<%BANNER%>

Using Molecular Analysis to Investigate Phylogenetic Relationships in Two Tropical Pathosystems

Permanent Link: http://ufdc.ufl.edu/UFE0041003/00001

Material Information

Title: Using Molecular Analysis to Investigate Phylogenetic Relationships in Two Tropical Pathosystems Witches' Broom of Cacao, Caused by Moniliophthora Perniciosa, and Mango Anthracnose, Caused by Colletotrichum Spp.
Physical Description: 1 online resource (236 p.)
Language: english
Creator: Tarnowski, Tara
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anthracnose, cacao, colletotrichum, mangifera, mango, moniliophthora, theobroma, witches
Plant Pathology -- Dissertations, Academic -- UF
Genre: Plant Pathology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: USING MOLECULAR ANALYSIS TO INVESTIGATE PHYLOGENETIC RELATIONSHIPS IN TWO TROPICAL PATHOSYSTEMS: WITCHES? BROOM OF CACAO, CAUSED BY MONILIOPHTHORA PERNICIOSA, AND MANGO ANTHRACNOSE, CAUSED BY COLLETOTRICHUM SPP. By Tara Luana Barrett Tarnowski December 2009 Chair: Randy C. Ploetz Cochair: Jeffrey R. Rollins Major: Plant Pathology The increasing availability of DNA sequence data has enabled rapid advances in molecular systematics. This is especially true for the Fungi, where systematics and taxonomy relied previously on largely artificial, morphologically based systems. Many of the advances in fungal systematics have been made with phytopathogens. Characterizing the relationships of lineages among phytopathogenic genera provides insight into not only basic systematic information, but can also address applied questions on host specificity, life strategy and pathogenicity. With DNA sequence-based phylogenetic analyses, two broad topics were investigated: 1) relationships among biotypes of Moniliophthora perniciosa, causal agent of witches? broom of cacao, and 2) host and tissue specificity of Colletotrichum gloeosporioides sensu lato on mango. Sequences from the ITS1-5.8S-ITS2 ribosomal DNA region (ITS), intergenic spacer (IGS) region, and RNA polymerase large subunit (RPB1) regions were analyzed for 36 accessions of M. perniciosa from all reported biotypes of the pathogen. Maximum parsimony and maximum likelihood analyses resolved three major clades within the species: a clade that contained all isolates from Theobroma spp. and most isolates from Solanum spp.; a clade that contained most isolates from malpigheaceous hosts; and a clade that contained three isolates from S. cernum, a bignoniaceous liana, and an unknown host. Analysis of morphological characters did not reveal striking differences among the clades. The molecular findings indicate that M. perniciosa likely evolved from a saprophytic ancestor and that pathogenicity may have evolved with a switch from a heterothallic to homothallic lifestyle. Host jumps have resulted in distinct lineages within the pathogen. Sequences from the ITS, mating type (MAT) 2 gene, and a cloned region from a randomly amplified DNA fragment were analyzed to examine 58 accessions of C. gloeosporioides sensu lato that represented all anthracnose-affected organs of mango, as well as avocado, banana, carambola and guava. Phylogenies from maximum parsimony and maximum likelihood analyses revealed a mango-specific clade that comprised all blossom blight and leaf anthracnose agents, and some fruit anthracnose agents. Other mango fruit anthracnose and peduncle isolates resolved in two general clades that also contained isolates from other fruit hosts (avocado, guava, carambola, banana). The pathogenicity of representive isolates supported the phylogenetic findings, in that the mango-specific clade isolates cause blossom blight, leaf anthracnose, and fruit anthracnose, and the general clade isolates caused only fruit anthracnose. There were no differences among clades with respect to conidium size and shape, and hyphopodium size. Hyphopodia produced by isolates from the mango-specific clade were clavate and smooth, while isolates from other clades produced irregular, lobed hyphopodia.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Tara Tarnowski.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Ploetz, Randy C.
Local: Co-adviser: Rollins, Jeffrey A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041003:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041003/00001

Material Information

Title: Using Molecular Analysis to Investigate Phylogenetic Relationships in Two Tropical Pathosystems Witches' Broom of Cacao, Caused by Moniliophthora Perniciosa, and Mango Anthracnose, Caused by Colletotrichum Spp.
Physical Description: 1 online resource (236 p.)
Language: english
Creator: Tarnowski, Tara
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: anthracnose, cacao, colletotrichum, mangifera, mango, moniliophthora, theobroma, witches
Plant Pathology -- Dissertations, Academic -- UF
Genre: Plant Pathology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: USING MOLECULAR ANALYSIS TO INVESTIGATE PHYLOGENETIC RELATIONSHIPS IN TWO TROPICAL PATHOSYSTEMS: WITCHES? BROOM OF CACAO, CAUSED BY MONILIOPHTHORA PERNICIOSA, AND MANGO ANTHRACNOSE, CAUSED BY COLLETOTRICHUM SPP. By Tara Luana Barrett Tarnowski December 2009 Chair: Randy C. Ploetz Cochair: Jeffrey R. Rollins Major: Plant Pathology The increasing availability of DNA sequence data has enabled rapid advances in molecular systematics. This is especially true for the Fungi, where systematics and taxonomy relied previously on largely artificial, morphologically based systems. Many of the advances in fungal systematics have been made with phytopathogens. Characterizing the relationships of lineages among phytopathogenic genera provides insight into not only basic systematic information, but can also address applied questions on host specificity, life strategy and pathogenicity. With DNA sequence-based phylogenetic analyses, two broad topics were investigated: 1) relationships among biotypes of Moniliophthora perniciosa, causal agent of witches? broom of cacao, and 2) host and tissue specificity of Colletotrichum gloeosporioides sensu lato on mango. Sequences from the ITS1-5.8S-ITS2 ribosomal DNA region (ITS), intergenic spacer (IGS) region, and RNA polymerase large subunit (RPB1) regions were analyzed for 36 accessions of M. perniciosa from all reported biotypes of the pathogen. Maximum parsimony and maximum likelihood analyses resolved three major clades within the species: a clade that contained all isolates from Theobroma spp. and most isolates from Solanum spp.; a clade that contained most isolates from malpigheaceous hosts; and a clade that contained three isolates from S. cernum, a bignoniaceous liana, and an unknown host. Analysis of morphological characters did not reveal striking differences among the clades. The molecular findings indicate that M. perniciosa likely evolved from a saprophytic ancestor and that pathogenicity may have evolved with a switch from a heterothallic to homothallic lifestyle. Host jumps have resulted in distinct lineages within the pathogen. Sequences from the ITS, mating type (MAT) 2 gene, and a cloned region from a randomly amplified DNA fragment were analyzed to examine 58 accessions of C. gloeosporioides sensu lato that represented all anthracnose-affected organs of mango, as well as avocado, banana, carambola and guava. Phylogenies from maximum parsimony and maximum likelihood analyses revealed a mango-specific clade that comprised all blossom blight and leaf anthracnose agents, and some fruit anthracnose agents. Other mango fruit anthracnose and peduncle isolates resolved in two general clades that also contained isolates from other fruit hosts (avocado, guava, carambola, banana). The pathogenicity of representive isolates supported the phylogenetic findings, in that the mango-specific clade isolates cause blossom blight, leaf anthracnose, and fruit anthracnose, and the general clade isolates caused only fruit anthracnose. There were no differences among clades with respect to conidium size and shape, and hyphopodium size. Hyphopodia produced by isolates from the mango-specific clade were clavate and smooth, while isolates from other clades produced irregular, lobed hyphopodia.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Tara Tarnowski.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Ploetz, Randy C.
Local: Co-adviser: Rollins, Jeffrey A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0041003:00001


This item has the following downloads:


Full Text

PAGE 1

USING MOLECULAR ANALYSIS TO INVES T IGATE PHYLOGENETIC RELATIONSHIPS IN TWO TROPICAL PATHOSYSTEMS: WITC HES BROOM OF CACAO, CAUSED BY AND MANGO ANTHRACNOSE, CAUSED BY By TARA LUANA BARRETT TARNOWSKI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Tara Luana Barrett Tarnowski 2

PAGE 3

To David, for the endless love and support 3

PAGE 4

ACKNOWLEDGMENTS There are so m any people that contribute suppor t to a student during dissertation research, and I want to thank everyone who pitched in. I would especially like to thank my committee members Dr. Aime, Dr. Palmateer, Dr. Ploetz, Dr. Rollins, and Dr. Soltis for their instruction, advice, and support. I also apprecia te all the help I received from Jos Prez, Patricia Lopez, and Gail Harris. For the financial support I recei ved for research, I thank IF AS, the Florida Mango Forum, and the Redlands Citizen Associ ation. I also received assistan ce for travel to professional meetings from the Graduate School, the Depart ment of Plant Pathology, and the Mycological Society of America. Many people also contributed material suppor t for my research. Drs. Harry Evans and Robert Barreto donated isolates for phylogenetic analysis. Limeco, LLC and Brooks Tropicals, Inc. donated fr uit for pathogenicity experiments. Lastly I want to thank all th e friends and family for the love and support they gave me. I could not have completed all of this without th em. I want to especia lly thank David, Trevor, Mom and Dad, for always believing in me and giving me strength when I need it the most, and for always reminding me what is important in life. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.......................................................................................................................10 CHAPTER 1 INTRODUCTION................................................................................................................. .14 Current State of Fungal Systematics.......................................................................................14 Classification of Different Phyla of Fungi......................................................................15 Species Characterization in Fungi...................................................................................16 Species defined by morphology...............................................................................16 Biological species concept.......................................................................................19 Phylogenetic species concept...................................................................................21 Evolutionary species concept...................................................................................24 The Role of Plant Pathology in Fungal Systematics.......................................................25 Witches Broo m of Cacao......................................................................................................29 -induced Anthracnoses: Pla gues to Tropical Fruit.........................................36 Taxonomic History of ............................................................................36 as the Cause of Mango Anthracnose ..........................40 2 INVESTIGATING THE RELATI ONSHIPS AMONG BIOTYPES OF USING PHYLOGENIES FROM MULTIPLE DNA REGI ONS.................................................................................................................... .50 Introduction................................................................................................................... ..........50 Materials and Methods...........................................................................................................56 Isolate Selection...............................................................................................................56 DNA Extraction and PCR of ITS, IGS, and RPB1 Regions...........................................56 Phylogenetic Analysis.....................................................................................................58 Congruence of Molecular Data Sets and Combined Analysis........................................59 Morphological Description..............................................................................................61 Results.....................................................................................................................................62 Phylogenetic Analysis.....................................................................................................62 Congruence Tests and Combined Data Analysis............................................................65 Morphological Characterization......................................................................................66 Discussion...............................................................................................................................67 3 USE OF PHYLOGENIES FROM MULTI PLE DNA REGIONS TO ASSESS THE DIVERGENCE OF LINEAGES OF SENSU LATO THAT ARE ASSOCIATED WITH M ANGO ANTHRACNOSE IN SOUTH FLORIDA.................................................................................................................9 6 5

PAGE 6

Introduction................................................................................................................... ..........96 Materials and Methods...........................................................................................................99 Collection and Storage of Isolates...................................................................................99 DNA Extraction and PC R of the ITS, MAT1-2, and CGTT5 Regions.........................100 Phylogenetic Analyses...................................................................................................102 Congruence of Molecular Data Sets and Combined Analysis......................................103 Morphology Studies......................................................................................................104 Pathogenicity Studies....................................................................................................104 Blossom blight........................................................................................................105 Leaf anthracnose....................................................................................................105 Fruit pathogenicity.................................................................................................106 Results...................................................................................................................................107 Phylogenetic Analyses...................................................................................................107 Congruence ests and ombined nalysis..................................................................109 Morphology Studies......................................................................................................110 Pathogenicity studies.....................................................................................................111 Discussion.............................................................................................................................112 4 INVESTIGATING THE ROLE PLAYED BY IN THE DEVELOPMENT OF BLOSSOM BLIGHT AND LEAF AND FRUIT ANTHRACNOSE OF MANGO IN SOUTHERN FLORIDA............................................153 Introduction................................................................................................................... ........153 Materials and Methods........................................................................................................154 Leaf Anthracnose Survey..............................................................................................154 Pathogenicity Tests........................................................................................................157 Blossom blight tests...............................................................................................157 Leaf pathogenicity..................................................................................................158 Fruit pathogenicity.................................................................................................158 Results...................................................................................................................................159 Survey Results...............................................................................................................15 9 Pathogenicity of and on Mango Organs....................160 Blossom blight........................................................................................................160 Leaf anthracnose....................................................................................................160 Fruit anthracnose....................................................................................................161 Discussion.............................................................................................................................161 5 CONCLUSIONS.................................................................................................................. 175 APPENDIX A IDENTIFICATION AND CHARACTER IZATION OF FIVE RANDOMLY AMPLIFIED DNA REGIONS FLANKING MICROSATELLITE REPEATS FOR USE IN PHYLOGENETIC ANALYSIS OF ..........................................................................................................183 Introduction................................................................................................................... ........183 6

PAGE 7

Materials and Methods.........................................................................................................184 Results and Discussion......................................................................................................... 185 B FIRST REPORT OF CAUSING POSTHARVEST ANTHRACNOSE ON PAPAYA IN SOUTH FLORIDA..................191 C FIRST REPORT OF AND A SP. CAUSING POSTHARVEST ANTHRACNOSE ON PASSIONFRUIT ( SPP ) IN FLORIDA..................202 LIST OF REFERENCES.............................................................................................................208 BIOGRAPHICAL SKETCH.......................................................................................................236 7

PAGE 8

LIST OF TABLES Table page 1-1 von Arx classification of species of (1957), including anamorph, teleomorph (when described), numbers of synonyms von Arx placed in each taxon, and host plant(s).............................................................................................................. ...46 1-2 Major revisions of by von Arx (1957), Sutton (1980), Sutton (1992), and others..................................................................................................................... ......47 2-1 Description of isolat es used in the study............................................................................74 2-2 Maximum parsimony statistics for th e ITS, IGS, and RPB1 datasets...............................77 2-3 Statistics from phylogenetic analyses................................................................................78 2-4 Topological congruence tests for single region data sets...................................................79 2-5 Macroscopic morphological charac ters of select accessions of ...........................................................................................................................81 2-6 Cystidia, basidia, and basidiospore characters of select accessions of isolates...................................................................................83 2-7 Dimensions of basidiospo res of select accessions of ............88 3-1 Number of trees, samples and single-spor e isolates collected in the hierachical sampling scheme from various mango organs and tropical fruit hosts............................124 3-2 Accessions included in study...........................................................................................125 3-3 Details of phylogenetic analyses......................................................................................128 3-4 Number of isolates by tissu e type and phylogenetic clade..............................................129 3-5 Topography-based congruence analyses..........................................................................130 3-6 Morphological description of conidia and hyphopodia of 13 isolates.............................132 3-7 Statistical comparison of conidia and hyphopodia dimensions.......................................134 3-8 Area under the disease pr ogress curve (AUDPC) and ymax values for blossom blight experiments, 2009............................................................................................................135 3-9 Area under the disease pr ogress curve (AUDPC) and ymax values for leaf anthracnose experiments, 2009............................................................................................................136 8

