2012, 'Culture-based study of endophytes associated with rubber trees in Peru reveals a new class of Pezizomycotina: Xylonomycetes', Molecular Phylogenetics and Evolution, vol. 65, no. 1, pp. 294-304.

Global Invasive Species Database 2015, viewed 22 April 2015, .

110

Gray, MW, Sankoff, D & Cedergren, RJ 1984, 'On the evolutionary descent of organisms and organelles: a global phylogeny based on a highly conserved structural core in small subunit ribosomal RNA', Nucleic Acids Research, vol. 12, no. 14, pp. 5837-52.

Gomes, RR, Glienke, C, Videira, SIR, Lombard, L, Groenewald, JZ & Crous, PW 2013, 'Diaporthe: a genus of endophytic, saprobic and plant pathogenic fungi', Persoonia, vol. 31, pp. 1- 41.

Gross, A, Holdenrieder, O, Pautasso, M, Queloz, V & Sieber, TN 2014, 'Hymenoscyphus pseudoalbidus, the causal agent of European ash dieback', Molecular Plant Pathology, vol. 15, no. 1, pp. 5-21.

Guarro, J, Gene, J & Stchigel, AM 1999, 'Developments in Fungal Taxonomy', Clinical Microbiology Reviews, pp. 454-500.

Guo, LD, Hyde, KD & Liew, ECY 2000, 'Identification of endophytic fungi from Livistona chinensis based on morphology and rDNA sequences', New Phytologist, vol. 147, no. 3, pp. 617- 30.

Haque, A, Galea, V, Goulter, K & Van Klinken, RD 2013, 'Botryosphaeriaceae fungi as a potential mycoherbicide for Prickly acacia', paper presented to 19th Australasian Weeds Conference, Auckland, New Zealand.

Harley, KLS 1985, 'Suppression of reproduction of woody weeds using insects which destroy flower and seeds', paper presented to 6th International Symposium on Biological Control of Weeds, Vancouver, 19-25, August, 1984.

Harley, KLS & Forno, IW 1992, Biological Control of Weeds a handbook for practitioners and students, Inkata Press, Brisbane.

Harley, KLS, Gillett, J, Winder, J, Forno, IW, Segura, RPL, Miranda, H & Kassulke, R 1995, 'Natural Enemies of Mimosa pigra and M. berlandieri (Mimosaceae) and Prospects for Biological Control of Mimosa pigra', Environmental Entomology, pp. 1664-78.

Hawksworth, DL 1995, 'Evolution by association - A history of Symbiosis - Sapp,J', Nature, vol. 374, no. 6525, pp. 841-2.

Hawksworth, DL 2012, 'Managing and coping with names of pleomorphic fungi in a period of transition', International Mycological Association, vol. 3, no. 1, pp. 15-24.

111

Heard, TA, Burcher, JA & Forno, IW 1999, 'Chalcodermus serripes (Coleoptera : Curculionidae) for biological control of Mimosa pigra: Host relations and life cycle', Biological Control, vol. 15, no. 1, pp. 1-9.

Heard, TA 2004, 'The taxonomy of Mimosa pigra L', paper presented to Research and Management of Mimosa pigra, Darwin, <://WOS:000231874800001>.

Heard, T & Paynter, Q 2009, 'Mimosa pigra L. (Leguminosae)', in R Muniappan, G Reddy & A Raman (eds), Biological Control of Tropical Weeds using Arthropods, Cambridge University Press, United Kingdom.

Heard, TA, Elliott, LP, Anderson, B, White, L, Burrows, N, Mira, A, Zonneveld, R, Fichera, G, Chan, R & Segura, R 2010, 'Biology, host specificity, release and establishment of Macaria pallidata and Leuciris fimbriaria (Lepidoptera: Geometridae), biological control agents of the weed Mimosa pigra', Biological Control, vol. 55, no. 3, pp. 248-55. Cambridge University Press, United Kingdom.

Heard, TA 2012a, Mimosa pigra L. - Mimosa, Biological Control of Weeds in Australia.

Heard, TA 2012b, 'Mimosa pigra: In Biological Control of weeds in Australia', in M Julien, REC McFadyen & J Cullen (eds), CSIRO Publishing, Melbourne, p. 21.

Hennecke, BR 2002, 'Host-pathogen Interactions between the fungal pathogen-Mimosae pigrae and Mimosa pigra, giant sensitive plant', Doctor of Philosophy thesis, University of Queensland.

Hennecke, BR 2006, 'Failure of Diabole cubensis, a promising classical biological control agent, to establish in Australia', Biological Control, vol. 39, no. 2, pp. 121-7.

Hoagland, RE, Boyette, C, Weaver, M & Abbas, H 2007, '"Bioherbicides: research and risks', Toxin Reviews, vol. 26, pp. 313-42.

Hogg, EH, Brandt, JP & Kochtubajda, B 2002, 'Growth and dieback of Aspen forests in northwestern Alberta, Canada, in relation to climate and insects', Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere, vol. 32, no. 5, pp. 823-32.

Holland, KL, Telfer, AJ, Walker, GR, Tyerman, SD & Mensforth, LJ 2000, Salinity and floodplain tree health in the lower River Murray, Australia, Groundwater: Past Achievements and Future Challenges.

112

Houle, D, Tremblay, S & Ouimet, R 2007, 'Foliar and wood chemistry of sugar maple along a gradient of soil acidity and stand health', Plant and Soil, vol. 300, no. 1-2, pp. 173-83.

Houston, D 1992, 'A Host-Stress-Saprogen Model for Forest Dieback-Decline Diseases', in P Manion & D Lachance (eds), Forest Decline Concepts, The American Phytopathological Society, St Paul, Minnesota, USA.

Huelma, CC, Moody, K & Mew, TW 1996, 'Weed seeds in rice seed shipments: A case study', International Journal of Pest Management, vol. 42, no. 3, pp. 147-50.

Hyde, KD & Soytong, K 2008, 'The fungal endophyte dilemma', Fungal Diversity, vol. 33, pp. 163- 73.

Innes J. L. (1992) Forest decline. Prog. Phys. Geogr. 16, 1–64.

Jami, F, Slippers, B, Wingfield, MJ & Gryzenhout, M 2013, 'Five new species of the Botryosphaeriaceae from Acacia karroo in South Africa (vol 33, pg 245, 2012)', Cryptogamie Mycologie, vol. 34, no. 1.

Jeltsch, F & Wissel, C 1994, 'Modelling dieback phenomena in natural forests', Ecological Modelling, vol. 75–76, no. 0, pp. 111-21.

Jurskis, V 2005, 'Eucalypt decline in Australia, and a general concept of tree decline and dieback', Forest Ecology and Management, vol. 215, no. 1-3, pp. 1-20.

Kalliola, R, Salo, J, Puhakka, M & Rajasilta, M 1991, 'New site formation and colonizing vegetation in primary succession on the Western Amazon floodplains', Journal of Ecology, vol. 79, no. 4, pp. 877-901.

Kaplan, DA & Munoz-Carpena, R 2014, 'Groundwater salinity in a floodplain forest impacted by saltwater intrusion', Journal of Contaminant Hydrology, vol. 169, pp. 19-36.

Keane, RM & Crawley, MJ 2002, 'Exotic plant invasions and the enemy release hypothesis', Trends in Ecology & Evolution, vol. 17, no. 4, pp. 164-70.

Khakdaman, H, Pourmeidani, A & Salari, SA 2008, 'Investigation of causal agents in decline and mortality of Arizona cypress (Cupressus arizonica) in Qom', Iranian Journal of Forest and Range Protection Research, vol. 6, no. 1, p. 9.

113

Ko, TWK, Stephenson, SL, Bahkali, AH & Hyde, KD 2011, 'From morphology to molecular biology: can we use sequence data to identify fungal endophytes?', Fungal Diversity, vol. 50, no. 1, pp. 113-20.

Lacap, DC, Hyde, KD & Liew, ECY 2003, 'An evaluation of the fungal 'morphotype' concept based on ribosomal DNA sequences', Fungal Diversity, vol. 12, pp. 53-66.

Laurance, WF, Dell, B, Turton, SM, Lawes, MJ, Hutley, LB, McCallum, H, Dale, P, Bird, M, Hardy, G, Prideaux, G, Gawne, B, McMahon, CR, Yu, R, Hero, JM, Schwarzkop, L, Krockenberger, A, Douglas, M, Silvester, E, Mahony, M, Vella, K, Saikia, U, Wahren, CH, Xu, ZH, Smith, B & Cocklin, C 2011, 'The 10 Australian ecosystems most vulnerable to tipping points', Biological Conservation, vol. 144, no. 5, pp. 1472-80.

Lee, JS, Haack, RA & Choi, WI 2011, 'Attack Pattern of Platypus koryoensis (Coleoptera: Curculionidae: Platypodinae) in Relation to Crown Dieback of Mongolian Oak in Korea', Environmental Entomology, vol. 40, no. 6, pp. 1363-9.

Li, JM, Xiao, T, Zhang, Q & Dong, M 2013, 'Interactive Effect of Herbivory and Competition on the Invasive Plant Mikania micrantha', PloS one, vol. 8, no. 5.