PAGE 9

3-10 Anthracnose area under the disease progress curve (AUDPC) and lesion diam eters on detached Tommy Atkinsfruit...................................................................................137 3-11 Area under the disease pr ogress curve (AUDPC) and ymax values for attached fruit anthracnose experiments, 2009........................................................................................138 4-1 ITS sequences included in phylogenetic analysis............................................................165 4-2 Isolate, species and mango origin for isolat es that were used in pathogenicity tests......167 4-3 Results for leaf anthracnose surveys................................................................................168 4-4 Areas under the disease pr ogress curve (AUDPC) and ymax values for blossom blight experiments, 2008-2009...................................................................................................169 4-5 Areas under the disease pr ogress curves (AUDPC) and ymax values for leaf anthracnose experiments, 2008-2009...............................................................................170 4-6 Areas under the disease pr ogress curve (AUDPC) and ymax (lesion diameter) values for detached fruit experiments, 2008-2009......................................................................171 A-1 Origin of, and statistics for, CGTT loci...........................................................................187 A-2 PCR primers, product size, homologies, and the extent of polymorphism of CGTT regions..............................................................................................................................188 B-1 Number of lesions from which species we re recovered from fruit from two locations..194 B-2 ITS sequences included in phylogenetic analysis............................................................195 B-3 Lesion diameters from pa thogenicity experiments..........................................................197 C-1 Incidence of symptoms fro m two inoculation studies.....................................................205 9

PAGE 10

LIST OF FIGURES Figure page 1-1 Proposed relationships among fungal phyla as reported in the AFTOL Deep Hypha issue of (2006, v. 89, issue 6)..........................................................................49 2-1 Phylogenetic relationshi ps among 36 accessions of based on ITS sequence data.........................................................................................................89 2-2 Phylogenetic relationships among accessions of resolved using ML analysis of ITS sequence data...........................................................................90 2-3 Phylogenetic relationships among 36 accessions of based on IGS sequence data.........................................................................................................91 2-4 Phylogenetic relationships among accessions of resolved using ML analysis of IGS sequence data...........................................................................92 2-5 Phylogenetic relationshi ps among 35 accessions of based on RPB1 sequence data......................................................................................................93 2-6 Phylogenetic relationships among accessions of resolved using ML analysis of RPB1 sequence data........................................................................94 2-7 Microscopic characteristics of accessions of .........................95 3-1 Phylogenetic relationshi ps among 65 accessions of based on ITS sequence data.............................................................................................139 3-2 Phylogenetic relationships among accessions of using ML analysis of ITS sequence data.........................................................................140 3-3 Phylogenetic relationshi ps among 58 accessions of based on MAT1-2 sequence data.....................................................................................1 3-4 Phylogenetic relationships among accessions of using ML analysis of MAT1-2 sequence data.................................................................142 3-5 Phylogenetic relationshi ps among 58 accessions of based on CGTT5 sequence data.......................................................................................143 3-6 Phylogenetic relationships among accessions of using ML analysis of CGTT5 sequence data...................................................................144 3-7 Phylogenetic relationshi ps among 54 accessions of based on combined MAT1-2+CGTT5 sequence data set................................................145 10

PAGE 11

3-8 Phylogenetic relationships among accessions of using ML analysis of MAT1-2+CGTT5 data set.............................................................146 3-9 Micrographs of conidia of indicated isolates of ...........147 3-10 Micrographs of hyphopodia of indicated isolates of ....148 3-11 Blossom blight 28 days after inoculation.........................................................................149 3-12 Leaf anthracnose 10 days after inoculation.....................................................................150 3-13 Lesion development on artifi cially inoculated fruit.........................................................151 3-14 Graphical representation of the numbe r of polymorphisms (including nucleotide and indel sites) among four taxa within sensu lato and three distantly related species..........................................................................................152 4-1 Conidia of A) were slightly fusiform, whereas those of B) had cylindrical tips............................................................................172 4-2 Anthracnose lesions collected during the grove survey (A-D) and lesions that were produced during pathogenicity tests (E-F).......................................................................173 4-3 Phylogeny for 44 accessions of based on ITS sequence data.................1 A-1 Amplification of A) CGTT-1, B) CGTT3, C) CGTT-4, D) CGTT-5, and E) CGTT-6 loci with primers desc ribed in Table A-1........................................................................189 A-2 Amplification of A) CGTT-3, B) CG TT-4, C) CGTT-5, and D) CGTT-6 loci in isolates...................................................................................................190 B-1 Lesions associated with began as A) small brown and slightly sunken areas that became B) increasingly s unken and eventually covered with dark sporulation........................................................................................................................198 B-2 Morphological characteristics of recovered from papaya:.........199 B-3 One of 2,255 most parsimonious ITSbased trees (101 steps, CI=0.796, RI=0.967) for taxa recovered from papaya and other hosts..............................................................200 B-4 Lesion produced after artificial inoculation with A) and B) ................................................................................................................201 C-1 Anthracnose lesion development on inoculated yellow passionfruit.............................. C-2 Morphology of sp......................................................................................... 11

PAGE 12

Abstract of Dissertation Pres ented to the Graduate School of the University of Flor ida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy USING MOLECULAR ANALYSIS TO INVEST IGATE PHYLOGENETIC RELATIONSHIPS IN TWO TROPICAL PATHOSYSTEMS: WITC HES BROOM OF CACAO, CAUSED BY AND MANGO ANTHRACNOSE, CAUSED BY SPP By Tara Luana Barrett Tarnowski December 2009 Chair: Randy C. Ploetz Cochair: Jeffrey R. Rollins Major: Plant Pathology The increasing availability of DNA sequence data has enabled rapid advances in molecular systematics. This is especially true for the Fungi, where systematics and taxonomy relied previously on largely artificial, morphologically based systems. Many of the advances in fungal systematics have been made with phytopathogens Characterizing the relationships of lineages among phytopathogenic genera provid es insight into not only basi c systematic information, but can also address applied questions on host spec ificity, life strategy and pathogenicity. With DNA sequence-based phylogenetic analyses two broad topics were inve stigated: 1) relationships among biotypes of causal agent of witches broom of cacao, and 2) host and tissue specificity of sensu lato on mango. Sequences from the ITS1-5.8S-ITS2 ribosomal DNA region (ITS), intergenic spacer (IGS) region, and RNA polymerase large subunit (RPB1) regions were analyzed for 36 accessions of from all reported biotypes of the pathogen. Maximum parsimony and maximum likelihood analyses resolved three major clades within the species: a clade that contained all isolates from spp. and most isolates from spp.; a clade that contained most 12

PAGE 13

13 isolates from malpigheaceous hosts; and a clade that contained three isolates from a bignoniaceous liana, and an unknown host. Analysis of morphological charac ters did not reveal striking differences among the clades. Th e molecular findings indicate that likely evolved from a saprophytic ancestor and that pat hogenicity may have evol ved with a switch from a heterothallic to homothallic lifestyle. Host jump s have resulted in distinct lineages within the pathogen. Sequences from the ITS, mating type (MAT) 2 gene, and a cloned region from a randomly amplified DNA fragment were analyzed to examine 58 accessions of sensu lato that represented all anthracnose-affected organs of mango, as well as avocado, banana, carambola and guava. Phylogenies from maximu m parsimony and maximum likelihood analyses revealed a mango-specific clade that comprised all blossom blight and leaf anthracnose agents, and some fruit anthracnose agents. Other mango fr uit anthracnose and pedunc le isolates resolved in two general clades that also contained isolates from other fruit hosts (avocado, guava, carambola, banana). The pathogenicity of re presentive isolates s upported the phylogenetic findings, in that the mango-specifi c clade isolates cause blossom blight, leaf anthracnose, and fruit anthracnose, and the general clade isolates caused only fruit anthracnose. There were no differences among clades with respect to co nidium size and shape, and hyphopodium size. Hyphopodia produced by isolates from the mango-sp ecific clade were clavate and smooth, while isolates from other clades pr oduced irregular, lobed hyphopodia.

PAGE 14

CHAPTER 1 INTRODUCTION The Kingdom Fungi includes at least 74,000 describe d species, with estimates of the total number of species from one to nine million (Hawksworth 1991, 2001). Fungi possess diverse life strategies and inhabit diverse habitats (Hawksworth 2001). One of the major groups of fungi is the plant pathogens, which cause the majority of plant diseases and reduce yields of food crops worldwide by an estimated 10% (Oerke and Dehne 2004). A major challenge is the correct identification and characteriza tion of the species that are responsible for these diseases. Although fungal taxa should be classified on both phylogenetic and biological bases, morphological criteria have played a more si gnificant role historically. In the last 20 years, multilocus DNA se quence-based phylogenies have been used to build a tree of life for the Fungi that would not have been possible with morphological data alone (Lutzoni et al. 2004, James et al. 2006a). The group of scientists involved in the Assembli ng the Fungal Tree of Life program (AFTOL, http://aftol.org/ ) produced a kingdom-wide phylogeny based on ribosomal DNA data (Lutzoni et al. 2004), and mo re recently published the Deep Hypha issue of (2006, v. 89, issue 6), which contains the most comprehensive analyses to date of all major groups of Fungi. Especially important were the studies of the Chytridiomycota and Zygomycota, which elucidated the relationships of taxa within these phyla and addressed their phylogenetic placement in the Kingdom Fungi (O Donnell et al. 2001, Tanabe et al. 2004, James et al. 2006a,b, White et al. 2006). With gene sequence data, mycologists can now decipher relationships at every level of fungal systema tics, and coordinated projects like AFTOL are paving the way towards a less artifi cial fungal classification system that is based on phylogenetic relationships. 14

PAGE 15

Recent molecular k ingdom-wide studies suppor ted the recognition of two of the four traditional fungal phyla, the Basidiomycota and As comycota (Lutzoni et al. 2004, James et al. 2006a). The Chytridiomycota and Zygomycota are polyphyletic and resolved into several distinct, early-divergi ng lineages within the F ungi (ODonnell et al. 2001, Tanabe et al. 2004, White et al. 2006, James et al. 2006a). Tw o new phyla have been recognized, the Blastocladiomycota (formerly classified as the Bl astocladiales in the Chyt ridiomycota) (James et al. 2006b) and the Glomeromycota (the arbuscu lar mycorrhizal fungi, formerly in the Zygomycota) (Schssler et al. 2001 ). The AFTOL project has helped clarify relationships within each phylum at the subphylal, ordinal, or fam ilial level. The relationship among fungal phyla according to phylogenies presented in the D eep Hypha issue is shown in Figure 1-1. James et al. (2006b) described a core Chytridiomycota clade containing the majority of previously defined chytrid taxa with the following exceptions : the Blastocladiomycota; two species, which appeared to represent the oldest known lin eage in the true Fungi; and which was nested in a zygomycete lineag e. The paraphyly of the Zygomycota was detailed by White et al. (2006). Nine zygomycete orders formed at least thr ee paraphyletic lineages between the Chytridiomycota and the Glomeromycota+Ascomycota+ Basidiomycota clade. Although the Glomeromycota is hypothesized to be sister to the Ascomycota+Basidiomycota clade, protein coding-data do not always su pport this hypothesis (Redecker and Raab, 2006). The sister phyla Ascomycota and Basidiomycota are the most recently evolved lineages in the Fungi (Lutzoni et al. 2004, James et al. 2006a) and have a larger number of recognized taxa and a larger database of DNA sequences than th e other phyla. Within the Ascomycota, there are three traditionally recognized subphyla: the Taphrinomycotina, the Saccharomycotina, and the 15

PAGE 16

Pezizomycotina. The relationships among Taphrinom ycotina taxa were only partially resolved by Sugiyama et al. (2006), who suggested that addition al taxon sampling is needed for further resolution. The Pezizomycotina analys is utilized one of the most comprehensive data sets, which included five loci and 191 taxa (Spatafora et al. 2006). The study s upported the monophyly of the Pezizomycotina and the Sacch aromycotina, and resolved mo st of the classes in the Pezizomycotina. As had been suggested in prio r molecular phylogenies (Lutzoni et al. 2004), apothecia were indicated as the most ancient ascomal state. The Basidiomycota are also classified in three subphyla: the Agaricomycotina, Ustilagomycotina, and Pucciniomycotina. Recent phylogenetic analyses (Aime et al. 2006, Hibbett 2006, Matheny et al. 2007 ) suggest that the Pucciniom ycotina are basal to the Ustilaginomycotina and Agaricomycotina. This re lationship is supported by the simple septal pore structure and cell wall sugars present in the Pucciniomycotina compared to the latter subphyla (Lutzoni et al. 2004, Prillinger et al. 2002). Although much progress has been made in resolving deeper nodes in the tree, the leaves of the Fungal Tree of Life are still largely nonres olved (relationships below the familial level). This is where plant pathology has contributed most to fungal systematics. Molecular data have resolved relationships for groups of important pathogens, such as several species complexes (ODonnell et al. 2000, ODonnell et al. 2004) and (de Meyer et al. 2008, Roets et al. 2009). The extent to which these and other data may c ontribute to a species concept for the Fungi is discussed below. Fungal classification has traditionally been based mainly on morphology. These classifications often did not re flect evolutionary relationships among fungal taxa. There are 16

PAGE 17

significant problems when mo rphologi cal data are used in phylogenetic analyses of fungi, most importantly the scarcity of homologous characters that can be used across taxa. For example, yeasts in the order Saccharomycetales are quite divergent based on sequence data, despite having almost no morphological diffe rences (Suh et al. 2006). Morphological classifications a bove the ordinal level have often conflicted due to the use of different sets of morphologica l characters and the scarcity of characters that are present among all taxa. Thus, relationships among groups are gene rally not resolved above the ordinal level, especially in the Ascomycota (Alexopolous et al. 1996). At the genus or species level, morphological characters used for classification often lack consis tency. Due to the indeterminate nature of fungal growth and development, charact ers such as the shapes and sizes of spores, reproductive structures, and producti on of sterile hyphae often differ with a specimens age or under different environmental conditions (E kpo 1978, Sutton 1992, Andrews 1995, Spiers et al. 2000). An additional challenge for fungal systematic s has been the placement of asexual taxa. Many fungi have asexual (anamorph) and se xual (teleomorph) stages with dissimilar morphologies. The respective stages have often been classified separately, leading to an overestimation of species. In contrast, only the anamorph is known for other fungi. In the past, anamorphic fungi formed a separate taxon calle d Fungi Imperfecti or Deuteromycota, an artificial phylum that had no hypothesized relationship with other fungal groups (Gams 1995). DNA sequence data have established relationships between asexual and sexual taxa, and have enabled phylogenetic classification of difficult taxa and new understanding of fungal diversity and evolution (Taylor 1995). However, despite these clarific ations anamorph and teleomorph species names will probably not be merged in the near future, due to their long 17

PAGE 18

history, opposition to change and the confusion that would likely ensue in mycology and plant pathology should this occur. Given the scarcity of useful morphological ch aracters in fungi, cryp tic speciation, wherein phylogenetically distinct species cannot be di stinguished m orphologica lly, is widespread (Hawksworth 2001, Kohn 2005). These are likely lineag es that have not been isolated long enough for notable morphological differences to develop (Kohn 2005), or in which a morphological character has evolved c onvergently (i.e. ch lamydospores in or setose acervuli in ; Sutton 1992, ODonnell et al. 1998b) These lineages are especially common in asexual, clonal specie s. Although some cryptic species can be recognized as distinct vegetative compatibility groups or mating popula tions, the widespread availability of DNA sequence data has allowed crypt ic species to be more easil y detected and characterized (ODonnell et al. 2000, Hawksworth 2001, Baker et al. 2003, ODonnell et al. 2004). Molecular phylogenies also suggest that some morphological characters used to describe major lineages in Fungi are homoplasious and need to be revisited (e.g. ascocarp and ascus development in the Pezizomycotina; Spatafora 1995, Lutzoni et al. 2004, Spatafor a et al. 2006). What role can morphology play in fungal systematics? Undoubtedly, many morphological characters represent important evolutionary changes that are still useful in classifying lineages of fungi. Likewise, there is evidence that ultrastructural characters may be a future source of informative data (Celio et al 2006), and may be able to an swer phylogenetic questions that molecular data has left unclear. For ex ample, Lutzoni et al. (2004) suggested a (Pucciniomycotina, (Ustilaginomycotina, Agaricomycotina)) relationship among the basidiomycete subphyla based on septal morphology before molecular phylogenies supported the same relationship (Aime et al. 2006, Hibbett 2006, Matheny et al. 2007). Unfortunately, the 18

PAGE 19

scarcity of these characters ensures that only small data sets are possible and, since most morphological characters are not represented across broad taxono m ic groups, they cannot be used to discern relationships among distantly related groups. In addition, morphological data usually cannot be used to accurately predic t phylogenetic relationships due to limited understanding of the evoluti on of these characters. Molecular data can help reconstruct the evol ution of morphological traits and determine whether characters are inherited via common descen t or convergent evolu tion (McLaughlin et al. 1995). Although molecular data will likely continue to be the major future contributor to fungal systematic analysis, morphology cannot be ignored if we are to understand organismal and functional evolutionary changes. By using mo lecular data as a guide to more complete phylogenies, we can understand the evoluti on of the diverse forms of the Fungi. The biological species concept describes speci es as groups of in terbreeding populations that are reproductively isolated from other su ch groups (Mayr 2000). Repr oductive isolation can result from physiological, genetic, ecological, and behavioral factors. In fungi, interfertility is governed by mating-type loci that code proteins for all aspects of mating (Kronstad and Staben 1997). Applying the traditional biological species concept to fungi is problematic due to the complexity of these mating systems and the widespread phenomenon of sexual sterility. There are two principal mating strategies in fungi, heterothallism and homothallism. In heterothallic fungi, individuals are self-sterile and gamete nuc lei must come from different mating types for karyogamy and meiosis to occu r (Alexopolous et al. 1996, Kronstad and Staben 1997). Ascomycetes generally exhibit bipolar hete rothallism, where mating types are determined by idiomorphs at a single locus (Kronstad and Staben 1997). An example of bipolar heterothallism is exhibited by the yeast where mating type is 19

PAGE 20

determined by the MAT locus that usually has two idiomor phs, a and The MAT locus must be heterozygous (i.e. MATa and MAT individuals must fuse) in order for meiosis to take place (Moore 1998). Basidiomycetes generally exhibit a more complex tetrapolar mating system that involves two loci, with many combinations of alleles (Kronstad and Staben 1997). In there are two mating type loci, A and B, which are located on different chromosomes (Moore 1998). Subloci exist at each locus, and ma ny alleles exist for each sublocus. Compatible individuals must differ at both loci, and because there are several alleles for each locus, many compatible combinations exist for successful fusion and meiosis. To divide groups of morphol ogically similar fungi into mating groups or biological species, non-identified strains are crossed with fertile tester st rains in the laboratory. In the species complex (GFSC) nine mating populations (A-H) have been identified, each of which contains morphologically similar anamorphs that exhibit varying levels of host specifici ty (Leslie 1995, Britz et al. 1999, ODonnell et al. 1998a). Mating tests have also identified cryptic, biological species in where and have been descri bed (Turner et al. 2001). Although mating tests can be used to detect biological species, several factors limit their utility. In mating tests, isolates may fail to cro ss with standard testers because they belong to a non-characterized mating population, have lost the ability to undergo sexual reproduction, or due to poor fertility (Leslie 1991). Thus, highly fertil e mating testers must be developed to increase mating probabilities. In addition, it may be difficult to interpret results of mating tests. For example, Guerber et al. (2003) used a seven point scale to rate results from crosses among strains of which ranged from the production of no structures to 20

PAGE 21

perithecia that produced vi able ascospores. In contrast, Leslie (1995) considered crosses in the GFSCto be fertile when perithecia were formed and exuded asci. For both of these studies, empty perithecia were considered infertile, ev en though such a result m ight suggest that occasional outcrossing may occur in nature. Mating tests are also cumbersome and time consuming, and highly dependent on environmental factors (G uerber et al. 2003). Another drawback to the biological species concept is that it cannot be applied to asexual species. Phylogenetic data can help define new biological species by: a) indicating where genetic barriers exist; b) identifying which lineages should be cr ossed to check for mating compatibility, thereby reducing the numbers of crosses that are needed; and c) specif ying in which mating population non-characterized isolates or new lineages may reside. Currently, the phylogenetic speci es concept predominates in fungal systematics. During the past 20 years it has relied in creasingly on molecular data, and transitioned from the use of isozyme protein markers to an array of PCR-base d molecular markers. Large sets of multilocus DNA sequence data are now most common and se veral genome-level analyses have been completed (Hu and Leger 2004, Delsuc et al 2005, Wolfe 2006, Aguileta et al. 2008, Stuckenbrock et al. 2009). These technological adva nces have radically strengthened our ability to resolve relationships among evolutionary lineages, and en abled the identification and characterization of numerous cr yptic lineages. As expected, clos e relationships between mating populations (biological species) and phylogene tic species have been demonstrated experimentally (ODonnell et al. 1998a, Steenkamp et al. 2000). Th e new challenge is to recover accurate and meaningful phylogeni es with DNA sequence data. Phylogenetic resolution at different taxonomic levels depe nds on utilizing appropriate regions of DNA. Three are most common in th e phylogenetic analysis of fungi: nuclear 21

PAGE 22

ribosomal DNA (currently most co mmon), nuclear protein coding DNA, and mitochondrial (mt) DNA; the utility and attributes of each is discussed below. Nuclear ribosomal DNAs include the larg e subunit (26S, LSU), small subunit (18S, SSU), and the 5.8S genes, as well as non-coding re gions such as the intern al transcribed spacer (ITS1 and ITS2) and the intergenic spacer (IGS) regions (Bridge et al. 2005). The most widely used loci are the ITS1 and ITS2 regions that flank the 5.8 subunit, and the majority of phylogenetic studies published i nvolving fungi include ITS data (Hillis and Dixon 1991, Nilsson et al. 2008). The wide use of ribosomal DNAs in systematic studies in Fungi is due to high copy number that facilitates PCR amplification, the availability of universal primers, and the resolution they provide at di fferent taxonomic levels (lvarez and Wendel 2003, Bridge 2002). Whereas the ITS and IGS regions are generally less conserved and can be used to resolve species and some subspecific taxa, the SSU and LSU are more conserved and appr opriate for studies at the super-generic level. When the ITS is not su fficiently polymorphic to identify cryptic species (de Meyer et al. 2008, Crouch et al. 2009a), region s with stronger signals must be identified to resolve such lineages. A large variety of the second type of locus, nuc lear protein-coding genes, has been used in the past. Some of the most common include the elongation f actor 1 alpha (EF-1 ), beta-tubulin ( -tubulin), RNA polymeras e large subunit (RPB2), and the mating (MAT) loci (ODonnell et al. 1998, Poggeler 1999, ODonnell et al. 2000, Reeb et al. 2004, Du et al. 2005, Matheny et al. 2007). The resolution that these lo ci provide depends on the group th at is studied. For example, in (Du et al. 2005), (ODonnell et al. 2004),(Barve et al. 2003), (Turgeon 1998), and and (Poggeler 1999), intraspecific variability was low in the high mobility group (HMG) region of the MAT1-2 22

PAGE 23

mating locus, but interspecific variabili ty was high, com pared to the ITS. In for which the ITS regions cannot be used, EF-1 is informative (Geiser et al. 2004). And in studies of / taxa, the -tubulin gene has been used (De Meyer et al. 2008, Roets et al. 2009). While these genes provide valuable da ta for multilocus phylogenies, they are usually not as conserved as the ribosomal subunits. Ther efore, specific primers may be needed for the studied taxa. Also, since large da ta sets, such as those that are available for the ITS, LSU and SSU ribosomal regions, may not exist for many nuc lear loci, comparisons of these data with other studies or taxonomic groups may not be possible. Mitochondrial DNA has also been employed in phylogenetic studies in fungi. It is usually inherited uniparentally and, thus subject to low levels of r ecombination (Xu 2005). It also evolves faster than nuclear DNA and may be more useful at lower taxonomic levels than the above loci (Waugh 2007). Nearly all fungal mitoc hondrial genomes characteriz ed to date contain genes that code for cellular resp iration chain subu nits, including and genes and ATPsynthetase subunits ( genes) (Paquin et al. 1997). These gene sequences are often employed in phylogenetic analysis at the species level, especially the cytochrome oxidase 1 gene ( ) (Waugh 2007, Vialle et al. 2009). M itochondrial genes have been propos ed as alternatives to ITS for barcoding in basidiomycet es (Vialle et al. 2009) and (Seifert et al. 2007). Caution should be used when incorporating mtDNA data in fungal phylogenetic studies. The wide range in gene/genome size due to in tergenic regions and introns (Gray 1998) that occurs among and within species could make amplification of a gene region among taxa unreliable. For example, in a study of 23 taxa, only 29% of the sequences were successfully amplified (Via lle et al. 2009). In addition, hi gh rates of genome rearrangement have been observed in some fungi (Gray 1989), which could compli cate alignment of mtDNA 23

PAGE 24

datasets among taxa, especially at higher taxono mi c levels. Although it is assumed that mtDNA is inherited uniparentally with a low rate of recombination, biparental inheritance and genome recombination have been reported (May and Taylor 1988, Barroso and Laberre 1997, de la Bastide and Horgen 2003, Xu 2005). Lastly, the preferential inheritanc e of one mitochondrial genome over another has been demonstrated (X u et al. 2000, de la Bastide and Horgen 2003). Edwards (2009) discussed the effect of natural selection on phylogeny reconstruction, and concluded that although many t ypes of selection can be addr essed by analytic methods and should not affect tree topologies, balancing sel ection and selection-based convergence of aminoacid substitutions could resu lt in the identification of monophyletic lineages among evolutionarily discrete clades. Si nce some nuclear loci share this problem, consideration must be given to the poor understanding of the selective forces that act on the loci that are used in phylogenetic analyses and the impact that this ma y have on the resultant phylogenies. Testing for congruence among trees from several DNA re gions can help address this problem. Since different evolutionary f actors affect different classes of DNA, employing data from at least two classes may provide more robus t phylogenetic assessments (Taylor et al. 2000, ODonnell et al. 1998b). Depending on the level of taxonomic resolution that is desired, each DNA region has advantages and disadvantages. For example, highly conserved loci do not resolve closely related species bu t may be useful at the generic or family level. When such studies are begun, preliminary analys es of several loci in a few accessions are advisable as they can help focus more comprehensive, subsequent work. Whereas species that are defined by the morphological, biologi cal and phylogenetic species concepts do not always ove rlap, and in some cases directly disagree, the evolutionary species concept utilizes these concepts to identi fy a species as a single lineage of ancestral 24

PAGE 25

descendant populations of organisms which m ainta ins its identity from other such lineages and which has its own evolutionary tendencies and historical fate (Wiley 1978). The strength of the evolutionary species concept lies in this combination of morphological, biological and phylogenetic evidence to identify di vergent evolutionary lineages. Taylor et al. (2000) different iated between the theoretical (the evolutionary species concept) and the operational (morphological, bi ological, and phylogenetic) nature of these species concepts, and suggested that the operational concepts could be used to fulfill the theoretical. They recommended using a phylogenetic approach to define sp ecies that fit Wileys (1978) evolutionary species concept, and cited several studies with mo rphological data that underestimated lineage number, presumably because lineage divergen ce occurred before morphological differences develope d. Wiley (1978) recognized th e limited ability to recognize evolutionary species that existed at that time, and stated that rea l evolutionary lineages exist in nature outside mans ability to perceive these lineages. The ability to detect cryptic species has improved dramatically since Wileys (1978) publication, but the challenge now is to identify the levels at which satisfactory and useful lineage resolution occurs. This is where the op erational species concepts can play a role. Progress towards defining real evolutionary species would be made with holistic considerations of biological (mating compatibility, ecology, pa thology, etc.), phylogeneti c and other types of data. In the meantime, broad-based evaluations ar e preferable to a reliance on a single or limited sets of characters for species delineations. The correct identification and characterizati on of lineages within phytopathogenic genera has both practical and academic applications and has relied incr easingly on molecular phylogenetics. Molecular data can be especially useful to id entify cryptic species (Kohn 2005, 25

PAGE 26

Couch et al. 2005, de Meyer et al. 2008, Crouch et al. 2009b), wh ich in turn enhances our understanding of the complexity of and interactions among plant pathogenic populations. For example, DNA sequence data can address quest ions on disease ecology and epidem iology, host specificity, pathogen nutrition and life strategies, and the evolution of pathogenicity. Since human-derived selection pressure can be substantial in agroecosystems, it is important to be able to predic t these effects on pathogen evolu tion (Burdon and Thrall 2008). To determine these effects, pathoge n evolution is assessed over time The evolution of resistance genes in natural host populations, fungicide sens itivity in agricultural pathogen populations, and host specificity in natural and agricultural setti ngs are some of the proc esses that can be better understood by tracking with molecular phylogenetic analyses the e volution of genotypes that are associated with these phenotypes. Understanding host specificity in pathogenic species and populations is necessary to understand pathogen evolution and ultimately devise effective disease management strategies. Many phytopathogenic fungi that have broad host ranges (e.g. spp. spp and ) are now known to be comprised of cryptic species that are themselv es often host-specific (Salazar et al. 2000, Steenkamp et al. 2002, ODonnell et al. 2004, Fourni er et al. 2005, Peever et al. 2005, Peres et al. 2008, Crouch et al. 2009b). Recognizing and characterizing these lineages can help understand the epidemiology of a disease, measur e infection of hosts and nonhosts, and regulate pathogens that may or may not be morphologically distinguishable. National and international quarantines for these pathogens will depend increasingly on rapid and reliable molecular tools for identification. 26

PAGE 27

To investigate host specifi city and coevolution in Couch et al. (2005) used phylogenetic analysis combined with constr uction of haplotype networks that used sequence data from 10 l oci. They were able to tr ack host shifts in the pa thogen to rice and other grassy weeds that were associated with rice cul tivation. They combined th e molecular data with pathogencity tests to associate host specificity with haplotype lineages. Another study investigated host specificity in the species complex. Crouch et al. (2006) used sequence data from the MAT1-2 HMG-box locus, the ITS region, and a superoxide dismutase gene to de monstrate that the cl osely related species and two divergent lineages of all formed wellsupported, host specific lineages. Further study of the lineages in with the same three loci and the MAT1-flanking apurinic DNA-lyase ge ne enabled the identifi cation of two lineages associated with C3 and C4 hosts, respectively, and the resolu tion of several new host-specific cryptic species (Crouch et al. 2009b). Phylogenetic analysis can also be used to investigate whether fungal symbionts are latent pathogens or saprotrophs. Promputtha et al. (200 7) used a phylogenetic approach to evaluate relationships between endophytic and saprotrophic strains of and Although endophytic strains often fell in the same clade as saprotrophic and pathogenic strains, they did not test the abil ities of endophytes to cause disease or colonize senescent tissue. As discussed above, the va lue of phylogenetic studies can be enhanced considerably with corresponding biological information for a fungus. The evolution of pathogenicity has been inferred using molecular phylogenetics. Studies involving the origin of several pathogenic forma speciales are the best example of this application. Baayen et al. ( 2000) constructed phylogenies of several formae 27