Liu, JK, Phookamsak, R, Doilom, M, Wikee, S, Li, YM, Ariyawansha, H, Boonmee, S, Chomnunti, P, Dai, DQ, Bhat, JD, Romero, AI, Zhuang, WY, Monkai, J, Jones, EBG, Chukeatirote, E, Ko, TWK, Zhao, YC, Wang, Y & Hyde, KD 2012, 'Towards a natural classification of Botryosphaeriales', Fungal Diversity, vol. 57, no. 1, pp. 149-210.

Liu, SS, Chen, FZ & Zalucki, MP 2002, 'Development and survival of the diamondback moth (Lepidoptera : Plutellidae) at constant and alternating temperatures', Environmental Entomology, vol. 31, no. 2, pp. 221-31.

Lonsdale, WM 1988, 'Litterfall in an Australian population of Mimosa pigra an invasive tropical shrub', Journal of Tropical Ecology, vol. 4, pp. 381-92.

Lonsdale, WM & Abrecht, DG 1989, 'Seedling Mortality in Mimosa pigra, an invasive tropical shrub', Journal of Ecology, vol. 77, pp. 371-85.

Lonsdale, WM 1992, The Biology of Mimosa pigra A Guide to Management of Mimosa pigra, CSIRO, Canberra.

114

Lotze, HK & Worm, B 2002, 'Complex interactions of climatic and ecological controls on macroalgal recruitment', Limnology and Oceanography, vol. 47, no. 6, pp. 1734-41.

Lygis, V, Bakys, R, Gustiene, A, Burokiene, D, Matelis, A & Vasaitis, R 2014, 'Forest self- regeneration following clear-felling of dieback-affected Fraxinus excelsior: focus on ash', European Journal of Forest Research, vol. 133, no. 3, pp. 501-10.

Mack, RN, Simberloff, D, Lonsdale, WM, Evans, H, Clout, M & Bazzaz, FA 2000, 'Biotic invasions: Causes, epidemiology, global consequences, and control', Ecological Applications, vol. 10, no. 3, pp. 689-710.

Madden, T 2002, 'The NCBI Handbook', in The BLAST sequence analysis tool.

Maharachchikumbura, SSN, Guo, LD, Chukeatirote, E, Bahkali, AH & Hyde, KD 2011, 'Pestalotiopsis-morphology, phylogeny, biochemistry and diversity', Fungal Diversity, vol. 50, no. 1, pp. 167-87.

Maharachchikumbura, SSN, Guo, L-D, Cai, L, Chukeatirote, E, Wu, WP, Sun, X, Crous, PW, Bhat, DJ, McKenzie, EHC, Bahkali, AH & Hyde, KD 2012, 'A multi-locus backbone tree for Pestalotiopsis, with a polyphasic characterization of 14 new species', Fungal Diversity, vol. 56, no. 1, pp. 95-129.

Malanson, GP 1993, Riparian landscapes, Riparian landscapes.

Manion, P & Lachance, D (eds) 1992, Forest Decline Concepts, A Host-Stress-Saprogen Model For Forest Dieback-Decline Diseases, The American Phytopathological Society, USA.

March, N 2004, Prickly acacia Ecology and Threat, Queensland, Australia.

Marques, MW, Lima, NB, Michereff, SJ, Camara, MPS & Souza, CRB 2012, 'First Report of Mango Dieback Caused by Pseudofusicoccum stromaticum in Brazil', Plant Disease, vol. 96, no. 1, pp. 144-5.

Maron, JL & Crone, E 2006, 'Herbivory: effects on plant abundance, distribution and population growth', Proceedings of the Royal Society B-Biological Sciences, vol. 273, no. 1601, pp. 2575-84.

Marsberg, A, Slippers, B, Wingfield, MJ & Gryzenhout, M 2014, 'Endophyte isolations from Syzygium cordatum and a Eucalyptus clone (Myrtaceae) reveal new host and

115

geographical reports for the Mycosphaerellaceae and Teratosphaeriaceae', Australasian Plant Pathology, vol. 43, no. 5, pp. 503-12.

Matusick, G, Ruthrof, KX, Brouwers, NC, Dell, B & Hardy, GS 2013, 'Sudden forest canopy collapse corresponding with extreme drought and heat in a mediterranean-type eucalypt forest in southwestern Australia', European Journal of Forest Research, vol. 132, no. 3, pp. 497- 510.

McClay, AS 1995, 'Beyond "Before-and-After." Experimental Design and evaluation in classical weed biocontrol.', paper presented to In Eighth International Symposium on Biological Control of Weeds, Melbourne, Asutralia.

McDonald, N & McAlpine, J (eds) 1991, Floods and droughts:the northern climate., Monsoonal Australia.Landscape, Ecology and Man in the Northern Lowlands.

McEvoy, PB & Coombs, EM 1999, 'Biological control of plant invaders: Regional patterns, field experiments, and structured population models', Ecological Applications, vol. 9, no. 2, pp. 387-401.

McFadyen, REC 1998, 'Biological control of weeds', Annual Review of Entomology, vol. 43, pp. 369- 93.

McIntyre, DL & Menges, CH 2004, A preliminary conceptual model of remote sensing for detecting small outbreaks of Mimosa pigra, Research and Management of Mimosa pigra.

Mejia, LC, Rojas, EI, Maynard, Z, Van Bael, S, Arnold, AE, Hebbar, P, Samuels, GJ, Robbins, N & Herre, EA 2008, 'Endophytic fungi as biocontrol agents of Theobroma cacao pathogens', Biological Control, vol. 46, no. 1, pp. 4-14.

Meisner, MH, Harmon, JP & Ives, AR 2014, 'Temperature effects on long-term population dynamics in a parasitoid-host system', Ecological Monographs, vol. 84, no. 3, pp. 457-76.

Michailides, TJ, Morgan, DP, Grant, JA & Olson, WH 1992, 'Shorter sprinkler irrigations reduce Botryosphaeria blight of pistachio', California Agriculture, vol. 46, no. 6, pp. 28-32.

Miller, I 1983, 'The distribution and threat of Mimosa pigra in Australia', paper presented to International Symposium on Mimosa pigra, Chiang Mai, Thailand.

116

Miller, IL 1987, 'Uses for Mimosa pigra', in Research and management of Mimosa pigra: papers presented at the 3rd International Symposium on the Management of Mimosa pigra, Darwin, Australia, 22-25 September 2002., pp. 63-7.

Miller, IL & Lonsdale, M 1987, 'Early Records of Mimosa pigra in the Northern Territory', Plant Protection Quarterly, vol. 2, pp. 140-2.

Miller, IL 1988, 'Aspects of the biology and control of Mimosa pigra L.', Master of Science in Agriculture thesis, University of Sydney.

Miller, IL 1991, 'Putting the heat on Mimosa', Plant Protection Quarterly, pp. 28-9.

Miller, TC, Gubler, WD, Geng, S & Rizzo, DM 2003, 'Effects of temperature and water vapor pressure on conidial germination and lesion expansion of Sphaerotheca macularis f. sp fragariae', Plant Disease, vol. 87, no. 5, pp. 484-92.

Mills, AJ 2006, 'The role of salinity and sodicity in the dieback of Acacia xanthophloea in Ngorongoro Caldera, Tanzania', African Journal of Ecology, vol. 44, no. 1, pp. 61-71.

Mitchell, JI, Roberts, PJ & Moss, ST 1995, 'Sequence or Structure? A Short Review on the application of Nucleic Acid Sequence Information to Fungal Taxonomy', Mycologist, vol. 9, no. 2, pp. 67-75.

Mitchell, PJ, Battaglia, M & Pinkard, EA 2013, 'Counting the costs of multiple stressors: is the whole greater than the sum of the parts?', Tree Physiology, vol. 33, no. 5, pp. 447-50.

Montagu, KD 1999, 'Effect of two of two pest on Acacia auriculiformis tree growth and form in Australia's Northern Territory', Journal of Tropical Forest Science vol. 11, no. 2, pp. 492-502.

Morin, L, Evans, KJ & Sheppard, AW 2006, 'Selection of pathogen agents in weed biological control: critical issues and peculiarities in relation to arthropod agents', Australian Journal of Entomology, vol. 45, no. 4, pp. 349-65.

Morin, L, Reid, AM, Sims-Chilton, NM, Buckley, YM, Dhileepan, K, Hastwell, GT, Nordblom, TL & Raghu, S 2009, 'Review of approaches to evaluate the effectiveness of weed biological control agents', Biological Control, vol. 51, no. 1, pp. 1-15.

117

Moricca, S & Ragazzi, A 2008, 'Fungal endophytes in Mediterranean oak forests: A lesson from Discula quercina', Phytopathology, vol. 98, no. 4, pp. 380-6.

Muellerdombois, D 1987, 'Natural dieback in forests', BioScience, vol. 37, no. 8, pp. 575-83.

Mulrennan, ME & Woodroffe, CD 1998, 'Saltwater intrusion into the coastal plains of the Lower Mary River, Northern Territory, Australia', Journal of Environmental Management, vol. 54, no. 3, pp. 169-88.