PAGE 28

speciales that cause wilt disease and bulb rot (f.sp. and ). They indicated that several of the formae speciales were polyphyletic and each appear to have evolved from at least two different lineages. Similar studies have found multiple evolutionary origins for other formae speciales: two races of f. sp. from lettuce in Arizona appear to have evol ved separately (Mbofung et al. 2007); five distinct lineages of the Panama disease pathogen of banana, f. sp. exist from different geographic origins (ODonnell et al. 1998b); and two formae sp ecialis affecting cucurbits, f. sp. and f. sp. form separate lineages (Vak alounakis and Fragkiadakis 1999). Recent genomic studies suggest that pathogenicity evolves in through horizontal gene transfer of pathogenicity factors, partic ularly on the supernumerary chromosomes of the genome (Coleman et al. 2009, Michielse and Rep 2009). Another example of the use of phylogenetic analys is to track the evolu tion of pathogenicity is in which causes tan spot of wheat. It has been shown that the fungus gained its pathogenicity th rough the horizontal gene transf er of a peptide toxin-encoding gene, from another wheat pathogen (Friesen et al. 2006). This hypothesis was supported through analysis of the gene sequences in isolates of both species, where displayed more diversity in the gene (11 haplotypes) than (1 haplotype), coupled with the fact th at the gene was requi red in both species for disease development. Phylogenetic analysis of sequence data can be used to track the geographic spread of pathogens. A four-gene phylogenetic an alysis of a global collection of indicated that the pathogen originated in Eurasi a (where a total of six lineages were found) and then spread to North America (two lineages) (Linzer et al. 2008). They also identified a likely 28

PAGE 29

human-derived introduction into a region of Italy, based on the reso lution of isolates from Italy in the North Am erican clade. The studies described above illustrate how mol ecular phylogenetic analysis can be used to answer specific questions about phytopathogenic fungi. As DNA sequencing becomes easier and more affordable, and more loci are developed for studies at the species and intraspecific level, additional applications will be found for these types of data. As we enter the era of phylogenomics, these possibilities grow. The overall objective of this dissertation research was to use DNA sequence data and phylogenetic analyses to address question of host a nd tissue specificity on two important tropical plant pathogens. The pathogens chosen cause witches broom of cacao, and anthracnose of tropical fruit crops, spp. In the first study, phylogenies of were used to test the validity of biotype designations that are currently used for the species. These phylogenies were used to determine if should be considered a single species or a species complex, and to hypothesize on whether different lineages and biotypes might affect other host taxa. The second study focused on the as a mango pathogen in south Florida. P hylogenetic and pathological data were used to determine: if a mangospecific population exists; whethe r this population should be considered a separate species; and if there are tissue-specific lineages within the mango population. Cacao, is a neotropical crop with two reported centers of diversity (Bartley 2005). The primary center of diversity is in the vast Amazon region, extending in the north from southern Colombia and Venezuela to eastern Peru and Bolivia in the south, and from east of the Andes to northern Brazil in the west Due to pre-Columbian cultivation of cacao, a 29

PAGE 30

secondary area of diversity exis ts in the Caribbean region, in cluding Mexico, Central America and the Caribbean islands. The Mayans and Azte cs prized cacao as a foodstuff and currency (Coe and Coe 1996). The first European contac t with cacao was by Colombus in 1503, and it quickly gained popularity in Sp ain and then the rest of Eu rope (Coe and Coe 1996). Today chocolate is one of the most beloved foodstuffs in the world. Significant cacao production occurs in tropical Africa, Asia, and America. According to World Cocoa Foundation statistics (2008, http://www.worldcocoafoundation.org/infocenter/statistics.asp ), the 2006/2007 crop was almost 3.5 m illion tons of cacao beans. Most cacao was produced in Central and South America before the 20th century, but West Africa has now become the world leader (Gray 2001). West Af rica produces about 70% of the world cacao crop, with Cte dIvoire being the largest producer by far. South Am erica currently produces just above 10% of the global crop, largely due to loss es and management costs that are associated with witches broom and frosty pod diseases (Gray 2001). The greatest limiting factor in cacao production is a trilogy of important diseases: black pod, witches broom, and frosty pod (Fult on 1989, Evans 2006). Black pod, caused by and (Appiah et al. 2004), is the most economically damaging disease glob ally (Taylor 1994, Bowers et al. 2001, Evans 2006). However, witches broom (caused by ) and frosty pod ( ), have the potential to devastat e the world cacao supply if they were to spread outside their current American range. The cacao industry was d ecimated in Ecuador af ter the appearance of witches broom in the early 1900s (Evans 1981b). The potential for catastrophic damage from the disease is evident in Bahi a, the largest cacao-producing re gion of Brazil. After witches broom appeared in Uruuca and Camacan in 1989 (Pereira et al. 2006), it spread rapidly 30

PAGE 31

throughout the region and reduced production in Brazil from 378,000 tons in 1990 to less than 120,000 tons in 1999 (Gray 2001). L osses due to frosty pod are similarly dramatic (Evans et al. 1998). The disease spread from Ecuador and Colomb ia in the early 1900s s outh into Peru (Evans et al. 1998), and then recently into Central Amer ica and southern Mexico (Phillips-Mora et al. 2006a, b). and are closely related basidiomycetes in the Marasmiaceae and Agaricales (Aime and Philip s-Mora 2005, Matheny et al. 2006). Both fungi: a) are indigenous to South America, b) affect c) are hemibiotrophs, d) influence the hormonal balance of the host, e) share a mushroom odor, f) have the dolipore septum characteristic of basidiomycetes, g) produce incrusted red pigmented hyphae, and e) produce similar early symptoms in affected pods. Due to these similaritie s a close relationship between the two fungi was proposed (Evans et al. 1978, Evans 1981a, Evans et al. 2002), and has been confirmed with molecular data. Based on a five-locus molecular phylogeny, Aime and Phillips-Mora (2005) found that and were sibling species in the Marasmiaceae, and that was more closely related to than to other species. The pathogen was theref ore transferred to the genus Classification of proved difficult, due to its ex ceptional biology. The fungus was first described as (Ciferri and Parodi 1933), an anamorphic discomycete genus in the Ascomycota. This classification was based on the production of conidia in long chains. However, Evans et al. (1978) discovered that produced conidia basi petally rather than acropetally, which is a characteristic of and produced dolipore hyphal septa, a characeteristic of the Basidiomyc ota. They erected a new genus, to accommodate the pathogen, and described it as an anamorphic (asexual) basidiomycete. Evans et 31

PAGE 32

al. (2002) reported that the supposed conidia of were actually dikary otic probasidia that have undergone the first but not th e second meiotic division; thus, appears to undergo sexual reproduction without form ing basidiocarps. These meios pores are produced in white powdery masses on white, fleshy pseudostroma on the surface of affect ed fruit (Evans 1981a, 2006). Spores are dispersed by wind and rain and infect immature cacao pods, cause swelling and premature ripening of infected areas, and eventually produce abundant inoculum for new infections (up to 44 million spores per square centimeter) (Evans 1981a). (= ) was originally described as by Stahel (1915), and transferred to the genus (Singer 1942) due to the presence of long thick-walled pileal hair s. The name of the disease it causes, witches broom, refers to the characteristic proliferati on of swollen and twisted shoots that develop on cacao, which were first described by Went (1904) in Surinam. Cacao pods and flower cushions (woody meristematic tissue above le af scars; de Almeida and Valle 2007) may also be affected, resulting in large yield losses. Only basidios pores, which are produced on basidiocarps on necrotic host tissue, are capable of infection. Much progress has been made in understanding the biology of and its interaction with the host. It is a hemibiotroph (Pegus 1972), w ith a biotrophic phase that consists of primary, monokaryotic mycelia that infect and colonize host tissue. As the host tissue necroses, the fungus switches to a saprotroph that produces seconda ry, dikaryotic mycelia. These changes have been observed in histological studies that show swollen, thick-walled, monokaryotic hyphae growing interc ellularly in infected host tissue until brooms necrose; thereafter, the fungus switches to a dikaryotic, fine mycelium that colonizes the necrotic tissue (Evans 1980, Ceita et al. 2007). 32

PAGE 33

It is assumed that uses endogenous or host ho rmone production to induce broom formation. The formation of brooms is the result of cell enlargem ent rather than cell proliferation (Orchard et al. 1994) leading to less organized tissu e. Indoleacetic acid (IAA) and salicylic acid (SA) levels are elevated in inf ected leaf tissue (Kilaru 2007), however, there is little evidence of the invo lvement of cytokinins (Orchard et al. 1994). Infected cacao tissue undergoes programm ed cell death as it necroses, which is accompanied by DNA degradation, calcium oxalate production, and increased concentrations of ascorbic acid and hydrogen peroxide (Ceita et al. 2007). resists oxidative stress and utilizes the breakdown products of host cells as nutrients as it transitions to its saprotrophic form (Ceita et al. 2007, Santos et al. 2008). Recent work has also shed light on mechanis ms of pathogenicity. Multiple biotypes of produce degradative enzymes, including cellulase, protease, lipase, amylase, esterase, and peroxidase (Bastos 2005). Several necros is and ethylene-induci ng proteins (MpNEP1, MpNEP2, MpNEP3) have been identified in th e fungus, which may play a role in disease development (Garcia et al. 2007), as well as cerato -platanin protein (MpCP1), which is similar to that produced by (Zaparoli et al. 2009). The synergistic effect of MpNep2 and MpCP1 produced symptoms similar those caused naturally by (Zaparoli et al. 2009). Witches broom is difficult to manage. Severa l fungicides reduced the incidence of pod rot, but were not able to protect vegetative shoot s (Laker and Ram 1992). In addition, the high cost of frequent applications that are needed to protect growing pods has reduced the utility of fungicide application (Laker and Ram 1992). Pod removal d ecreases diseas e incidence (Soberanis et al. 1999), but is also labor intensive. Alternative measures ar e key to the future of 33

PAGE 34

cacao production in the Americas. In this regard biocontrol and disease resistance have received much recent attention. Potential biological control agents utilize antibiosis, parasitism and competition, and include (Bastos et al. 1981), (Macagnan et al. 2006, Macagnan et al. 2008), several spp. (Krauss and Soberanis 2002, Sanogo et al. 2002, Aneja et al. 2005, Bailey et al. 2006, Bailey et al. 2008), and even and (Meja et al. 2008). Despite th e amount of work that has been devoted to this area (Coe et al. 2006, Bastos et al. 1981, Aneja et al. 2005, de Marco and Felix 2007, Bailey et al. 2008, Macagnan et al. 2008, Meji a et al. 2008), limited success has been observed in field trials (Aneja et al. 2005, Mej a et al. 2008). The best hope for cacao in South and Central America lies with the development of cacao clones with durable resistance to witches broom, frosty pod, and black pod. Breeding programs exist in many producing regions, and often focu s on resistance to the most important local disease(s). The first step in breeding resistance to witches broom is to identify potential resistant parents in the wild and in co llections. Bartley (2005) gives a detailed description of known diversity worldwide, including the historical provenance of major groups of cacao clones. The International Cocoa Ge rmplasm Database (ICGD, http://www.icgd.rdg.ac.uk/ ), based at the University of Reading, UK, is available online and gives detailed descriptions of most known clones, including compiled photographs, physiological characteristics, SSR profiles, and available disease resistance data. Several independ ent studies have been u ndertaken in Brazil to characterize genetic diversity of cacao collections with a focu s on witches broom resistance (Marita et al. 2001, Paim et al. 2006). Through breeding efforts, several progeny display useful 34

PAGE 35

tolerance to witches br oom. Several resistance QTLs have been identified, with the ultimate goal of pyramiding resistance loci to produce clones with durable resistance (Queiroz et al. 2003, Brown et al. 2005, Faleiro et al. 2006, Pa im et al. 2006, Santos et al. 2007). The diversity of has been studied, and molecu lar tools are yielding further insight into the origins and evol ution of the pathogen. Basidiocar p variation was first described by Pegler (1978), who identified three varieties: a) var. with a crimson pileus in the center and fading to white at the margins, id entified in Trinidad and Surinam; b) var. with a uniformly red pileus, identified in Ecuador; and c) var. with a yellow pileus and identified from a single broom in Ecuador. This varietal classification has given way to a host-based system of biotypes. Fi ve biotypes have been de scribed: a) C-biotype, which affects spp. (including cacao) and spp. (Evans 1981b); b) S-biotype, which affects spp. ( and ) (Bastos and Evans 1985, Pereira et al. 1997, Rincones et al. 2006); c) B-biotype, which affects (Bastos and Anderbrhan 1986); d) L-biotype, which includes presumed saprotrophs found on lianas in th e Bignoniaceae and Malpighiaceae (Evans 1978, Pegler 1978, Griffith and Hedge r 1994b); and e) H-biotype from (Resende et al. 2000). Mati ng studies have suggested that the Cand S-biotypes are homothallic, but the L-biotype is heterothallic (Griffith and Hedger 1994a,b). Molecular data have been used to investig ate relationships between biotypes. RAPD and RFLP (restriction fragment length polymorphism ) fingerprints support the hypothesis that the S, C, H, and L-biotypes represent distinct populations (Anderbrhan and Furtek 1994, de Arruda et al. 2003a,b, Rincones et al. 2006). Ho wever, there have been only two phylogenetic analyses for One, which included an analysis of the IG S region of 14 isolates in the C, S, and 35

PAGE 36

H-biotypes, indicated that the biotypes formed well-supported cl ades (de Arruda et al. 2003a). Another analysis of the ITS region in a similar ra nge of isolates, found that an H-biotype isolate resolved sis ter to a clad e containing isolates from and(de Arruda et al. 2005). A multilocus phylogenetic analysis, whic h would facilitate a more complete picture of lineages within has not been conducted. Colletotrichum Colletotrichum Corda erected the genus in 1837 with the type species (Corda 1831). Many genera have been placed in synonymy with including The two genera were previously disti nguished by the presen ce of setae in the acervulus of but not Before the 1950s, species of were erected based on host substrate, leading to a proliferation of species in the genus with little biological or phylogenetic basis. In his revision of the genus in 1957, von Arx reduced more than 750 species of and 40 other genera1 to only 19 species of (Table 1-1). This classification was based solely on morphology, and von Arx used type specimens (approx. 25% of the taxa that were reviewed), pure cultures, and descriptions to group taxa into species. Morphological characters that were used to distinguish species were conidium shape and size, the pr esence of sclerotia in culture, chlamydospore development, host specificity, and acervulus size. Most notable about th is classification are / which von Arx synonymized with over 600 previously described species. He also distinguished several forms of with anamorphs that differ from 1 36

PAGE 37

, including from (bean), from Cucurbitaceae (cucurbits), from (banana), from (cotton), from from from from and (leguminous forages), and from The next revision of was by Sutton (1980), who de scribed 22 species based on conidium shape and size, presence of sclero tia, appressorium shape and size, and host specificity (Table 1-2). and were included as newly described species, and several species that were synonymized by von Arx were distinguished: and from ; and from ; and and from A standardized descrip tion was not given for as Sutton considered such a description meaningless considering the species wide variation. Due to intrataxon variation, and spp. on graminaceous hosts were each considered species complexes. Sutton (1992) provided a subsequent description of 33 widely accepted species (Table 1-2). Several of the additional taxa were described by von Arx (1957) ( , and ), whereas others were new ( , and ). Since Suttons (1992) lis t was published, several new species have been described2. 2 : in Farr, Aime, Rossman & Palm, (12): 1401 (2006) 37

PAGE 38

Currently there are 674 entries of taxa in the Index Fungorum ( http://www.indexfungorum.org/ ). a species with stra ight conidia that was described by Moriwaki et al (2003), is differentiated from by having wider conidia, and by its phylogene tic placement outside the large sensu lato clade. Since was erected, it has been associat ed worldwide with hosts in , and the Agavaceae and Proteaceae (Moriwaki et al. : (4): 594 (1981) : (1): 48 (2003) : Wang & Li, (4): 212 (1987) : (suppl. 1): 12 (1988) : (8): 993 (1993) : (2): 309 (2002) : in Fa rr, Aime, Rossman & Palm, (12): 1401 (2006) : (6): 647 (1997) : [as '], : 6 (1996) : in Far r, Aime, Rossman & Palm, (12): 1403 (2006) : (2 0 06) Index of Fungi 7: 929 : Chevassut & Pellicier, (3): 197 (2002) [2001] : (5): 149 (2001) : (1 ) : 66 (1992) : (12 ) : 1525 (1993) : (3): 280 (1998) : : 139 (2007) 38

PAGE 39

2003, Lubbe et al. 2004, Farr et al. 2006, Avila-Q uezada et al. 2007). Based on its m orphology and phylogenetic placement outside Farr et al. (2006) re-established for setose taxa found on and (the species had been included in by von Arx). Farr et al. (2006) also erected two new species to accommodate host specific taxa on the Agavaceae: (on ) and (on ). Studies such as these that comb ine morphological examinations of new isolates and herbarium materials with robust mu lti-locus phylogenetic analyses will be key to meaningful classifications in the genus. Sutton (1980) considered to be a species complex, and wrote later that no progress in the systematics and identification of isolates belonging to this complex is likely to be made based on morphology alone (Sutton 1992). Wide ranges in have been reported for colony color and growth, conidium dimensions, appressorium and hyphopodium shape and size, and host range. Fo r example, Sutton (1980) reported that hyphopodia can be clavate, ovate to lobed, and th at conidia can measure 6-20 x 4-12 m, a far greater range than is recognized for other species, such as (conidia 8.5-16 x 2.5-4 m) or (conidia 18-23 x 3.5-4 m). The / complex contains several sp ecies that have not been distinguished with mo lecular phylogenies: cause of banana anthracnose; cause of coffee berry disease; and cause of crown rot of st rawberry. Unfortunately, there are few informative sites in DNA regions that have been used to date that differentiate these taxa, and the relationships among th e three species and their relationship to sensu stricto is not clear (Sreenivasapra sad et al. 1993, Sreeni vasaprasad et al. 1996, Munaut et al. 2002, Martinez-Culeb ras et al. 2003, Du et al. 2005). 39

PAGE 40

As phylogenetic analyses with DNA sequence data become commonplace, the systematics of and other species in this genus will recognize true evolutionary relationships in these fungi. Understanding th e different diseases that are caused by these pathogens is not possible wit hout recognizing the different ca usal agents. Recent work has shown that is comprised of distinct genetic clades that often display different host specificities (Aba ng et al. 2002, 2005, Munaut et al. 2002, MacKenzie et al. 2007). A systematic review of the species complex that uses both molecular and biological data would help clarif y the evolutionary history of th is important plant pathogen and enable more focused work on the epidemiology and management of the corresponding diseases. Genetic diversity studies on include comparisons of electrophoretic patterns of molecular markers (mostly random amplification of polymorphic DNAs, RAPDs), occasionally accompanied by construction of dendrograms based on UPGMA (unweighted pair group method with arithmetic mean ) analysis. These studies include the characterization of: a) two populations from yam that differ in pathoge nicity (Abang et al. 2002); b) populations on different species in Mexico, Braz il and Colombia (Munaut et al. 2002, Weeds et al. 2003); c) distinct mango and avocado isolat es in Australia (Giblin 2005); and d) and on strawberry (Martinez-Culebr as et al. 2002, Ur ea-Padilla et al. 2002, Xiao et al. 2004, MacKenzie et al. 2007). Use of sequence data is rare, and phylogenetic analysis to determine taxonomic relationships among lineages in the species complex is lacking. Colletotrichum gloeosporioides Almost every fruit and vegetable crop is affected by postharvest anthracnose caused by spp. is most often associated with tropical hosts, and causes severe economic damage on a number of important fruit crops worldwide (Freeman 40

PAGE 41

et al. 1998). On mango, the pathogen causes bl ossom blight, leaf anthracnose and fruit anthracnose. Three taxa have been associated with these diseases: (worldwide), var. (Australia, Simmonds 1965), and (Australia, Fitzell 1979; Taiwan, Weng and Chuang 1995; and Homestead, Florida, Riveras-Vargas et al. 2006). Several groups have studied the influen ce of relative humidity and temperature on conidium germination and appressorium form ation (Dodd et al. 1991b, Estrada et al. 2000, Dinh et al. 2003). According to Dinh et al. (2003), conidia germinat e 48 hr after inoculation onto mango peels, and appressoria form 30-96 hr af ter germination. High relative humidity of 95100% is necessary for germination, and appre ssorium formation peak s at 25C (Dodd et al. 1991b, Estrada et al. 2000). Reservoi rs of inoculum exist on fallen leaves and fruit, as well as necrotic peduncles that failed to set fruit in th e previous season (Arauz 2000). Affected leaf flushes also serve as important sources of inoculum (Fitzell and Peak 1984, Dodd et al. 1991b). Newly emerged inflorescences and leaves are sus ceptible, resulting in bl ossom blight or leaf anthracnose. Conidia are most pr evalent during rainy periods, and rain events are important for disease development (Fitzell and Peak 1984). The role of the teleomorph in the epidemiology of the disease is no t clear. Fitzell and Peak (1984) were able to produce ascospores in culture, but they were never found in the field. Although more than one cycle of infection can occur on leaves and blossoms, fruit anthracnose is monocyclic (Arauz 2000). Infectio ns remain latent on fr uit until they begin to ripen. Most work studying latent infection on fruit by has been done on avocado (Prusky and Lichter 2007, 2008). After fruit set, conidia on the fr uit surface germinate and form appressoria that penetrate the fruit cuticle and form a biot rophic latent infection. 41

PAGE 42

Colonization of the fruit is i nhibited by host defenses (such as the production of reactive oxygen species, and the presence of an antifungal dien e) until fruit maturation (Beno-Moualem and Prusky 2000, Guetsky et al. 2005), when hydrolytic enzymes secreted by the fungus enable tissue maceration and saprotrophic colonization (Prusky and Lichter 2007, 2008). Guetsky et al. (2005) showed that laccase produced by breaks down epicatechin, which enables the break down of the an tifungal diene, activa ting the latent infection. Pectate lyase secreted by and other species has been shown to be an important factor for host penetration by and pathogenicity on avocad o (Yakoby et al. 2000a, 2001). The regulation of synthesis and secretion of this pectate lyase is dependent on alkalization of the host tissue by the pathogen (Yakoby et al. 2000b), and the fungus secr etes ammonia to enab le this pH change (Prusky et al. 2001, Kramer-Haimovi ch et al. 2006). The presence of sugars, such as fructose, glucose and sucrose, also play a role in enzyme secretion (Miyara et al. 2008). Droby et al. (1986, 1987) have invest igated latent infections of on mango and found that, like the -avocado pathosystem, mango peels contain antifungal compounds (resorcinols) that limit colonization of the pathogen and symptom development until fruit maturation. To date, fifteen resorcinols have been identified in mango peels (Kndler et al. 2007), and a direct correlation has been demonstrated between the presence of resorcinols and anthracnose lesion si ze on mango fruits (Hassan et al. 2007). Due to the latent infection exhibited by the pathogen, it has been suggested that coevolved as a climacteric fruit pa thogen (Flaishman and Kolattukudy 1994, Arauz 2000). The latent infection exhibited by th e pathogen works to the advantage of both host and pathogen. By not causing disease until the fru it is mature, anthracnos e allows the host to 42

PAGE 43

reproduce and then facilitates ra pid seed germination by decomposing the fruit flesh (Arauz 2000). Anthracnose ma nagement in humid growing regi ons consists mainly of pre and postharvest fungicide treatment, and is especially necessary when fruit develops in humid, wet conditions (Arauz 2000). Frequent applications continue from shortly be fore bloom until shortly before harvest. A number of chemistries have shown varying levels of effi cacy, including thiophanatemethyl, benomyl, copper, prochloraz, sulfur, ferbam, and azoxystrobin (McMillan 1984, Dodd et al. 1997, Ploetz and Prakash 1997, Mossler and Nesheim 2002, Sundravadana et al. 2006). The classes at highest risk for resistance are th e benzimidazoles and thiophanates (benomyl and thiophanate-methyl) (Sanders et al. 2000), wher eas inorganic compounds such as copper and sulfur carry little risk of resistance (Anonymous 2009). The use of chemistries such as prochloraz and thiabendazole has no t been shown to lead to the development of resistance in (Kuo 2001, Sanders et al. 2000), even though resistance has developed in other fungi (Dyer et al. 2000, Smilanick et al. 2003). When there is a risk of re sistance developing in a given fungicide, its use should be: a) alternated with pr oducts with different modes of action; b) limited; and c) according to the manufacturer la bel (do not use reduced rates); recommendations for specific fungicides are reporte d by Brent and Hollomon (2007). The use of weather-based models that pred icted the likelihood of disease development reduced the numbers of sprays that were needed for effective anthracnose management in the Philippines (Dodd et al. 1991b, Estrada et al. 1996). And postharvest treatments are as important as field treatments to manage anthracnose on fr uit. Immersion in hot water, especially when accompanied by fungicides, is very effective in lowering anthracnose on fruit (Muirhead 1976, 43

PAGE 44

Thompson 1987, Dodd et al. 1991a, McGuire and Cam pbell 1993). However, no fungicides are currently labeled for postharvest application in the United States (Prusky et al. 2009). Several studies have attempted to identify and develop biocontrol agents for anthracnose. Koomen and Jeffries (1993) identified 121 mi croorganisms that were antagonistic to and showed that a single strain of decreased anthracnose levels on fruit in the Philippines. Vivekananthan et al. (2004) found strains of and that decreased anthracnose incidence on fruit when applied with a chitin formulation. Most recently, a formulated product containing was as effective as benomyl in decreasing anthracnose in Haden and Kent in the field (Patio-Vera et al. 2005). In addition, B ugante and Lizada (1996) found that bagging immature fruit with brown paper bags decreased disease levels by 50%. No complete resistance has b een identified in mango, although partial resistance has been noted for some cultivars. Resistance has been quan tified for several cultivar s. Knight (1993) used a 1-9 scale, with 1 being the most susceptible, and 9 the most resistant. He reported that the cultivars Tommy Atkins, K eitt, Kensington, and Van D yke were the most resistant (ratings of 7 or higher), and cultivars Pope and Alfonso were the most susceptible (ratings of 2-3). Crane et al. (2003) descri bed characteristics of 28 cultivar s and gave them one of four resistance ratings: moderately resistant, m oderately susceptible, susceptible, and very susceptible. Ten cultivars were moderately resi stant, including Edward, Florigon, Keitt, Tommy Atkins, and Van Dyke, whereas Irwin and Kent were very susceptible. Dinh et al. (2003) investigated anthracnose resistance in Thai mango cultivars and found that Rad, Kaew, and Chok Anan were more resistant than Nom Doc Mai and Nan Klang Wang. 44

PAGE 45

Mango-specific populations of have been identified in several studies. With RFLPs and isolates from four tropical fruit species, Hodson et al. (1993) showed that only isolates from mango were genetically similar. Other RAPD and RFLP studies corroborated the presence of a distinct mango population in worldw ide collections from tropical fruits (Alakahoon et al. 1994, Hayden et al. 1994). Cross inoculatio n studies have shown that isolates of usually cause symptoms on all fruit, rega rdless of the original host, but that lesion diameter is generally greater on the original host (Alakahoon et al. 1995, Sanders and Korsten 2003). Additional work is warranted to qualify and quantify the phylogeny and biology of populations from mango and other hos ts. Contemporary investigations of the tissue-specificity of mango populations ar e also needed. Although pectic zymograms (Gantotti and Davis 1993) and RAPDs (Davis 1 999) demonstrated some correlation between isolate and host organ (leaf, in florescence and fruit), modern data are needed to clarify the existence of these biotypes. 45

PAGE 46

46 Table 1-1. von Arx classification of species of (1957), including anamorph, teleomorph (when described), numbers of synonyms von Arx placed in each taxon, and host plant(s). Species (anamorph) Species (teliomorph) # synonyms Host 600 Numerous 3 12 Cucurbits 2 5 Malvaceae 1 1 1 and (legume forage) 1 --18 Numerous --4 --2 Legumes --3 --3 --2 --1 --15 Numerous 41 Graminaceae --89 Numerous

PAGE 47

Table 1-2. Major revisions of by von Arx (1957), Sutton (1980), Sutton (1992), and others. Sutton (1980) Currenta von Arx (1957) Anamorph Teleomorphb Anamorph Teleomorph Anamorph Teleomorph c ----------------------------------------------d ----------d e ------f ----------47

PAGE 48

48 Table 1-2. Continued. a Based on Sutton (1992) except where footnoted b Dashed line denotes no tele omorph described for the taxon c Guerber and Correll 2001 d Farr et al. 2006 e Moriwaki et al. 2003 f Waller et al. 1993 Anamorph Teleomorph Anamorph Teleomorph Anamorph Teleomorph ------------------d ----------

PAGE 49

Saccharomycotina Pezizomycotina Taphrinomycotina Ascomycota Agaricomycotina Pucciniomycotina Ustilagomycotina Basidiomycota Glomeromycota Zygomycete 3 (Mucorales+Endogonales) Zygomycete 2 (Mortierellatales) Zygomycete 1 (Kickxellales, Harpellales, Dimargaritales, Zoopagales, Entomophthorales, Basidiobolaceae) Blastocladiomycota Chytridiomycota Rozella Animals (Outgroup) Figure 1-1. Proposed relationships among fungal phyla as reported in the AFTOL Deep Hypha issue of (2006, v. 89, issue 6). 49

PAGE 50

CHAPTER 2 INVESTIGATING TH E RELATIONS HIPS AMONG BIOTYPES OF USING PHYLOGENIES FROM MULTIPLE DNA REGIONS Cacao ( ) is an important tropical crop worldwide. It is native to the headwaters of the Amazon, with a secondary area of diversity in Mesoamerica due to extensive prehistoric cultivation (Bartley 2005). The Mayans and Aztecs used cacao not only as a popular beverage, but also as a form of currency. The Maya ns held the crop in reli gious esteem (Coe and Coe 1996). After introduction to Europe in the 16th century its popularity grew, and chocolate has become one of the worlds most loved foods. World production in 2006-2007 was estimated at 3.5 million tons, 70% of which occurs in West Africa (World Cocoa Foundation statistics, 2008, http://www.worldcocoafoundation.or g/info-center/ statistics.asp ). The greatest limiting factor in cacao production is disease. The three mo st important diseases are black pod, caused by several specie (Appiah et al. 2004); witches broom, caused by (Stahel) Aime & Phillips-Mora; and frosty pod, caused by (Cif.) H.C. Evans, Stalpers, Samson & Benny (Fulton 1989, Evans 2006). Black pod is the most economically damaging disease globally (Bowers et al. 2001, Evans 2006), but the two diseases caused by spp. have the potentia l to devastate the world cacao supply if they spread to the larg est production areas in Africa and Indonesia. The devastating effect of these two di seases is still being played out in South and Central America. After their appearance in Ecua dor in the early 1900s, the c acao industry was decimated, and many plantations abandoned (Evans 1981b). The potential for witches broom to cause catastrophic damage is evident from production nu mbers in the largest cacao-growing region of Brazil, Bahia. The disease initially appeared in Bahia in 1989 (Pereira et al. 2006), purportedly after the deliberate introduction by bioterrorists of inoculum from the st ate of Rondonia (Junior 50