Nepstad, DC, Stickler, CM, Soares, B & Merry, F 2008, 'Interactions among Amazon land use, forests and climate: prospects for a near-term forest tipping point', Philosophical Transactions of the Royal Society B-Biological Sciences, vol. 363, no. 1498, pp. 1737-46.

Nicolson, TH & Johnston, C 1979, 'Mychorrizhae in the Gramineae 3. Glomus fasciculatus as the endophyte of pioneer grasses in a maritime sand dune', Transactions of the British Mycological Society, vol. 72, no. Apr, pp. 261-8.

Niinemets, U 2010, 'Responses of forest trees to single and multiple environmental stresses from seedlings to mature plants: Past stress history, stress interactions, tolerance and acclimation', Forest Ecology and Management, vol. 260, no. 10, pp. 1623-39.

Northern Territory Government 2003, Mimosa (Mimosa pigra) Weed Management Guide, NRETAS, Northern Territory.

Oberhofer, M & Leuchtmann, A 2014, 'Horizontal transmission, persistence and competition capabilities of Epichloe endophytes in Hordelymus europaeus grass hosts using dual endophyte inocula', Fungal Ecology, vol. 11, pp. 37-49.

Oono, R, Lutzoni, F, Arnold, AE, Kaye, L, U'Ren, JM, May, G & Carbone, I 2014, 'Genetic variation in horizontally transmitted fungal endophyres of pine needles reveals population structure in cryptic species', American Journal of Botany, vol. 101, no. 8, pp. 1362-74.

Ostermeyer, N & Grace, BS 2007, 'Establishment, distribution and abundance of Mimosa pigra biological control agents in northern Australia: implications for biological control', Biocontrol, vol. 52, no. 5, pp. 703-20.

Ostry, ME, Venette, RC & Juzwik, J 2011, 'Decline as a Disease Category: Is it Helpful?', Phytopathology, vol. 101, no. 4, pp. 404-9.

118

Page, AR & Lacey, KL 2006, Economic impact assessment of Australian weed biological control, Australian Weed Management, Adelaide, Australia.

Palik, BJ, Ostry, ME, Venette, RC & Abdela, E 2011, 'Fraxinus nigra (black ash) dieback in Minnesota: Regional variation and potential contributing factors', Forest Ecology and Management, vol. 261, no. 1, pp. 128-35.

Pascoe, I & Cottral, E 2000, 'Developments in grapevine trunk diseases research in Australia', Phytopathologia Mediterranea, vol. 39, no. 1, pp. 68-75.

Pavlic, D, Slippers, B, Coutinho, TA & Wingfield, MJ 2007, 'Botryosphaeriaceae occurring on native Syzygium cordatum in South Africa and their potential threat to Eucalyptus', Plant Pathology, vol. 56, no. 4, pp. 624-36.

Pavlic, D, Wingfield, BD, Barber, P, Slippers, B, Hardy, GES & Burgess, TI 2008, 'Seven New Species of the Botryosphaeriaceae from Baobab and Other Native Trees in Western Australia', Mycologia, vol. 100, no. 6, pp. 851-66.

Paynter, Q & Hennecke, B 2001, 'Competition between two biological control agents, Neurostrota gunniella and Phloeospora mimosae-pigrae, and their impact on the invasive tropical shrub Mimosa pigra', Biocontrol Science and Technology, vol. 11, no. 5, pp. 575-82.

Paynter, Q 2005, 'Evaluating the impact of a biological control agent Carmenta mimosa on the woody wetland weed Mimosa pigra in Australia', Journal of Applied Ecology, vol. 42, no. 6, pp. 1054-62.

Paynter, Q 2006, 'Evaluating the impact of biological control against Mimosa pigra in Australia: Comparing litterfall before and after the introduction of biologial control agents', Biological Control, vol. 38, pp. 166-73.

Pearson, RG & Dawson, TP 2003, 'Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful?', Global Ecology and Biogeography, vol. 12, no. 5, pp. 361-71.

Perez, CA, Wingfield, MJ, Slippers, B, Altier, NA & Blanchette, RA 2010, 'Endophytic and canker- associated Botryosphaeriaceae occurring on non-native Eucalyptus and native Myrtaceae trees in Uruguay', Fungal Diversity, vol. 41, no. 1, pp. 53-69.

119

Petrini, O, Sieber, TN, Toti, L & Viret, O 1992, 'Ecology, metabolite production, and substrate utilization in endophytic fungi', Natural Toxins, vol. 1, no. 3, pp. 185-96.

Phillips, AJL, Alves, A, Abdollahzadeh, J, Slippers, B, Wingfield, MJ, Groenewald, JZ & Crous, PW 2013, 'The Botryosphaeriaceae: genera and species known from culture', Studies in Mycology, no. 76, pp. 51-167.

Photita, W, Lumyong, S, Lumyong, P, McKenzie, EHC & Hyde, KD 2004, 'Are some endophytes of Musa acuminata latent pathogens?', Fungal Diversity, vol. 16, pp. 131-40.

Pitt, WM, Huang, R, Steel, CC & Savocchia, S 2013, 'Pathogenicity and epidemiology of Botryosphaeriaceae species isolated from grapevines in Australia', Australasian Plant Pathology, vol. 42, no. 5, pp. 573-82.

Promputtha, I, Lumyong, S, Dhanasekaran, V, McKenzie, EHC, Hyde, KD & Jeewon, R 2007, 'A phylogenetic evaluation of whether endophytes become saprotrophs at host senescence', Microbial Ecology, vol. 53, no. 4, pp. 579-90.

Ploch, S & Thines, M 2011, 'Obligate biotrophic pathogens of the genus Albugo are widespread as asymptomatic endophytes in natural populations of Brassicaceae', Molecular Ecology, vol. 20, no. 17, pp. 3692-9.

Podger, FD 1981, 'Some difficulties in the diagnosis of drought as the cause of dieback', in KM Old, GA Kile & CP Ohmart (eds), Eucalypt Dieback in Forests and Woodlands, CSIRO, Australia, pp. 167-73.

Raffa, KF, Aukema, BH, Bentz, BJ, Carroll, AL, Hicke, JA, Turner, MG & Romme, WH 2008, 'Cross-scale drivers of natural disturbances prone to anthropogenic amplification: The dynamics of bark beetle eruptions', BioScience, vol. 58, no. 6, pp. 501-17.

Ramaling, A & Rati, E 1974, 'Simple Techniques for Spore Germination Studies', Current Science, vol. 43, no. 2, pp. 60.

Ramos, LJ, Lara, SP, McMillan, RT & Narayanan, KR 1991, 'Tip dieback of mango (Mangifera indica) caused by Botryosphaeria ribis', Plant Disease, vol. 75, no. 3, pp. 315-8.

120

Rayment, GE & Lyons, DJ 2011, Soil chemical methods: Australasia, vol. 3., CSIRO, Collingwood, Vic.

Rayner, RW 1970, A Mycological Colour Chart, CAB.

Redlin, S & Carris, L (eds) 1996, Endophytic Fungi in Grasses and Woody Plants, Systematics, Ecology and Evolution, Latent Infection vs Endophytic Colonization by Fungi, The American Phytopathological Society, Illinois.

Ren, D, Madsen, JS, de la Cruz-Perera, CI, Bergmark, L, Sorensen, SJ & Burmolle, M 2014, 'High- throughput screening of multispecies biofilm formation and quantitative PCR-based assessment of individual species proportions, useful for exploring interspecific bacterial interactions', Microb Ecol, vol. 68, no. 1, pp. 146-54.

Rodriguez, RJ, Redman, RS & Henson, JM 2004, 'The role of fungal symbioses in the adaptation of plants to high stress environments', Mitigation and adaptation strategies for global change, vol. 9, no. 3, pp. 261-72.

Rodriguez-Garcia, C, Medina, V, Alonso, A & Ayllon, MA 2014, 'Mycoviruses of Botrytis cinerea isolates from different hosts', Annals of Applied Biology, vol. 164, no. 1, pp. 46-61.

Rocha, ACS, Garcia, D, Uetanabaro, APT, Carneiro, RTO, Araujo, IS, Mattos, CRR & Goes-Neto, A 2011, 'Foliar endophytic fungi from Hevea brasiliensis and their antagonism on Microcyclus ulei', Fungal Diversity, vol. 47, no. 1, pp. 75-84.

Rodriguez, RJ, White, JF, Arnold, AE & Redman, RS 2009, 'Fungal endophytes: diversity and functional roles', New Phytologist, vol. 182, no. 2, pp. 314-30.

Rollinger, JL & Langenheim, JH 1993, 'Geographic survey of fungal endophyte community composition in leaves of Coastal Redwood', Mycologia, vol. 85, no. 2, pp. 149-56.

Rudgers, JA, Fischer, S & Clay, K 2010, 'Managing plant symbiosis: fungal endophyte genotype alters plant community composition', Journal of Applied Ecology, vol. 47, no. 2, pp. 468- 77.

Saikkonen, K, Wali, P, Helander, M & Faeth, SH 2004, 'Evolution of endophyte-plant symbioses', Trends in Plant Science, vol. 9, no. 6, pp. 275-80.