PAGE 51

2006). The rapid spread of witches broom in Bahi a resulted in a dram atic decrease in national production, from 378,000 tons in 1990 to less than 120,000 tons in 1999 (Gray 2001). Losses due to frosty pod are no less staggering (Evans et al. 1998). Frosty pod spread from Ecuador and Colombia in the early 1900s south into Peru (Evans et al. 1998), and north to Belize and southern Mexico (Phillips-Mora et al. 2006a,b, 2007), but is currently restricted to the west of the Andes, save for the Napo province of Ecuador (E vans 1981a, Phillips-Mora et al. 2007). and are basidiomycetes in the Marasmiaceae, a family in the Marasmioid clade in the Agaric ales (Aime & Phillips-Mora 2005, Matheny et al. 2006). As early as the 1970s, similarities between these two pathogens we re recognized (Evans et al. 1978, Evans 1981a, Evans et al. 2002). Both f ungi are indigenous to South America, have similar host ranges, are hemibiotrophs, and influen ce the hormonal balance of their hosts. Due to these similarities, a close relationship between the two fungi was proposed and recent molecular data has confirmed this hypothesis. (= ) was originally described as by Stahel (1915), and late r transferred to the genus (Singer 1942) due to the presence of l ong, thick-walled pileal hairs. Aime and Phillips-Mora (2005) investigat ed the molecular phylogeny of and with DNA sequences from five loci. They found that they were sibling species within the Marasmiaceae, and that was more closely related to than to other species. Thus, it was tr ansferred to the genus The name witches broom refers to the charac teristic proliferation of swollen and twisted shoots that develop on affected cacao trees, sympto ms that were first described by Went (1904) in Surinam. Cacao pods and flower cushions (w oody meristematic tissue above leaf scars; de Almeida and Valle 2007) may also be affected, resu lting in large yield losses. The sole inoculum 51

PAGE 52

for inf ection is basidiospores that are released by basidiocarps that ar e produced on necrotic, infected tissue. Singers (1976) monograph of th e Marasmiaceae indicated that basidiocarps had a crimson red pileus that bleached to whitish with a crimson spot in the center, and was 5-15 mm in diameter. Pileocystidia were 80-150 x 412 m, pseudoamyloid (i.e. dextrinoid) in Melzers reagent, had a red thick wall, and we re rounded-obtuse at the tip. Lamellae were white and distant (1-2 mm). Stipes were 5-10 x 0.40.7 mm and dark-brown red at the subbulbous base to white at the apex. Basidiospores were 7-11 x 4-5 m, ellipsoid, hyaline and smooth, and basidia were 31 x 8 m. Che ilocystidia were bottle-sh aped and 35-50 x 9-14 m. Morphological and pathological di versity has been studied in Basidiocarp variation was first described by Pegler (1978), who reported three varieties: a) var. which was identified in Trinidad and Surinam and had a pileus that was crimson in the center and faded to white at the margins; b) var. identified in Ecuador and with a uniformly red pileus; and c) var. from a single broom in Ecuador and with a yellow pileus. In addition to morphologically based vari etal classification, the pathogen has also been divided into biotypes based on host range. Five biotypes have been described; the: a) C-biotype affects species in the genera (including cacao) and (Malvaceae) (Evans 1981); b) S-biotype affects and (Solanaceae) (Basto s and Evans 1985, Pereira et al 1997, Rincones et al. 2006); c) B-biotype affects (Bixaceae) (Bastos and Anderbrhan 1986); d) L-biotype includes presumed saprotrophs on lianas (Bi gnoniaceae) (Evans 1978, Pegler 1978, Griffith and Hedger 1994b); and e) H-biotype affects (Malpighiaceae) (Resende et al. 2000). 52

PAGE 53

-type basidiocarps were first reported on lianas and other nonidentified hosts in 1949 along the Napo Rive r in Ecuador (Desrosiers and von Buchwald 1949). Evans (1978) made the first collections of th e L-biotype in Ecuador, and showed them to be nonpathogenic on cacao. The L-biotype is associat ed with dead or living liana vines, and due to limited symptom development on cacao (Evans 1978, Hedger et al. 1987, Purdy and Dickstein 1990), it is assumed that this biotype is either a non-pathogenic parasite or a non-host specific saprotroph (Hedger et al. 1987). Griffith and Hedger suggested that the L-bi otype is host-specific to (Bignoniaceae) (1994b), but basidio carps have since been found on other liana species (R. Barret o, personal communication). was first identified as a cau se of witches broom on the solanaceous hosts in Brazil, first on and by Bastos and Evans (1985), and later on and by Pereira et al. (1997) S-biotype isolates have been shown to be non-pathogenic on cacao, but have caused broom symptoms and necrosis on tomato plants ( ), and shortened internode s and galls on bell pepper ( ) (Bastos and Evans 1985). The B-biot ype was originally reported from (Bixaceae) (Bastos and Anderbhran 1986), but was shown to be genetically identical to C-biotype isolates, and has not been widely reported since. Therefore, its designation as a distinct biotype is questi onable (Meinhardt et al. 2008b). The most recently described biotype was found causing brooms on (Malpighiaceae), and is reported to be pathogenic on cacao (Resende et al. 2000). Another malpighiaceous host has been reported, cf. and isolates from this host are also pathogenic on cacao (Bastos et al. 1998). de Arruda et al. (2005) compared the morphology of C-biotype (nine collections), S-biotype (one collection), and one collection from 53

PAGE 54

They determined that the collection produced slightly la rger basidiospores (average 12 x 6.5 m) and obpyriform, lageniform to mucrona te cheilocystidia, comp ared with the other collections (10-11.5 x 5.5-6.5 m basidiospores with sub-cylindrical, lageniform to obclavate cheilocystidia). Based on these morphologica l characters and the resolution of the isolate outside of the other isolates in an ITSbased phylogeny, de Arruda et al. (2005) erected a new species name for the H-biotype, Arruda, G.F. Sepulveda, R.N.G. Mill., M.A. Ferreira & M.S. Felipe A fuller systematic analysis with more isolates from the various biotypes is needed to reso lve species level relationships in Mating studies indicated that th e Cand S-biotypes are homotha llic, but that the L-biotype is heterothallic, with a complex bifactorial, multiple allele mating system (Griffith and Hedger 1994a,b). Whereas single spore, uninucleate cultures of isolates of the C-, Band Sbiotype quickly developed clamp connections and binucleat e mycelia, isolates of the L-biotype persisted as uninucleate cultures without clamp connections (Griffith and Hedger 1994a). Molecular data have been used in severa l studies to investigate relationships among biotypes. A RAPD study with eight isolates (including one S-, one B-, and six C-biotype isolates) suggested that an isolate of the S-biotype was genetically di stinct from isolates of the Band C-biotypes (Anderbrhan and Furtek 1994). More recently, RFLP analyses of rDNA and mtDNA were largely unable to distinguish the Sand Cbiotypes, and found only one polymorphism for separating the H-biotype from the other biotypes (de Arruda et al. 2003a). When an almost identical set of isolates was analyzed with ERIC-PCR fi ngerprints, the C-, S-, and H-biotypes were distinguished (de Arr uda et al. 2003b). Rincone s et al. (2007) used microsatellite-primed PCR (mp-PCR) to inves tigate relationships between the C-, Sand L54

PAGE 55

biotypes. Although genetic divers ity was revealed within each biotype, the authors did not speculate on relationships among biotypes or specific lineages in the species. Only two exam ples of phylogenetic analyses of with DNA sequence data were found in the literature. de Arruda et al. (2003a) completed a pars imony analysis of the intergenic spacer (IGS) region of 14 isolates, including 10 isolates from two isolates from and two isolates from The three biotypes formed well-supported clades, one of which contained isolates from and and another that cont ained isolates from In a similar study using the ITS1-5.8S-ITS2 region, de Arruda et al. (2005) again found that an isolate from was distantly related to isolates from and All isolates resolved as a sister group to var. which was also supported in the multi-gene analys es of Aime & Phillips-Mora (2005) The isolate from was sister to a largely unresolved clade that contained all other isolates of A multiple regoion phylogenetic analysis comprise d of isolates from all biotypes and hosts has yet to be completed, and is needed to reso lve the conflicting evidence regarding relationships among biotypes and host-specific lineages. Clarificat ion of these relations hips would help to: estimate the threat that subspecific groups pose to the cacao crop; indicate whether as it is currently de fined, should be considered a si ngle species or species complex; and assess the utility of host pl ant associations to designate biotypes. In the present study, isolates of from a range of hosts in Brazil, many of them newly reported for the pathogen, were analyzed with isolates from and several South American countries. Three DNA regions were chosen for phylogenetic analyses: ITS, IGS, and the RNA polymerase 55

PAGE 56

II large subunit (RPB1). In addition, macroand mi croscopi c morphological characters were examined in a subset of these isolates to de termine if they reflected genetic relationships. Isolates of that were used in this study we re collected previously by several individuals and are stored in th e isolate collection of RCP in Homestead, FL (Table 2-1). All CPB isolates were collected from various states in Brazil by H. Evans and R. Barreto in MarchJune of 2006, and they represent the bulk of isolates collected from hosts other than Overall, 16 isolates of the C-biotype, 10 of the S-biotype, five of the L-biotype, one of the Hbiotype, and three from other host plants were studied. Isolates of the C-biotype were from Bolivia, Brazil, Colombia, Ecuador, Venezuela, Tr inidad, and Tobago; all other isolates were collected in Brazil. Isolate TC P7 was originally received as but phylogenetic analysis showed it to be consis tently distinct, therefore it wa s considered as an additional sp. isolate. Three taxa that are sister to were included in analyses: in the ITS, IGS and RPB1 analyses; var. in the ITS analysis; and sp. MCA 2500, a grass endophyte, in ITS and RPB1 analyses. For the IGS analysis, and var were used as outgroup taxa. which is sister to (Aime and Phillips-Mora 2005), was included as an outgroup taxon in the ITS and RPB1 analyses. Isolates of were grown at room temperature on PDA covered with a layer of sterile cellophane. After colonies had grown to at least 6 cm in diameter, mycelial mats were peeled from cellophane with sterile forceps, trip le rinsed with sterile deionized water and dried 56

PAGE 57

on sterile filter paper. DNA was extracted usi ng a DNA genomi c preparati on protocol from the University of Wisconsin Biotechnology Center. Briefly, mycelia were ground in 500 l Shorty DNA Extraction Buffer (0.2M Tris-HCl, pH 9.0, 0.4M LiCl, 25mM EDTA, 1% SDS) and incubated at 68C for 10 minutes. Tissue was centrifuged at 14,000 rpm for 5 minutes, 400 l of the supernatant was tr ansferred to new tube and the DNA was precipitated with 400 l 99% isopropanol. The tubes were centrifuged for 10 minutes at 14,000 rpm and the supernatant was decanted. The DNA pellets were air-dried for 5 minutes then resuspended in 400 l TE buffer (10 mM Tris-Cl, pH 7.5, 1 mM EDTA) for 30 minutes at room temperature. 2l DNA was used in PCR reactions. DNA sequence data from three nuclear regions were used in phylogenetic analyses: two rDNA sequences (the ITS1-5.8S-ITS2 and the in tergenic spacer, IGS), and a protein-coding gene, the RNA polymerase II large subunit, RPB1. In studies by de Arruda et al. (2003a, b), the IGS region was shown to have a stronger phylogenetic signal than ITS. Both ribosomal regions were included to provide two le vels of resolution. A lthough it is generally more conserved than the ITS and IGS, the RPB1 locus was included as a protein-coding nuclear ge ne that is not linked to the ITS and IGS. All PCR reactions were carried out in 50 l reactions that contained 38.25 l of sterile distilled, deionized water, 6.5 l ThermoPol Re action Buffer (New England Biolabs, Ipswich, MA), 1 l 10 mM dNTP mix (New Engla nd Biolabs, Ipswich, MA), 0.25 l Taq DNA polymerase (conc. 5,000 units/ml), 1 l each of 15 M primers and 2 l DNA template. Standard cycling parameters with a 55 C annealing temperature were used. The ITS region was amplified using the following primers: ITS1 (G ardes and Bruns 1993) and ITS4 (White et al. 1990). The IGS rDNA region was amplified usin g primers CNL12 (Appel and Gordon 1995) and 57

PAGE 58

O-1 (Duchesne and Anderson 1995). The RPB1 locu s was amplified with primers RAS.RPB1-F2 and RAS.RPB1-R2 (Aim e and Phillips-Mora 2005) Direct sequencing of the PCR products was performed at the University of Florida I CBR (Interdisciplinary Center for Biotechnology Research) Facility (Gainesvil le, FL) with the same primers used for amplification. For ambiguous bases, the base that agreed with the consensus sequence was retained. Because sequencing signal from direct sequencing was st rong, cloning was not performed. For isolates with unexpected phylogenetic placements, se quences were confirmed by repeated DNA extraction, PCR reactions and direct sequencing. Sequence alignments were done using CLUS TALX (Thompson et al. 1994) in Mega4 using default parameters, and adjusted manually. In the IGS data set, a single indel of 170bp present only in isolate TCP7 was excluded from all analyses. For the maximum parsimony analysis, two data sets were used: one which in cluded indels but with indel sites treated as missing data, and a second in which indels were coded as separate characters following a coding system modified from Simmons and Ochoteren a (2001). Indel sites were removed from the alignments, and indels were coded as unordered, numerical characters. Single site indels were coded as binary presence/absence characters. Conti guous indels that did not share 5' or 3' termini were considered as separate mutation events, and were coded as separate presence/absences characters. Contiguous indels that share a 5' or 3' terminus were coded as unordered multistate characters. Because no gaps in the data set we re imbedded (one gap falls completely within another contiguous gap), step matrices (as out lined by Simmons and Ochoterena 2001) were not utilized. A microsatellite region in the IGS locus (CAAA)x was also coded as an unordered, multistate character reflecting the number of rep eats of the microsatellite. Coded indels were excised from the data set, unless the nucleo tide sequence among isolates not containing the 58

PAGE 59

deletion contained inf ormative sites, in which case the indel was left in the data set but gaps were coded as missing data. Alignments were submitted to TreeBASE (http://www.treebase.org ) Both maximum parsimony (MP, using PAUP* v.4.0b; Swofford 2000) and maximum likelihood (ML, using GARLI v.0.946) analyses were done on all data sets. Both data sets for each locus (indels treated as missing data or indels coded as separate characters) were used for MP analysis. Parsimony analysis was conducted using heuristic searches with 100 random addition replicates, with tree bisection reconnection (TBR) swapping, saving no more than two trees with tree scores greater than five per repetition. Consistency index (CI) and retention index (RI) values were calculated by PAUP*. To determine statistical support for groups in the phylogenies, the non-parametric bootstrap test (Felsenstein 1985) was performed using 1,000 repetitions and heuristic search criteria as described above except that 10 random addition replicates were used, saving no more than two trees with scores greater than five per repetition. ML analysis was run using the data set with gaps treated as missing data only. Analysis was run on GARLI v.0.946, using default parameters and a randomly generated starting topology. Analysis was stopped after 5x106 generations or a 0.01 decrease in ML score. Three independent runs were performed and resul ting topologies compared in TreeView v1.6.6. If similar topologies were generated, bootstrap anal ysis was performed using the same parameters with 1,000 repetitions. Tree files were imported in to MEGA 4.0 for visualization and editing of phylograms. For the ITS and RPB1 analyses, was the outgroup; for the IGS analysis, C21 and Dis116a were outgroups. In order to determine if combining data from the three loci was appropriate, the level of congruence was first tested betw een pairs of the three data se ts, testing either congruence between character sets or between topologies of phylogenetic trees (Seelanen et al. 1997, 59

PAGE 60

Johnson and Soltis 1998). Overall incongruence b etween data sets was quantified using the Mickevich-Farris index (IMF), which calculates a standardized value for the increase in the tree length of the most parsimonious tree from a combin ed analysis compared to the sum of the tree lengths of the two loci analy zed individually (Mickevich and Farris 1991). To determine the significance of the quantified incongruence, a con catenated data set was constructed with all three loci, and partitions were defined for each locus. The partition homogeneity test was implemented in PAUP* comparing all three lo ci, to determine the incongruence length difference (ILD) between loci pairs (Farris et al1994). -values less than 0.05 contain significant incongruence between the data sets. The ILD test is known to be very sensitive (Yoder et al. 2001; Darlu and Lecointre 2002), so the decision as to whether to combine data sets was largely based on examination of topology and bootstrap support. To assess topological congruence, an approach similar to that used by Seelanen et al. (1997) was followed. First, consensus trees for each of the three loci were compared empirically to determine if instances of t opological incongruence were hard (isolates resolve in different well-supported clades in each topology), or soft (incongruence due to weak phylogenetic signal in one of the data sets). To determine the stat istical significance of t opological incongruence, we used 70% bootstrap trees from one region as a topological constraint in analysis of each additional region, and the Wilcoxon signed-rank (WSR) test to dete rmine if the number of steps that were gained or lost in the re sulting tree was significantly greater ( <0.05) than the original trees, an approach that was first described by Templeton (1983). In addition, strongly supported clades that were resolved in each single locus anal ysis were used as constraints for the other loci (Table 2-4). This approach allowed us to test th e congruence of specific clades in each data set, enabling the identification of individual clad es and/or accessions that cause incongruence 60

PAGE 61

between data sets. This approach allowed congruence of specific clades to be tested in each data set and enab led the identification of individual clades and accessi ons that caused incongruence. This approach helps determine if data sets co uld be combined and provides information about evolutionary events that may have caused incongruence, such as cryptic outcrossing, hybridization, introgre ssion, or horizontal gene tran sfer (Wendel and Doyle 1998). Basidiocarps of isolates CPB1, CPB 2, CPB5, CPB9, CPB15 and TRDC74 were successfully produced using the bran media cookie method (Griffith and Hedger 1993, Niella et al. 1999). Preserved basidi ocarps of CPB10 and CPB12 we re provided by H. Evans. Fisherbrand 500ml polypropylene scre w top jars were filled two-thir ds full with bran media of 400 g vermiculite, 500 g wheat bran flakes cereal, 120 g CaSO4. 2H2O, 15 g CaCO3 and 1200 ml de-ionized water and autoclaved for 1 h at 0, 24 and 72 h, then seeded with mycelial plugs of isolates. Once media was fully colonized, it was covered with a sterile casing of 200g mown turf, 50 g vermiculite and 50 g CaCO3. After the casing was colonized, cookies were removed from plastic jars, cut into quarters to maximize su rface area for basidiocarp production, and hung in a Plexiglass closet under a 12 hr lig ht/dark cycle and and high humid ity; cookies were misted daily with de-ionized water. Basidiocarps were harvested 1-2 days after opening, and dried (50 C for approx. 20 min) for subsequent micromorphological characterization. For one to 10 ba sidiocarps of each isolate, the following macromorphological ch aracteristics were recorded: pileus diameter and color; presence of lamellulae; lamellae spacing and color; and stipe color, length, and diameter at base and tip. The following microscopic characters were examined for two basidiocarps per isolate: pileocystidium pigmentation, shape and dimensi ons (15 cystidia per fruiting body); basidium 61

PAGE 62

size and number of sterigmata (15 basidia per fruiting body); cheilocystidium shape and dimensions (15 cystidia per fr uiting body); and basidiospore shap e and dimensions (30 spores per fruiting body). In addition, tissue was staine d with Melzers reagent and pseudoamyloid reactions were recorded. Mean length, width, and length/width ratios of basidiospores were separated using Fishers LSD te st (PROC GLM in SAS v. 0.1.3; SAS Institute Inc., Cary, NC). Coding indels as separate unor dered characters increased re solution and bootstrap support for the majority of nodes in the ITS analysis, and approximately half of the nodes in the IGS analyses. Thus, indels may be important in the ev olutionary histories of these loci (Table 2-2). The indel characters in the RPB1 locus only ma rginally improved statis tical support for nodes, and most separated the outgroup taxon from the isolates. The ITS dataset contained 40 accessions, including one of two of var. and two of sp. isolates. was used as an outgroup. There were 678 total characters (35 in del characters), of which 122 were parsimonyinformative. Within isolates of there were 22 parsimony informative characters (PIC). Maximum parsimony (MP) an alysis resulted in 18 most pa rsimonious trees (MPT), with a tree length of 391 (CI=0.903, RI=0.893) (Figure 21). The maximum likelihood (ML) tree had a ln score of -2472.65, and largely similar topolo gy to the 50% consensus tree from parsimony analysis, except that clade 2 was not completely resolved in ML (Figure 2-2). Both ML and MP analyses resolved isolates of as a highly supported sister clade to Isolates of were resolved into three main clad es in both trees (Figures 2-1 and 22). In the MP phylogeny, Clade 1 resolved into two subclades. Subclade 1a contained all isolates from and isolate 73-62, collected from an unide ntified solanaceous weed (the species 62

PAGE 63

was not identified). Subclade 1b contained most isolates from solanaceous hosts, and also included a broom isolate from a liana, CPB 17, and a C-biotype isolate, CPB2, from Subclades 1a and 1b were resolved in the 50% parsimony consensus tree, but did not have high bootstrap support. Examination of character state cha nges in the analysis revealed that only three and one synapomorphies supporte d the resolution of the 1a and 1b subclades, respectively. Clade 1 is sister to Clade 2 (59% in MP), which contains three isolates, CPB5 ( ), TCP8 (unknown host), and CPB8 (bignoniaceous liana). It had 55% bootstrap support in the MP analysis, but isolates CPB5 and TCP8 were sister to each ot her with higher bootstrap support in both MP and ML analyses (60% and 62%, respectively). Clade 3 contained two liana isolates, CPB7 and CPB10, and three isolates fr om malpighiaceous shrubs and trees, CPB6, CPB9, and CPB12. It was highly supported in both analyses (90% MP, 88% ML), and was supported by 17 synapomorphies. The IGS dataset included 38 accessions of and C21 and var. Dis116e were used as outgroup taxa. Th ere were 861 total characters (41 indel characters) and 300 PIC, 68 within isolates of MP analysis resulted in 200 MPT, with a tree length of 475 (CI=0.931, RI=0.934) (Figur e 2-3). The ML tree had a ln score of 3023.85, and similar topology to the 50% consensu s tree from MP analysis, except for the placement of Clade 2 (Figure 2-4). The resulti ng phylogenies resolved with high support the three main clades: Clade 1 contained all C-biotype and S-biotype isolates, without subclade resolution; Clade 2 contained CPB5, CPB8, a nd TCP8; and Clade 3 contained CPB6, CPB10, and CPB12 (Figures 2-3 and 2-4). The placement of Clade 2 differed from that in the ITS phylogeny, which placed it sister to Clade 1. Isolates CPB7 and CPB9 resolved in Clade 1 in the IGS phylogeny rather than Clade 3, where they were placed in the ITS analysis. Within Clade 1, 63

PAGE 64

they were part of a large polytom y that contained all other Clade 1 isolates. This was an example of hard incongruence between th e two loci (Seelanen et al. 19 97). Support for Clade 1 was high (95% MP), and the group shared only two syna pomorphies. Support for Clade 2 (isolates CPB5, CPB8, and TCP8) was high in both ML (92%) and MP (83%) analys es and the three isolates shared six synapomorphies. Cl ade 3 contained only CPB6, CPB12, and CPB10, and was highly supported (100% MP, 99% ML). The RPB1 analysis included 36 accessions of and single accessions of (C21), var. (Dis116a), and sp. (MCA 2500). was used as an outgroup. There were 747 total characters, and 58 PIC. Within there were only 12 PIC. This locu s is largely conserved, hence the weak phylogenetic signal it gave among closely re lated taxa. MP analysis resulted in 200 MPT with a tree length of 230 (CI=0.948, RI=0.911) (Figur e 2-5). The ML tree had a ln score of 2124.22, and a topology that was largely similar to the 50% consensus tree from parsimony analysis (Figure 2-6). The ML tree showed less resolution than the MP phylogeny, with no support for resolving Clade 2. There was less reso lution in the RPB1 phylogeny compared to the other loci, likely due to the low number of parsimony informative sites. Subclades 1a (one synapomorphy) and 1b (two synapomorphies, and c ontained most S-biotype isolates and CPB17) had 64% MP and 87% MP, and 84% ML bootstrap support, respectively. Isolate CPB7 (which resolved into Clade 3 in the ITS phylogeny, and Clade 1 in the IGS phylogeny), was sister to subclade 1b, with moderate to high support (8 7% MP, 79% ML). The other incongruent taxon, CPB9 (Clade 3 in ITS, Clade 1 in IGS), resolved in Clade 3 in the RPB1 phylogeny. CPB8 (Clade 2 in ITS and IGS phylogenies) resolved in Clade 3, with 87% MP and 82% ML bootstrap support. 64

PAGE 65

Partition homogeneity analysis in P AUP revealed significant incongruence ( =0.002) among all data set comp arisons (Table 2-3). To identify the source of in congruence, we tested several topological constraints on each data set, with WSR tests to estimate the significance of incongruence of alternate topologies (Table 2-4). In general, enforcing RPB1 clades onto the ITS and IGS data sets resulted in significant incongruence ( <0.05), with the exception of RPB1 subclade 1b. Alternatively, enforcing ITS and IGS-supported clades onto the RPB1 data set resulted in less significant incongruence. This is likely due to fewer parsimony informative characters in the RPB1 data se t; thus, part of the resulting phy logenetic signal may be due to chance. Clade 2 (CPB5, TCP8, and CPB8) was comp letely congruent in the IGS and ITS data set, and led to a four-step increase in the RPB1 phylogeny ( 0.0455). Clade 3 (CPB6, CPB10, and CPB12) was congruent in all data sets. The addition of CPB7 and CPB9 (which were resolved in Clade 3 in the ITS set, but not in the IGS set) made the clade significantly incongruent in the IGS set ( 0.0010), but not in the RPB1 set (0 length increase). Based on the results of the topological incong ruence analyses, it was concluded that most of the incongruence between data sets was due to the placement of CPB7 and CPB9, as well as the low phylogenetic signal in the RPB1 set. Other instances of incong ruence occurred in the placement of isolates in subclades 1a and 1b, and again could be due to a weak phylogenetic signal as the resolution of these two clades was only supported by one or tw o synapomorphies in each data set. Because significant incongruence existed among all data sets, data were not combined for phylogenetic analysis. 65

PAGE 66

Basidiocarps were produced for four isol ates from Clade 1 (CPB1, TRDC74, CPB2 and CPB15), one from Clade 2 (CPB5), and isolate CP B9. In addition, preserved basidiocarps of two Clade 3 isolates (CPB10 and CPB12) from original host material were examined. The examined characters largely agreed with Singers description of the species (1972) (Tables 2-5 and 2-6). Pileus diameter generally ranged from 7-25 m, with larger pilei produced by CPB5 (Clade 2), which ranged 11-32 m. Pileus color was crimson, and faded to crimson at the umbo to cream at the margins as basidiocarps aged, except in alcohol-p reserved basidiocarps of CPB12, in which pigment had leached to result in a uniform cream color. Isolates from Clade 3 (CPB10, CPB12) and CPB5 had significantly larger basidiospores (mean size of 10.7x6.0 m, 10.9x5.8 m, and 10.5x5.4 m, respectively) than isolates from other clades (generally 9-9.5x5-5.3 m) (Table 27). Isolate CPB9 had smaller basidiospores, 8.4x4.7 m. Cheilocystidial shape, wh ich de Arruda et al. (2005) indicated distinguished the Hbiotype, was variable, even between basidiocarps of the same isolat e. CPB5 and isolates in Clade 1 tended to have numerous lageniform (bottle-shaped) cheilocystidia in older basidiocarps (collected on 2nd day after opening) (Figure 2-7). Among isolates in Clade 3 (CPB10, CPB12) and CPB9, cheilocystidia were rare and had an obclavate to pyriform shape. Pileocystidium pigmentation was cytoplasmic and often produced as extracellular incrustations. Pileocystidia were produced in a trichodermal layer, and were mostly clavate to cylindrical, but sometimes mucronate (Figure 2-7) and clamped. CPB5 had distinct pileocystidia, in that the first cell was swollen clavate, with a largel y swollen second cell (Figure 2-7). Pigment was concentrated in the tip of th e apical cell. Pileocystidia of a ll isolates were generally shorter than those reported by Singer (1976). 66

PAGE 67

The multilocus phylogenetic analys is presented he re is the most complete analysis for to date. It included isolates from four of the five reported biotypes; the fifth, Bbiotype is thought to be c onspecific with the C-biotype (Meinhardt et al. 2008). Although Hedger et al. (1987) suggest ed that isolates of the Lbiotype were saprotrophs or nonpathogenic endophytes, thei r isolates were from (Bignoniaceae). Interestingly, H. Evans and R. Barreto (unpublis hed) have collected isolates from brooms on malpighiaceous lianas in Brazil, and one of these isolates, CPB17, was examined in the present work. Thus, there may be pathogenic lineages of on lianas in the Malpighiaceae, but not the Bignoniaceae. More work is needed to distinguish saprophytic and pathogenic populations of that reside on tropical vines. Morphological differences were slight among th e isolates that were examined, and do not appear to distinguish subspeci fic groups. However, subspecifi c lineages were resolved in phylogenetic analyses. Clade 1 contained most isolates, including all from and hosts, as well as several from lianaisolates, including CPB17, which was rec overed from a broom. Isolates from all had similar sequences, and resolved into a single subclade (1a) with low bootstrap support (40-60%) in the ITS and RPB1 analyses. Isolates from solanaceous hosts (subclade 1b) were resolved as a sister to subcla de 1a in the ITS and RPB1 analyses, with low to high support (49 and 87%, respectively). CPB2, CPB15, and 92-10-7 had equivocal resolution in Clade 1, in that they were alternately placed in subclade 1a or 1b in the single locus phylogenies. Subclades 1a and 1b may be distinct, but recently diverged lineages in This hypothesis is supported by low bootst rap support and the limited numbe r of synapomorphies that resolved the and subclades. 67

PAGE 68

Isolate TCP7, sp., was distantly rela ted to isolates of although it was more closely related to than the frosty pod pathogen, TCP7 was collected from a liana in 2001 in Brazil. Based on the phylogenetic evidence presented in this study, it could be considered a separate species, but fruiting bodies need to be examined for a complete taxomonic assessment. Isolate CPB5 from CPB8 from a bignoniaceous liana, and TCP8 from an unknown host resolved into Clade 2 in the IT S and IGS analyses, but not in the RPB1 phylogeny. In the RPB1 analysis, CPB8 fell into Clade 3, and CPB5 into subclade 1a. Hard incongruence in the RPB1 phylogenies may be du e to insufficient sampling, but the jumping of so many isolates between Clades 1 and 2 may also indicate that the genetic isolation between these lineages is not absolute and that outcrossing occurs between lineages. Further taxon sampling may increase the collection of isolates in this clade, which could provide more information about its ecology. Clade 3 contained malpighiaceous isolates CPB6, CPB10, and CPB12, as well as CPB7 and CPB9 for at least one locus. It was well re solved as sister to Cl ades 1 and/or 2 in all analyses, and could be considered a separate taxon. There was little morphological difference between isolates of Clade 3 and Clades 1 and 2, but isolates CPB10 and CPB12 had slightly, but significantly larger basidiospores, and slightly differently shaped cheilocystidia. Our observation of obclavate to pyriform cheilocystidia produced by malpigheaceous isolates contradicts that of de Arruda et al. (2005), who i ndicated that the H-biotype ( ) produced lageniform cheilocystidia However, at least one study has sh own that cheilocystidia can be morphologically variable, even within a single pileus and depe nding on basidiocarp age, humidity, and other environmental factors (Aime 2001). 68