121

Scarlett, K, Guest, D & Daniel, R 2013, 'Elevated soil nitrogen increases the severity of dieback due to Phytophthora cinnamomi', Australasian Plant Pathology, vol. 42, no. 2, pp. 155-62.

Schulz, B & Boyle, C 2005, 'The endophytic continuum', Mycological Research, vol. 109, pp. 661- 86.

Schulz, B, Wanke, U, Draeger, S & Aust, HJ 1993, 'Endophytes from herbaceous plants and shrubs - effectiveness of surface sterilization methods', Mycological Research, vol. 97, pp. 1447-50.

Schoch, CL, Seifert, KA, Huhndorf, S, Robert, V, Spouge, JL, Levesque, CA, Chen, W, Bolchacova, E, Voigt, K, Crous, PW, Miller, AN, Wingfield, MJ, Aime, MC, An, KD, Bai, FY, Barreto, RW, Begerow, D, Bergeron, MJ, Blackwell, M, Boekhout, T, Bogale, M, Boonyuen, N, Burgaz, AR, Buyck, B, Cai, L, Cai, Q, Cardinali, G, Chaverri, P, Coppins, BJ, Crespo, A, Cubas P, P, Cummings, C, Damm, U, de Beer, ZW, de Hoog, GS, Del- Prado, R, Dentinger, B, Dieguez-Uribeondo, J, Divakar, PK, Douglas, B, Duenas, M, Duong, TA, Eberhardt, U, Edwards, JE, Elshahed, MS, Fliegerova, K, Furtado, M, Garcia, MA, Ge, ZW, Griffith, GW, Griffiths, K, Groenewald, JZ, Groenewald, M, Grube, M, Gryzenhout, M, Guo, LD, Hagen, F, Hambleton, S, Hamelin, RC, Hansen, K, Harrold, P, Heller, G, Herrera, G, Hirayama, K, Hirooka, Y, Ho, HM, Hoffmann, K, Hofstetter, V, Hognabba, F, Hollingsworth, PM, Hong, SB, Hosaka, K, Houbraken, J, Hughes, K, Huhtinen, S, Hyde, KD, James, T, Johnson, EM, Johnson, JE, Johnston, PR, Jones, EB, Kelly, LJ, Kirk, PM, Knapp, DG, Koljalg, U, Kovacs, GM, Kurtzman, CP, Landvik, S, Leavitt, SD, Liggenstoffer, AS, Liimatainen, K, Lombard, L, Luangsa-Ard, JJ, Lumbsch, HT, Maganti, H, Maharachchikumbura, SS, Martin, MP, May, TW, McTaggart, AR, Methven, AS, Meyer, W, Moncalvo, JM, Mongkolsamrit, S, Nagy, LG, Nilsson, RH, Niskanen, T, Nyilasi, I, Okada, G, Okane, I, Olariaga, I, Otte, J, Papp, T, Park, D, Petkovits, T, Pino-Bodas, R, Quaedvlieg, W, Raja, HA, Redecker, D, Rintoul, T, Ruibal, C, Sarmiento-Ramirez, JM, Schmitt, I, Schussler, A, Shearer, C, Sotome, K, Stefani, FO, Stenroos, S, Stielow, B, Stockinger, H, Suetrong, S, Suh, SO, Sung, GH, Suzuki, M, Tanaka, K, Tedersoo, L, Telleria, MT, Tretter, E, Untereiner, WA, Urbina, H, Vagvolgyi, C, Vialle, A, Vu, TD, Walther, G, Wang, QM, Wang, Y, Weir, BS, Weiss, M, White, MM, Xu, J, Yahr, R, Yang, ZL, Yurkov, A, Zamora, JC, Zhang, N, Zhuang, WY, Schindel, D & Fungal Barcoding, C 2012, 'Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi', Proceedings of the

122

National Academy of Sciences of the United States of America, vol. 109, no. 16, pp. 6241-6.

Selosse, MA & Schardl, CL 2007, 'Fungal endophytes of grasses: hybrids rescued by vertical transmission? An evolutionary perspective', New Phytologist, vol. 173, no. 3, pp. 452-8.

Shah, MD, Verma, KS, Singh, K & Kaur, R 2010, 'Morphological, pathological and molecular variability in Botryodiplodia theobromae (Botryosphaeriaceae) isolates associated with die-back and bark canker of pear trees in Punjab, India', Genetics and Molecular Research, vol. 9, no. 2, pp. 1217-28.

Simberloff, D & Gibbons, L 2004, 'Now you see them, now you don't - population crashes of established introduced species', Biological Invasions, vol. 6, no. 2, pp. 161-72.

Sims-Chilton, NM, Zalucki, MP & Buckley, YM 2009, 'Patchy herbivore and pathogen damage throughout the introduced Australian range of groundsel bush, Baccharis halimifolia, is influenced by rainfall, elevation, temperature, plant density and size', Biological Control, vol. 50, no. 1, pp. 13-20.

Sims-Chilton, NM, Zalucki, MP & Buckley, YM 2010, 'Long term climate effects are confounded with the biological control programme against the invasive weed Baccharis halimifolia in Australia', Biological Invasions, vol. 12, no. 9, pp. 3145-55.

Sinclair, JB & Cerkauskas, RF 1996, Latent infection vs. endophytic colonization by fungi, Endophytic fungi in grasses and woody plants: Systematics, ecology, and evolution.

Slepecky, RA & Starmer, WT 2009, 'Phenotypic Plasticity in Fungi: A Review with Observations on Aureobasidium pullulans', Mycologia, vol. 101, no. 6, pp. 823-32.

Smith, C & Wilson, C 1995, 'Close to the edge: microhabitat selection by Neurostrota gunniella: (Lepidoptera: Gracillariidae), a biological control agent for Mimosa pigra L. in Australia', J. Aust. Entmol. Soc, vol. 34, pp. 177-80.

Smith, D & Onions, HS 1983, The preservation and maintenance of living fungi, Commonwealth Mycological Institute, Kew.

Smith, SM, Medeiros, KC & Tyrrell, MC 2012, 'Hydrology, Herbivory, and the Decline of Spartina patens (Aiton) Muhl. in Outer Cape Cod Salt Marshes (Massachusetts, USA)', Journal of Coastal Research, vol. 28, no. 3, pp. 602-12.

123

Steinbauer, MJ, Edwards, PB, Hoskins, M, Schatz, T & Forno, IW 2000, 'Seasonal abundance of insect biocontrol agents of Mimosa pigra in the Northern Territory', Australian Journal of Entomology, vol. 39, pp. 328-35.

Steinbauer, MJ 1998, 'Host plant phenotype and the impact and development of Carmenta mimosa, a biological control agent of Mimosa pigra in Australia', Biological Control, vol. 13, no. 3, pp. 182-9.

Strobel, G, Yang, XS, Sears, J, Kramer, R, Sidhu, RS & Hess, WM 1996, 'Taxol from Pestalotiopsis microspora, an endophytic fungus of Taxus wallachiana', Microbiology-Uk, vol. 142, pp. 435-40.

Strobel, G & Daisy, B 2003, 'Bioprospecting for microbial endophytes and their natural products', Microbiology and Molecular Biology Reviews, vol. 67, no. 4, pp. 491-495.

Strobel, G, Daisy, B, Castillo, U & Harper, J 2004, 'Natural products from endophytic microorganisms', Journal of Natural Products, vol. 67, no. 2, pp. 257-68.

Strobel, GA 2003, 'Endophytes as sources of bioactive products', Microbes and Infection, vol. 5, no. 6, pp. 535-44.

Strobel, GA, Hess, WM, Ford, E, Sidhu, RS & Yang, X 1996, 'Taxol from fungal endophytes and the issue of biodiversity', Journal of Industrial Microbiology & Biotechnology, vol. 17, no. 5-6, pp. 417-23.

Su, YY, Guo, LD & Hyde, KD 2010, 'Response of endophytic fungi of Stipa grandis to experimental plant function group removal in Inner Mongolia steppe, China', Fungal Diversity, vol. 43, no. 1, pp. 93-101.

Suciatmih & Rahmansyah, M 2013, 'Endophytic fungi isolated from mangrove plant and have antagonism role against Fusarium wilt', Journal of Agricultural and Biological Science, vol. 8, no. 3, pp. 251-7.

Sutherst, RW & Maywald, GF 1985, 'A computerized system for matching climates in Ecology', Agriculture Ecosystems & Environment, vol. 13, no. 3-4, pp. 281-99.

Swofford, D 2003, PAUP*. Phylogenetic analysis using parsimony., 4 edn.

Tanada, Y. Kaya, H.K., 1993. Insect Pathology. Academic Press, Boston.

124

Tang, L, Gao, Y, Wang, C-h, Zhao, B & Li, B 2012, 'A plant invader declines through its modification to habitats: A case study of a 16-year chronosequence of Spartina alterniflora invasion in a salt marsh', Ecological Engineering, vol. 49, pp. 181-5.

Taylor, K, Barber, PA, Hardy, GES & Burgess, TI 2009, 'Botryosphaeriaceae from tuart (Eucalyptus gomphocephala) woodland, including descriptions of four new species', Mycological Research, vol. 113, pp. 337-53.

Thomas, MB & Reid, AM 2007, 'Are exotic natural enemies an effective way of controlling invasive plants?', Trends in Ecology & Evolution, vol. 22, no. 9, pp. 447-53.

Thompson, S, Tan, Y, Shivas, R, Neate, S, Morin, L, Bissett, A & Aitken, E 2014, 'Green and brown bridges between weeds and crops reveal novel Diaporthe species in Australia', Persoonia-Molecular Phylogeny and Evolution of Fungi.

Thorp, J & Lynch, R 2000, National Weeds Strategy Executive Committee, Launceston.

Thorp, J & Lynch, R 2000, The Determination of the Weeds of National Significance, Launceston, Tasmania: National Weeds Executive Committee.

Tintjer, T, Leuchtmann, A & Clay, K 2008, 'Variation in horizontal and vertical transmission of the endophyte Epichloe elymi infecting the grass Elymus hystrix', New Phytologist, vol. 179, no. 1, pp. 236-46.

Underwood, AJ 2009, 'Components of design in ecological field experiments', Annales Zoologici Fennici, vol. 46, no. 2, pp. 93-111.

Upadhyay, MK, Jain, D, Tiwari, K, Singh, A & Verma, HN 2011, 'Exploitation of Fungi: A potential approach for the management of weeds', Proceedings of the National Academy of Sciences India Section Biological Sciences, vol. 81, pp. 69-75.

Urbez-Torres, JR, Peduto, F, Vossen, PM, Krueger, WH & Gubler, WD 2013, 'Olive Twig and Branch Dieback: Etiology, Incidence, and Distribution in California', Plant Disease, vol. 97, no. 2, pp. 231-44.

Urbez-Torres, JR & Gubler, WD 2009, 'Pathogenicity of Botryosphaeriaceae Species Isolated from Grapevine Cankers in California', Plant Disease, vol. 93, no. 6, pp. 584-92.

125

van Klinken, RD & Raghu, S 2006, 'A scientific approach to agent selection', Australian Journal of Entomology, vol. 45, pp. 253-8. van Klinken, RD, Lawson, BE & Zalucki, MP 2009, 'Predicting invasions in Australia by a Neotropical shrub under climate change: the challenge of novel climates and parameter estimation', Global Ecology and Biogeography, vol. 18, no. 6, pp. 688-700.

Volkmar, KM, Hu, Y & Steppuhn, H 1998, 'Physiological responses of plants to salinity: A review', Canadian Journal of Plant Science, vol. 78, no. 1, pp. 19-27.

Walden, D, Dam, Rv, Finlayson, M, Storrs, M, Lowry, J & Kriticos, D 2002, 'A risk assessment of the tropical wetland weed Mimosa pigra in northern Australia', in Research and management of Mimosa pigra: papers presented at the 3rd International Symposium on the Management of Mimosa pigra, Darwin, Australia, 22-25 September 2002., pp. 11- 21.

Wanichanantakul, P & Chinawong, S 1979, 'Some aspects of the biology of Mimosa pigra in northern Thailand', paper presented to Proceedings of the Seventh Asian Pacific Weed Science Society Conference, Thailand.

Wargo, PM 1996, 'Consequences of environmental stress on oak: Predisposition to pathogens', Annales Des Sciences Forestieres, vol. 53, no. 2-3, pp. 359-68.

Watmough, SA 2010, 'Assessment of the potential role of metals in sugar maple (Acer saccharum Marsh) decline in Ontario, Canada', Plant and Soil, vol. 332, no. 1-2, pp. 463-74.

Watson, A (ed.) 1991, Microbial Control of Weeds, The Classical Approach with Plant Pathogens, Chapman and Hall, New York.

Wilkinson, HH, Siegel, MR, Blankenship, JD, Mallory, AC, Bush, LP & Schardl, CL 2000, 'Contribution of fungal loline alkaloids to protection from aphids in a grass-endophyte mutualism', Molecular Plant-Microbe Interactions, vol. 13, no. 10, pp. 1027-33.

Whipps, JM 1987, 'Effect of media on growths and interactions between a range of soil‐borne glasshouse pathogens and antagonistic fungi', New Phytologist, vol. 107, no. 1, pp. 127-42.

Wilson, CG & Flanagan, GJ 1991, 'Establishment of Acanthoscelides-Quadridentatus (Schaeffer) and a Puniceus Johnson (Coleoptera, Bruchidae) on Mimosa pigra in Northern Australia', Journal of the Australian Entomological Society, vol. 30, pp. 279-80.

126

Wilson, CG & Pitkethley, RN 1992, 'Botryodiplodia dieback of Mimosa pigra, a noxious weed in Northern Australia', Plant Pathology, vol. 41, no. 6, pp. 777-9.

Wilson, D 1995, 'Endophyte-the evolution of term, a classification of its use and definition', Oikos, vol. 73, pp. 274-6.

Woodroffe, CD 1995, 'Response of tide-dominated mangrove shorelines in Northern Australia to anticipated sea-level rise', Earth Surface Processes and Landforms, vol. 20, no. 1, pp. 65-85.

Yan, J-Y, Xie, Y, Zhang, W, Wang, Y, Liu, J-K, Hyde, KD, Seem, RC, Zhang, G-Z, Wang, Z-Y, Yao, S-W, Bai, X-J, Dissanayake, AJ, Peng, Y-L & Li, X-H 2013, 'Species of Botryosphaeriaceae involved in grapevine dieback in China', Fungal Diversity, vol. 61, no. 1, pp. 221-36.

Yeoh, PB, Fontanini, L, Cranfield, RJ & Scott, JK 2006, 'How bad is blackberry in southwest Australia?', in 15th Australasian Weeds Conference, Adelaide, Australia, p. 293.

Zhang, LY, Fu, YP, Xie, JT, Jiang, DH, Li, GQ & Yi, XH 2009, 'A novel virus that infecting hypovirulent strain XG36-1 of plant fungal pathogen Sclerotinia sclerotiorum', Virology Journal, vol. 6.

127

Appendices

Appendix A: Permit to use a declared weed from Northern Territory Government (Department of Land Resource Management)- scanned document

128

129

Appendix B: Ecological Studies Data Sheets ( Chapter 3) Data sheet(Mimosa pigra)

Quadrat ID:______Date: ______

Samplers: ______Time of sampling: ______

Coordinates:______Time Ended: ______

SITE/QUADRAT DESCRIPTION

Disturbance & weather conditions: ______

Presence of other native plants: ______

Density and extent of weed population: ______

Herbaceous cover: ______

Is the site representative of surrounding mimosa: ______

Photo captured: ______

WHOLE PLOT COUNT

Class Size Count Alive Dead

Seedlings < 20cm Juveniles 0.2 – 1.5m Adult > 1.5

130

INDIVIDUAL PLANT GROWTH AND SURVIVAL

Characteristics Plant A Plant B Plant C Plant D Plant E Tag Label Disease Rating (1-10) Pathogen Activity (1-3) Height of Living Tree Height of Dead Tree Width of Living Tree 1 Width of Living Tree 2 Width of Dead Tree 1 Width of Dead Tree 2 % of stems living in top

1m height Presence of dead wood thicker than 2cm Presence of dead wood thicker than 5cm Presence of dropped leaves but green stems Diameter of 3 main stems @30cm Other descriptions Pathogen photos INSECT SURVEY

InsectType Plant 1 Plant2 Plant 3 Plant 4 Plant 5 Count of Carmenta with frass Count of Carmenta without frass Count of Maroga with frass Count of Maroga without frass Count of Platyomois present on stems Specimen collection

Type of sample  Collected  Labelled Soil for pH and EC Stems for isolation 50-cm long stem tip for Neurostrota count Loggers set-up and labels

131

Appendix C: Soil pH and Electrical Conductivity (EC) Determination (1:5 soil/water method)

Label all containers using masking tape with your initials, date and sample name on it.

METHOD:

1. Weigh out 20g of air dry soil into a labelled 250ml wide mouth plastic screw capped bottle. 2. Screw lid on tightly and place on shaker. 3. Shake at 160rpm for 60 minutes. 4. Allow to stand undisturbed for 30 mins 5. Rinse conductivity probes in the undisturbed supernatant. Insert probes and read EC off meter. Rinse probes between samples with deionised water. 6. Replace lid and shake bottle. 7. Insert pH probes into the freshly mixed sample and read pH off meter. Complete EC measurement within 3-4 hours of obtaining the aqueous solution.

Method adapted from “Australian Laboratory Handbook of Soil and Water Chemical Methods” (Rayment & Higginson, 2011).

132

Appendix D: Representative conidia of Botryosphaeriaceae morphotypes

Appendix D are photomicrographs of different Botryosphaeriaceae,codes are: (a=MP1296; b=MP1286;c=MP569d=MP1134;e=MP1122;f=MP586;g=MP482;h=MP184;i=MP760;j=MP020;k=MP1255;l=MP876; m=MP554;n=MP722;o=MP564;p=MP433;q=MP321;r=MP389;s=MP143.

133

Appendix E: Rationalisation table used to compare different characters.

Morphotype Cultural characters Sporulation Conidia size ITS ID EF1 ID When compared to Final Designation s holotypes and ex-type isolates

Bortyo Fluffy, thick aerial Yes, mature 21-25µm length, Lasidiplodia Lasiodiplodia Lasiodiplodia iraniensis Lasiodiplodia Group A mycelium,slow growing, conidia on pine theobromae iraniensis iraniensis striate, grey olivaceous to black needles under after 8 days incubation; black light after brown, carpet-like with age 45 days dark-walled

13-15µm width

Botryo Cottony grey on margin, No, immature n/a Lasidiplodia Lasidiplodia Lasiodiplodia Lasiodiplodia Group B fast-growing compared to hyaline conidia pseudotheobrom iraniensis iraniemsis iraniensis Botryo A, greenish on pine needles ae olivaceous to black; not under black carpet-like with age light

Botryo Fast-growing, and has a No, immature n/a Lasiodiplodia Lasiodiplodia Lasiodiplodia Lasiodiplodia Group C unique polygon-shaped spores hyaline theobromae hormozganensis pseudotheobromae pseudotheobrom

green pigmentation in the conidia under ae

middle, white at start and black light

becomes dark mouse grey Lasiodiplodia Lasiodiplodia Lasiodiplodia iraniensis with age pseudotheobrom iraniensis ae Lasiodiplodia iraniensis Lasiodiplodia pseudotheobro mae

134

Botryo Very fine aerial mycelium Yes, mature 19-23µm length, Lasiodiplodia Lasiodiplodia Lasiodiplodia iraniensis Lasiodiplodia Group D concentrated on the conidia striate, theobromae iraniensis iraiensis centre of the plate, Brown, cottony edge, has a dark slate blue; irregular Dark-walled margin

10-14µm

Botryo Slowest growing (30mm Unable to n/a Cophinforma Cophinforma Cophinforma eucalypti Cophinforma Group E after 2 days sporulate under eucalypti atrovirens eucalypti

incubation)among all black light Cophinforma Botryo groups, irregular atrovirens margin and less aerial Cophinforma Cophinforma

mycelium, whitish at start atrovirens atrovirens and turned to olivaceous

after 3 days MP 1209 is M. phaseolina (sample 43)

Botryo Culturally similar to E but Yes. Mature 13-22µm length, Lasiodiplodia Lasiodiplodia Lasiodiplodia iraniensis Lasiodiplodia Group F fast growing, colour conidia striate, theobromae iraniensis iraniensis change from white to Dark-brown, mouse grey. Lasiodiplodia pseudotheobrom

ae 10-14µm

Botryo Very fast growing, reached Unable to n/a Sequence failed Sequence failed Group G 90mm after 3 days, grey sporulate under olovaceous margin and black light olivaceous black reverse colour.

135

Botryo Very cottony, aerial No. immature n/a Lasiodiplodia Lasiodiplodia Lasiodipodia Group H mycelium thin, not carpet- conidia theobromae iraniensis iraniensis like, reverse and front colour is leaden black and a smoke grey cottony margin

Botryo The same with Group H Unable to n/a Lasiodiplodia Lasiodiplodia Lasiodiplodia Group I culturally except that sporulate theobromae iraniensis iraniensis

reverse colour is lighter;

olivaceous grey front and reverse color Neofusicoccum Neofusicoccum Neofusicoccum Neofusicoccum parvum brasiliense brasiliense parvum

Botryo Flat and very slow- Unable to n/a Sequencing failed Sequencing n/a Group J growing, with greenish sporulate failed black irregular margin and smoke grey front colour

Pestalotiopsi Fast-growing, flat, Acervullus 25-31 µm length, Pestalotiopsis Pestalotiopsis P. ellipsospora s Group A irregular margin, started (pionnote) microspora sp. 5-8µm width as powdery white and produced after

changed to pale yellow 2 weeks 3-4 conidiia cells, after 5 days of incubation incubation in terminal Pestalotiopsis sp. Pestalotiospsis P. ellipsospora in the dark the dark on ½ appendages 2-4 glandicola PDA

Pestalototio Slow-growing, circular, Black oily 19-25µm length, Pestalotiopsis Pestalotiopsis Tree and holotypes psis Group B flat, crateriform. Starts as acervullus 6-8µm width, mangiferae sp.

very white colony then observed Conidial changes to pale luteus appendage 2-3 Pestalotiospsis Pestalotiopsis P. clavispora microspora clavispora

136

Pestalotiopsis Pestalotiopsis P. clavispora microspora clavispora

Pestalotiopsis Pestalotiopsis P. clavipsora clavispora clavispora

Pestalotiopsis Pestalotiopsis clavispora sp. P. clavispora

Diaporthe Slow growing, usually Yes, Beta Diaporthe Diaporthe D. phaseolorum Group A whitish in colour, flat and conidia on phaseolorum pseudophoenico

has irregular margin. oatmeal agar la

Produced pycnidia on OA after 35 days

after 40days Gibberella Phomopsis sp. moniliformis Gibberella moniliformis

Diaporthe Slow-growing, usually Yes, Beta Diaporthe Diaporthe sp. Diaporthe sp. Group B whitish in colour on PDA. conidia phaseolorum Flat, irregular margin media becomes light yellow to brown with age. It produced pycnidia on OA after 35 days of incubation in the dark.

137

Aternaria Dark herbage green, fast- Yes, it Alternaria A. alternat like growing, ebtire margin sporulated afte alternata a 25 d incubation

Cladosporiu Slow-growing, green color, Did not Cladosporium B. gloedpo m like flat, irregular margin sporulate gloeosporioides riodes

Colletotrichu Circular, white at start and Yes, sporulates. Not sequenced n/a m becomes grey with age. Produced orange mass of spores

Epicoccum Slow-growing, produces Yes, sporulated Sequencing failed Sequencing n/a yellow pigmentation when under black failed sporulated in the dark light, spore is circular, dark- brown colour

Fusarium Flower-like, pinkish to Yes, banana- Fusarium Fusarium Fusarium sp. violet, flat mycelium goes shaped spore verticillioides pseudocircinatu inside agar. observes-used m scotch-tape technique

Unidentified White at start, with yellow No, did not Diaporthe Diaporthe sp. 1 pigmentation, raised on sporulates after musigena the edge and has an many attempts irregular margin

Unidentified Slow-growing, greenish No, did not Lasiodiplodia Lasiodiplodia L. 2 black back colour and sporulate after theobromae hormozganensis hormozganensis olivaceous grey front many attempts colour, it has a very irregular margin; not fluffy

Unidentified Slow-growing, 30mm after No, did not Diaporthe sp. Diaporthe sp. 3 3 days, irregular and sporulate after many attempts

138

filamentous; not fluffy or cottony

Unidentified Slow-growing, and with No, did not Lasiodiplodia Lasiodiplodia L. iraniensis 4 irregular margin, colour is sporulate after pseudotheobrom iraniensis

white at start and many attempts ae

becomes light grey with

age. Cophinforma C.atrovirens Cophinforma atrovirens atrovirens

Cophinforma Cophinforma eucalypti eucalypti C. eucalyp ti

Unidentified Green pigmentation, slow No, did not Not sequenced Not sequenced n/a 5 growing, flat and has very sporulate after less aerial mycelium. many attempts

139

Appendix F: Maximum parsimony of Pesalotia EF sequences.

(P. clavispora)

(P.ellipsospora)

(P. clavispora) Pestalotiopsis

(Pestalotiopsis sp.)

spp.

(P. ellipsospora)

(Pestalotiopsis sp.)

140

Appendix G: Nursery Propagation Media

Constituents for 1 cubic metre:

 600L Medium Grade Perlite  300L Grade 3 Vermiculite  100L Sphagnum Peat  2kg Basacote Mini 3M Slow-Release Fertiliser (lasts approximately 3 months, less in hot weather). NPK Ratio: 13-6-16.  1kg Dolomite Lime (to raise pH level due to the acidity of the sphagnum peat).

The above constituents are placed into the media mixer together and the mixer is turned on for about 30 seconds to combine. Then, approximately 6-7 minutes of running water from a hose is added. The mixer is turned on for a further 2 minutes and the media is then ready to use.

Nursery Potting Media

The nursery’s potting media consists of the following constituents per cubic metre;

 1.0m3 0-10mm composted bark

Nutritional Additives;

 1kg- Osmocote Exact 3-4 month release NPK: 16+4.4+8.3 plus TE.  2kg Osmocote Plus 8-9 month release. NPK: 16+309+9.1 plus 1.5 Mg and TE.  2 kg Nitrocote 7 month release. NPK 1.8+2.2+11 plus TE  1.3 kg- Coated Iron. 28% Iron (Fe) and 17% Sulphur (S).  1.2 kg Saturaid. Granular wetting agent, helps with wetting of media.  1.3kg Dolomite The composted bark and nutritional additives are added to the media mixer together and mix for approximately 5-6 minutes, water may be required if the bark feels dry.

141

Appendix H: Sequence Data Identification and its corresponding MP (Mimosa pigra) Codes.

142

26 MP 1260 Lasiodiplodia pseudotheobromae 99% 36 MP 1260 Lasiodiodiplodia iraniensis 99% 27 MP 853 Lasiodiplodia theobromae 100% 37 MP 853 Lasiodiodiplodia iraniensis 99% 28 MP 312 Lasiodiplodia theobromae 100% 38 MP 312 Lasiodiodiplodia iraniensis 100% 29 MP 19 Lasiodiplodia theobromae 100% 39 MP 19 Lasiodiodiplodia iraniensis 100% 30 MP 1611 Diaporthe phaseolourum 99% 40 MP 1611 sequence failed 31 MP 1464 sequence failed 100% 41 MP 1464 C. eucalypti 100% 32 MP 1790 sequence failed 100% 42 MP 1790 C. eucalypti 100% 33 MP 1402 Fusarium verticillioides 100% 43 MP 1402 Fusarium pesudocircinatum 99% 34 MP 586 L. theobromae 99% 44 MP 586 Lasiodiodiplodia iraniensis 100% 35 MP 1682 Pestalotiopsis microspora 100% 45 MP 1682 Pestalotiopsis clavispora 97% 36 MP 1202 Lasiodiplodia theobromae 99% 46 MP 1202 Lasidiplodia pseudotheobromae 99% 37 MP 1770 Pestalotiopsis clavispora 47 MP 1770 Pestalotiopsis foedans 98% 38 MP 1748 Pestalotiopsis sp. 100% 48 MP 1748 Pestalotiopsis sp. 99% 39 MP 1678 Pestalotiopsis sp. 100% 49 MP 1678 Pestalotiopsis sp. 100% 40 MP 1716 Pestalotiopsis sp. 99% 6 MP 1716 Pestalotiopsis clavispora 98% 41 MP 272 Lasiodiplodia pseudotheobromae 100% 7 MP 272 Lasidiplodia pseudotheobromae 100% 42 MP 18 Lasiodiplodia theobromae 100% 10 MP 18 Lasiodiodiplodia iraniensis 100% 43 MP 1209 B. mamane 99% 9 MP 1209 C. eucalypti 100% 44 MP 1126 Fusarium verticillioides 100% 5 MP 1126 Fusarium pesudocircinatum 100% 45 MP 544b Fusarium verticillioides 100% 8 MP 544b Fusarium pesudocircinatum 100% 46 MP 212 Lasiodiplodia theobromae 100% 1 MP 212 Lasiodiodiplodia iraniensis 100% 47 MP 432 Lasiodiplodia pseudotheobromae 99% 2 MP 432 Lasiodiodiplodia iraniensis 99% 48 MP 384 Lasiodiplodia theobromae 99% 3 MP 384 Lasidiplodia pseudotheobromae 99% 49 MP 1263 sequence failed 99% 4 MP 1263 sequence failed 93% 50 MP 10 Lasiodiplodia theobromae 100% 60 MP 10 Lasiodiodiplodia iraniensis 99% 51 MP 68 Lasiodiplodia theobromae 100% 61 MP 68 Lasiodiodiplodia iraniensis 99% 52 MP 1171 Lasiodiplodia theobromae 98% 62 MP 1171 Lasiodiodiplodia iraniensis 99% 53 MP 1321 Lasiodiplodia theobromae 100% 63 MP 1321 Lasiodiodiplodia iraniensis 99% 54 MP 1319 Lasiodiplodia theobromae 100% 64 MP 1319 Lasiodiodiplodia iraniensis 99%

143

Appendix I: List of isolates used (numbers and descriptions) for each of the three experiments

A. List of putative pathogenic isolates used in the glasshouse pathogenicity studies

MP Code Isolate ID (nearest taxa) MP117 Lasiodiplodia iraniensis MP129 Lasiodiplodia iraniensis MP 115 Lasiodiplodia theobromae MP 036 Lasiodiplodia theobromae MP 290 Lasiodiplodia iraniensis MP 372 Lasiodiplodia pseudotheobromae MP 330 Lasiodiplodia theobromae MP 647 Lasiodiplodia theobromae MP 684 Lasiodiplodia iraniensis

MP 808 Lasiodiplodia iraniensis MP 582 Lasiodiplodia iraniensis MP 813 Lasiodiplodia pseudotheobromae MP 431 Lasiodiplodia theobromae MP 698 Lasiodiplodia iraniensis MP 107 Lasiodiplodia iraniensis MP 157 Neofusicoccum parvum MP 490 Neofusicoccum parvum MP 596 Cophinforma eucalypti MP 315 Lasiodiplodia hormozganensis MP 310 Lasiodiplodia pseudotheobromae MP 690 Cophinforma eucalypti

144

B. List of isolates used in the seedling pathogenicity trials

MP Code Isolate ID (nearest taxa) MP Code Isolate ID (nearest taxa) MP 206 Phomopsis sp. MP 206305 PhomopsisCladosporium sp cladosporioides MP 112 Lasiodiplodia theobromae MP 112307 LasiodiplodiaFusarium sp. theobromae

MP 088 Lasiodiplodia pseudotheobromae MP 088308 LasiodiplodiaPhomopsis sp. pseudotheobromae MP 020 Phomopsis sp. MP 020311 PhomopsisCladosporium sp. sp. MP 117 Lasiodiplodia iraniensis MP 117312 LasiodiplodiaDiaporthe phaseolorum iraniensis

MP 220 Lasiodiplodia theobromae MP 117220 LasiodiplodiaL. theobromae theobromae MP 011 Phomopsis sp. MP 213011 PhomopsisDiaporthe sp. sp. MP 078 Phomopsis sp. MP 048078 PhomopsisL. theobromae sp.

MP 046 Lasiodiplodia theobromae MP 107046 LasiodiplodiaL. theobromae theobromae MP 285 Lasiodiplodia theobromae MP 036285 LasiodiplodiaL. theobromae theobromae MP 107 Lasiodiplodia theobromae MP 107 Lasiodiplodia theobromae MP 115 L. theobromae MP 287 Uncultured MP 194287 UnculturedL. theobromae MP 288 Diaporthe phaseolorum MP 129288 DiaportheL. pseudotheobromae phaseolorum MP 289 Diaporthe phaseolorum MP 188289 DiaportheL. theobromae phaseolorum MP 290 Diaporthe phaseolorum MP 020290 DiaportheLasiodiplodia phaseolorum sp.

MP 291 Lasiodiplodia theobromae MP 290291 LasiodiplodiaB. mamane theobromae MP 292 Fusarium sp. MP 372292 FusariumL. iraniensis sp MP 293 Cladosporium cladosporioides MP 330293 CladosporiumL. hormozganensis cladosporioides

MP 294 Cladosporium cladosporioides MP 647294 CladosporiumL. iraniensis cladosporioides MP 684 L.pseudotheobromae MP 295 Phomopsis sp. MP 295 Phomopsis sp. MP 539 L. iraniensis MP 296 Alternaria alternata MP 296 Alternaria alternata MP 808 L. iraniensis MP 286 Fusarium sp. MP 286 Fusarium sp. MP 582 L. iraniensis MP 115 Lasiodiplodia theobromae MP 115 Lasiodiplodia theobromae MP 813 L. pseudotheobromae MP 188 Lasiodiplodia theobromae MP 188 Lasiodiplodia theobromae MP 431 L. pseudotheobromae MP 036 Lasiodiplodia theobromae MP 036 Lasiodiplodia theobromae MP 698039 LasiodiplodiaL. iraniensis theobromae MP 039 Lasiodiplodia theobromae MP 341284 LasiodiplodiaN. parvum theobromae MP 284 Lasiodiplodia theobromae MP 157298 LasiodiplodiaN. parvum theobromae MP 298 Lasiodiplodia theobromae MP 490218 PestalotiopsisL. iraniensis mangiferae MP 218 Pestalotiopsis mangiferae MP 596092 DiaportheL. iraniensis phaseolorum MP 092 Diaporthe phaseolorum MP 315299 DiaportheL. theobromae phaseolorum MP 299 Diaporthe phaseolorum MP 310300 DiaportheL. hormozganensis phaseolorum MP 300 Diaporthe phaseolorum MP 864 L. iraniensis MP 301 Fusarium sp. MP 301 Fusarium sp. MP 690 B. rhodina MP 302 Lasiodiplodia theobromae MP 302 Lasiodiplodia theobromae MP 361 Lasiodiplodia sp. MP 303 Phomopsis sp. MP 303 Phomopsis sp. MP 862 Lasiodiplodia sp. MP 304 Lasiodiplodia theobromae MP 304 Lasiodiplodia theobromae MP 610 Lasiodiplodia sp.

145

MP Code Isolate ID (nearest taxa) MP Code Isolate ID (nearest taxa) MP 483206 LasiodiplodiPhomopsisa sp.sp MP 573206 LasiodiplodiaPhomopsis sp.sp MP 723112 LasiodiplodiLasiodiplodiaa sp. theobromae MP 610112 LasiodiplodiaLasiodiplodia sp. theobromae MP 690088 LasiodiplodiaLasiodiplodia sp. pseudotheobromaeMP 586088 BotryoLasiodiplodia Group A pseudotheobromae MP 559020 LasiodiplodiaPhomopsis sp.sp. MP 723020 BotryoPhomopsis Group Asp. MP 612117 LasiodiplodiLasiodiplodiaa sp. iraniensis MP 571117 BotryoLasiodiplodia Group D iraniensis MP 378220 LasiodiplodiaLasiodiplodia sp. theobromae MP 569220 BotryoLasiodiplodia Group D theobromae MP 760011 LasiodiplodiaPhomopsis sp.sp. MP 665011 BotryoPhomopsis Group Dsp. MP 480078 B.Phomopsis mamane sp. MP 596078 BotryoPhomopsis Group Asp. MP 442046 B.mamaneLasiodiplodia theobromae MP 610046 BotryoLasiodiplodia Group A theobromae MP 342285 LasiodiplodiaLasiodiplodia sp. theobromae MP 585285 UnidentifiedLasiodiplodia theobromae MP 597107 LasiodiplodiaLasiodiplodia sp. theobromae MP 593107 BotryoLasiodiplodia Group A theobromae MP 577287 LasiodiplodiaUncultured sp. MP 608287 BotryoUncultured Group A MP 373288 LasiodiplodiaDiaporthe phaseolorum sp. MP 371288 BotryoDiaporthe Group phaseolorum B MP 371289 LasiodiplodiaDiaporthe phaseolorum sp. MP 483289 BotryoDiaporthe Group phaseolorum D MP 585290 LasiodiplodiaDiaporthe phaseolorum sp. MP 669290 BotryoDiaporthe Group phaseolorum A MP 608291 LasiodiplodiaLasiodiplodia sp. theobromae MP 611291 BotryoLasiodiplodia Group A theobromae MP 306292 ColletotrichumFusarium sp sp. MP 724292 BotryoFusarium Group sp B MP 573293 LasiodiplodiaCladosporium sp. cladosporioidesMP 761293 BotryoCladosporium Group A cladosporioides MP 559294 LasiodiplodiaCladosporium sp. cladosporioidesMP 767294 nonCladosporium sporulating Phomopsis cladosporioides like MP 571295 LasiodiplodiaPhomopsis sp. MP 696295 BotryoPhomopsis Group sp.A MP 689296 L.Alternaria pseudotheobromae alternata MP 373296 BotryoAlternaria Group alternata B MP 583286 LasiodiplodiaFusarium sp. sp. MP 573286 BotryoFusarium Group sp. A MP 442 Lasiodiplodia sp. MP 559 Botryo Group A MP 115 Lasiodiplodia theobromae MP 115 Lasiodiplodia theobromae MP 575 Lasiodiplodia sp. MP 681 Botryo Group A MP 188 Lasiodiplodia theobromae MP 188 Lasiodiplodia theobromae MP 577 Lasiodiplodia sp. MP 575 Botryo Group A MP 036 Lasiodiplodia theobromae MP 036 Lasiodiplodia theobromae MP 611 Lasiodiplodia sp. MP 576 Botryo Group A MP 039 Lasiodiplodia theobromae MP 039 Lasiodiplodia theobromae MP 483 Lasiodiplodia sp. MP 597 Botryo Group A MP 284 Lasiodiplodia theobromae MP 284 Lasiodiplodia theobromae MP 669 Lasiodiplodia sp. MP 577 Botryo Group A MP 298 Lasiodiplodia theobromae MP 298 Lasiodiplodia theobromae MP 597 Lasiodiplodia sp. MP 307 Colletotrichum sp. MP 218 Pestalotiopsis mangiferae MP 218 Pestalotiopsis mangiferae MP 576 Lasiodiplodia sp. MP 306 Colletotrichum sp. MP 092 Diaporthe phaseolorum MP 092 Diaporthe phaseolorum MP 597 Lasiodiplodia sp. MP 241A Pestalotiopsis sp. MP 299 Diaporthe phaseolorum MP 299 Diaporthe phaseolorum MP 787 Lasiodiplodia sp. MP 107A Botryo Group B MP 300 Diaporthe phaseolorum MP 300 Diaporthe phaseolorum MP 611 Lasiodiplodia sp. MP 013 Phomopsis sp. MP 301 Fusarium sp. MP 301 Fusarium sp. MP 580 Lasiodiplodia sp. MP 016 Phomopsis sp. MP 302 Lasiodiplodia theobromae MP 302 Lasiodiplodia theobromae MP 442 B.mamane MP 274 Pestalotiopsis sp. MP 303 Phomopsis sp. MP 303 Phomopsis sp. MP 770 Lasiodiplodia sp. MP 143 Botryo Group F MP 304 Lasiodiplodia theobromae MP 696 Lasiodiplodia sp. MP 278304 PhomopsisLasiodiplodia sp. theobromae

146

MP Code Isolate ID (nearest taxa) MP Code Isolate ID (nearest taxa)

MP 300206 PestalotiopsisPhomopsis spsp. MP 205206 BotryoPhomopsis Group Esp MP 242112 PestalotiopsisLasiodiplodia sp. theobromae MP 849112 BotryoLasiodiplodia Group J theobromae MP 169088 BotryoLasiodiplodia Group G pseudotheobromaeMP 116088 BotryoLasiodiplodia Group B pseudotheobromae MP 078020 BotryoPhomopsis Group Fsp. MP 190B020 BotryoPhomopsis Group Bsp. MP 291117 PhomopsisLasiodiplodia sp. iraniensis MP 849117 BotryoLasiodiplodia Group B iraniensis MP 301220 UnidentifiedLasiodiplodia theobromae MP 060220 BotryoLasiodiplodia Group D theobromae MP 044011 BotryoPhomopsis Group Hsp. MP 357011 BotryoPhomopsis Group Dsp. MP 091078 BotryoPhomopsis Group Hsp. MP 157078 PhomopsisPhomopsis sp. MP 088046 BotryoLasiodiplodia Group B theobromae MP 050046 BotryoLasiodiplodia Group J theobromae MP 180285 PhomopsisLasiodiplodia sp. theobromae MP 264285 PestalotiopsisLasiodiplodia sp. theobromae MP 015107 PhomopsisLasiodiplodia sp. theobromae MP 174107 PestalotiopsisLasiodiplodia sp. theobromae MP 201287 PhomopsisUncultured sp. MP 234287 PestalotiopsisUncultured sp. MP 068288 BotryoDiaporthe Group phaseolorum C MP 118288 PestalotiopsisDiaporthe phaseolorum sp. MP 026289 BotryoDiaporthe Group phaseolorum F MP 240289 PestalotiopsisDiaporthe phaseolorum sp. MP 005290 BotryoDiaporthe Group phaseolorum E MP 216290 PestalotiopsisDiaporthe phaseolorum sp. MP 170291 BotryoLasiodiplodia Group G theobromae MP 034291 PhomopsisLasiodiplodia sp. theobromae MP 203292 BotryoFusarium Group sp C MP 042292 BotryoFusarium Group sp A MP 221 Botryo Group E MP 023 Phomopsis sp. MP 293 Cladosporium cladosporioidesMP 293 Cladosporium cladosporioides MP 359 Botryo Group A MP 052 Botryo Group J MP 294 Cladosporium cladosporioidesMP 294 Cladosporium cladosporioides MP 187 Botryo Group A MP 043 Botryo Group A MP 295 Phomopsis sp. MP 295 Phomopsis sp. MP 061 Botryo Group A MP 131 Pestalotiopsis sp. MP 296 Alternaria alternata MP 296 Alternaria alternata MP 211 Botryo Group A MP 286 Fusarium sp. MP 286 Fusarium sp. MP 183 Botryo Group F MP 115 Lasiodiplodia theobromae MP 115 Lasiodiplodia theobromae MP 154 Botryo Group F MP 188 Lasiodiplodia theobromae MP 188 Lasiodiplodia theobromae MP 062 Botryo Group F MP 036 Lasiodiplodia theobromae MP 036 Lasiodiplodia theobromae MP 063 Botryo group C MP 039 Lasiodiplodia theobromae MP 039 Lasiodiplodia theobromae MP 356 Botryo Group H MP 284 Lasiodiplodia theobromae MP 196284 BotryoLasiodiplodia Group H theobromae MP 298 Lasiodiplodia theobromae MP 120298 BotryoLasiodiplodia Group B theobromae MP 218 Pestalotiopsis mangiferae MP 181218 PhomopsisPestalotiopsis sp. mangiferae MP 092 Diaporthe phaseolorum MP 059092 PhomopsisDiaporthe sp. phaseolorum MP 299 Diaporthe phaseolorum MP 344299 PhomopsisDiaporthe sp. phaseolorum MP 300 Diaporthe phaseolorum MP 195300 BotryoDiaporthe Group phaseolorum C MP 301 Fusarium sp. MP 204301 BotryoFusarium Group sp. F MP 302 Lasiodiplodia theobromae MP 190302 BotryoLasiodiplodia Group E theobromae MP 303 Phomopsis sp. MP 121303 BotryoPhomopsis Group Gsp. MP 047304 BotryoLasiodiplodia G theobromae MP 304 Lasiodiplodia theobromae

147