PAGE 69

The phylogenetic analyses presented here larg ely agree with previ ous studies comparing populations of different biotypes us ing mo lecular markers, in that the Sand C-biotypes were genetically distinct (de Arruda et al. 2003a,b, Rincones et al. 2007). Assuming that the isolates that were included in this study were correctly labeled with respect to host, the revealed phylogenetic lineages are pathogens of or malpighiaceous hosts with a few exceptions: a liana isolate (CPB17) and isolate (CPB2) fell in the clade (1b), and a isolate (73-62) fell in the clade (1a). Although the Cand Sbiotypes appear to define biologically meaningful subspecific populations of the fungus, the present results indicate that the Land Hbiotypes as they are currently defined are not valid. Biotypes of should be redefined once the hos t ranges of these populations are better understood. The host specificity of the different biotypes is of practical importance to cacao growers. As cacao cultivation expands in Brazil, ther e is concern about introducing the crop where is found on solanaceous hosts. Previous pathogenicity tests indicated that host specificity exists among biotypes, in that cross inoculations of the Lor S-biotypes onto yielded no symptoms or only slig ht swelling of inoculated node s (Evans 1978, Evans and Bastos 1985). The present phylogenetic analyses indicate that these biotypes may be host specific, but that specificity may not be abso lute. The extent to which phylogene tically distinct isolates of would cause disease on cacao needs to be better understood. In addition, virtually nothing is known of the potential for host range expansion to occur in this fungus via gene flow. Although a homomictic, homothallic breeding strategy prevails in incongruent phylogenies for several isolates suggests that hybridization or outcrossing may have occurred among the lineages. 69

PAGE 70

is a relatively new genus that was erected in 1978 to accommodate a single sp ecies, (Evans et al. 1978). It remained a monospecific genus until 2005, when was transferred to (Aime and Phillips-Mora 2005). While both species include cacao pathoge ns and are hemibiotrophs, their respective biologies exhibit striking differences. was erected to accommodate an anamorphic basidiomycete, which does not produce basidiocarps. Evans et al. (2002) recently s uggested that the conidia of are actually meiospores. Aime and Phillips-Mora (2005) conjectured that other species of exist as nondescribed biotroph s and that many species of currently placed in Section Iopodinae are likely congeners with Based on molecular phylogenies, it appears that at least seven other species of are now known: a grass endophyte, isolate MCA2500 (Aime and Phillips-Mora 2005), us ed in this study; a lianaassociated species, isolate TCP7, used in the present study; (Aime unpubl.); and four species (three transferred from to and one new species) from Southeast Asia (Kerekes and Desjardin 2009). It is possible that isolate TCP7 is a previously described species in section Iopodinae, but ba sidiocarps need to be obtained and studied for this isolate to determine its affinity wi th other species. Because the lineages that were resolved in the present study may have out crossed and lack distinctive morphological differences, they should remain subspecific groups of The Sand Cbiotypes are homothallic and th e L-biotype heterothal lic. It has been proposed that the divergence of lineages within the species was accompanied by a switch from a heterothallic to homothallic mating strategy (Griffith and Hedger 1994a,b). Homothallism can be advantageous for plant pathogens. With no enfo rced outcrossing, populations can become highly 70

PAGE 71

specialized to a given environm ent or lifestyle. T here is also no need for the presence of compatible individuals to complete a life cycle. The classic hypothesis of the evolution of mating strategies is the evolution from homothallism to heterothallism (Whitehouse, 1949). Phylogenetic an alysis of several species supported a single switch from homothallism to heterothallism (Geiser et al. 1998), and examination of mating type gene sequences in several species suggest that the switch was caused by the loss of mating genes (Galagan et al. 2005). However, other homothallic fungi appear to be derived from heterothallic taxa. Homothallic species of often contain both mating type loci, but their function is not known (Kronstad and Staben 1997). These loci may be artifacts from a heterothallic ancestor. The occurrence of homothallic species at the geographic limits of the genus also suggests that homothallism may have evolved where opposite mating types were not present. Phylogenies of homothallic and heterothallic and species support a single switch from heterothallism to ho mothallism in both genera (Pogge ler et al. 1999). A shift from heterothallism to homothallism has also been proposed for (Yun et al. 1999) and (ODonnell et al. 2004). In ,there is evidence that homothallism evolved not once, but several times (Turgeon et al. 1998 ). A similar situation could be proposed for where pathogenicity and homothallism e volved to enable the survival of these specialized populations. L-biotype is olates from Ecuador that were reported to be heterothallic and saprotrophs or endophytes were not included in this study. It is possible that they represent another species of that differs from the pathogenic It has been speculated that both and evolved from a common forest endophyte (Evans 1981, Griffith et al. 1994, Evans et al. 2002), and that co71

PAGE 72

evolved with and on the eastern side of the Andes (Pound 1943, Baker and Holliday 1957). However, Evans (1981) and then Gr iffith et al. (1994) p roposed an alternative hypothesis to coevolution with cacao, suggesting that the homothallic Cand Sbiotypes evolved from the heterothallic L-biotype in separate dive rgence events (Griffith et al. 1994). They studied the mating systems of the L-, S-, and Cbiotype s and determined the L-biotype to be more diverse than the other biotype s, based on banding patterns of mtDNA and isozyme studies. They also noted that the S-biotype profiles were more similar to th e L-biotype than the C-biotype. They also speculated that additional biotype s would be found that would represent older divergences from a heteroth allic endophytic ancestor. Recent evidence, including the phylogenies constr ucted in this study, seems to support an evolutionary scenario similar to th at proposed by Griffith et al. (1994) Recent collections of in the Brazilian Amazon have uncovere d new host species in the Malpighiaceae, which in the present analyses fall in Clade 3 with bignoniaceous liana isolate (L-biotype) CPB7. Thus, there is less phylogenetic support fo r distinct Hand L-biotypes than for a malpighiaceous clade that affects that fam ily and contains bignoniaceous saprobes. The present phylogenies also support the s uggestion by Hedger et al. (1994) that the and solanaceous lineages diverged more recen tly compared to other clades. Lineages that diverged earlier have mostly malpighea ceous and bignoniaceous hosts. Interpretation of these phylogenies suggests that the homothallic, highly specialized pathogen could have evolved from a heterothallic, saprotrophic or endophytic ancestor, and subs equent host jumps resulted in the lineages presently seen in as a pathogen on cacao therefore likely represents not a pathosystem involving coevolution, but a series of host jumps to plants in several different families. As more host species are identified in the Amazon, and more 72

PAGE 73

73 taxa are described and added to phylogenetic studies, the evolution of both and should become clearer.

PAGE 74

Table 2-1. Description of isolates used in the study. Isolate Collector/Donator Host origin Collection tissue Geographic origin Biotype Genbank Accession numbe r ITS IGS RPB1 4. M. Shaw Broo m Colombia C G Q 919116 G Q 919081 G Q 919152 M. Shaw Broo m Venezuela C G Q 919117 G Q 919080 G Q 919153 7. M. Shaw Broo m Colombia C G Q 919118 G Q 919082 G Q 919156 17. M. Shaw Broo m Colombia C G Q 919119 G Q 919083 G Q 919154 28. M. Shaw Broo m Trinidad C G Q 919120 G Q 919084 G Q 919155 31. 46PR H. Purd y Broo m Ecuador C G Q 919137 G Q 919085 G Q 919157 73-31 M&M/Penn State Broo m Brazil C G Q 919121 G Q 919094 G Q 919158 73-62 DIFIP/Penn State Weed, Solanaceae Broo m Brazil S G Q 919122 G Q 919108 G Q 919159 92-10-7 ACRI/PennState Weed, Solanaceae Broo m Brazil S G Q 919123 G Q 919096 G Q 919160 CPB1 H. Evans/ R. Barreto Broo m Brazil C G Q 919141 G Q 919092 G Q 919161 CPB2 H. Evans/ R. Barreto Broo m Brazil C G Q 919124 G Q 919101 G Q 919162 CPB3 H. Evans/ R. Barreto Frui t Brazil S G Q 919125 G Q 919102 G Q 919163 CPB4 H. Evans/ R. Barreto Broo m Brazil S G Q 919126 G Q 919103 G Q 919164 CPB5 H. Evans/ R. Barreto Broo m Brazil S G Q 919127 G Q 919104 G Q 919165 CPB6 H. Evans/ R. Barreto Tree, Mal p i g hiaceae Broo m Brazil ? G Q 919147 G Q 919113 --CPB7 H. Evans/ R. Barreto Liana, Bi g noniaceae Han g in g litte r Brazil L G Q 919146 G Q 919106 G Q 919166 CPB8 H. Evans/ R. Barreto Liana, Bi g noniaceae Han g in g litte r Brazil L G Q 919128 G Q 919111 G Q 919167 CPB9 H. Evans/ R. Barreto Tree, Mal p i g hiaceae Broo m Brazil ? G Q 919148 G Q 919100 G Q 919168 CPB10 H. Evans/ R. Barreto Liana, Basidios p ore Brazil L G Q 919149 G Q 919114 G Q 919169 CPB12 H. Evans/ R. Barreto Broo m Brazil L G Q 919150 G Q 919112 G Q 919171 CPB14 H. Evans/ R. Barreto Broo m Brazil S G Q 919129 G Q 919104 G Q 919172 74

PAGE 75

Table 2-1. Continued. Isolate Collector/Donator Host origin Collection tissue Geographic origin Biotype Genbank ITS IGS RPB1 CPB17 H. Evans/ R. Barreto Liana Broo m Brazil L G Q 919131 G Q 919097 G Q 919173 CPB20 H. Evans/ R. Barreto Broo m Brazil S G Q 919132 G Q 919098 G Q 919174 CPB21 H. Evans/ R. Barreto Broo m Brazil S G Q 919133 G Q 919099 G Q 919175 CPB22 H. Evans/ R. Barreto Broo m Brazil S G Q 919134 G Q 919105 G Q 919176 TBOC3-2 L. Johnson Broo m Toba g o C G Q 919145 G Q 919091 G Q 919177 TCP3-2 R. Schnell Broo m Bolivia C G Q 919139 G Q 919090 G Q 919178 TCP8 A. Pommella Unknown p lant --Brazil ? G Q 919136 G Q 919109 G Q 919179 TCP24-1 unknown Broo m Brazil S G Q 919135 G Q 919107 G Q 919181 TCP33-1 R. Ploetz Broo m Brazil C G Q 919138 G Q 919088 G Q 919180 R. Ploetz Broo m Brazil C G Q 919140 G Q 919093 G Q 919182 TCP44-5 TCP90 C. Suarez-Ca p ello Broo m Ecuador C G Q 919142 G Q 919089 G Q 919183 TRDC15 L. Johnson Broo m Trinidad C G Q 919144 G Q 919086 G Q 919184 TRDC74 L. Johnson Broo m Trinidad C G Q 919143 G Q 919087 G Q 919185 A. Pommella Liana Hanging litter Brazil? L GQ919151 GQ919115 GQ919186 s p ( isolate C21 H. Evans --Costa Rica N/A AY916746 GU183377 AY91747 ----Ecuador N/A GU183375 GU183376 GU183378 var. DIS116e H. Evans --Ecuador N/A AY230255 ----var. IMI 389649 75

PAGE 76

76 not known not known ----N/A AY571032 --AY916683 C. Aime --New Mexico N/A AY916754 --AY916755 IGS RPB1 Genbank ITS Collector/Donator Host origin Collection tissue Geographic origin Biotype Table 2-1. Continued. sp. MCA2500 (Outgroup) Isolate

PAGE 77

Table 2-2. Maximum parsimony statistics fo r the ITS, IGS, and RPB1 datasets. # accessions # total characters # indel characters # PICa # nodes with increased supportb Mean increase in supportc Locus ITS 40 678 35 122 (18.0) 10 (83) 3.9 IGS 36 861 41 300 (34.8) 3 (43) 2.7 RPB1 39 747 --58 (7.8) 3 (27) -3.7 Number of parsimony informative characters, with percent of total charac ters in parentheses. Number of nodes that display increased bootstra p support after indel char acters were included, with percent of total nodes with >50% BS support in parentheses. c Mean percent increase of bootstrap support across all nodes with >50% BS 77

PAGE 78

Table 2-3. Statistics fr om phylogenetic analyses ITS IGS RPB1 ITS+IGS RPB1+ITS RPB1+IGS # characters 678 861 747 ------# PICa 122 300 45 ------# PIC in isolatesb 22 68 13 ------# trees 18 200 200 ------# steps 391 475 230 ------CI 0.903 0.931 0.948 ------RI 0.893 0.934 0.911 ------ILD -value c ------0.002 0.002 0.002 IMF d ------0.038 0.251 0.341 Clade Statistical supporte 1 59/95/63/------1a 41/--64/------1b 49/50 --87/84 ------2 55/62f83/92 -/------3 96/88 100/99 87/82 ------# parsimony informative characters. Excluding var. sp. MCA2500, and Incongruence length difference (ILD) d Mickevich-Farris index (IMF). e Based on 1000 boostrap replications (MP/ML) f ML analysis supported the grouping of TCP8 and CPB5 only 78

PAGE 79

Table 2-4. Topological congruence tests fo r single region data sets. Tree length Gain Loss Net a Data set with constraints 1. ITS 391 402121 11 0.0023 39980 8 0.0078 403131 12 0.0013 39100 0 ---b 391 0 0 0 --495 104 0 104 <0.0001 397 7 0 7 0.0384 411 20 0 20 0.0004 398 7 0 7 0.0384 392 1 0 1 --391 0 0 0 --2. IGS 475 510361 35 <0.0001 500250 25 <0.0001 47852 3 0.3340 47500 0 --47500 0 --489162 14 0.0010 571960 96 <0.0001 495233 20 0.0020 524565 61 <0.0001 504356 29 <0.0001 47621 1 0.5637 527531 52 <0.0001 3. RPB1 230--23880 8 0.0384 23880 8 0.0384 23550 5 0.0253 23440 4 0.0455 23000 0 --23770 7 0.0196 23550 5 0.0339 23330 3 0.0833 23000 0 --23440 4 0.0455 79

PAGE 80

80 Table 2-4. Continued. Tree length Gain Loss Net Data set with constraints a -values are based on the Wilcoxon signed-rank test of the number of steps gained or lost by using the corresponding t opological constraints. <0.05 indicate the topology is significantly less parsimonious than the trees from the data set being tested. b WSR cannot be used for topologies with 0 or 1 step changes. 230 0 0 0 ---

PAGE 81

Table 2-5. Macroscopic m orphological characte rs of select accessions of Pileus characteristics Accession Host Shape Diametera Color Margin Surface Flesh Conic turning extremely convex, broadly umbulate 7 (13) 20 Rose center turning to cream with white margins Margins inrolled, becoming planar to uprolled; entire becoming undulating or eroded Dull and dry, wrinkled Cream colored, yellows after harvest CPB1 C onic, broadly umbellate 7 (9) 16 Crimson center fading to creamyyellow margins Margins inrolled Dull and dry, wrinkled Creamy-yellow colored, y TRDC74 Conic to catenulate, becoming slightly uplifted/convex, broadly umbellate 4 (9) 22 Crimson center fading to cream margins Margins inrolled, becoming planar to uprolled; entire becoming undulating or eroded Dull and dry, wrinkled Cream colored, yellows after harvest CPB2 Conic turning extremely convex, broadly umbellate 7 (16) 25 Pale crimson center to light rose or cream margin Margins inrolled, becoming planar to uprolled; entire becoming undulating or eroded Dull and dry, wrinkled Cream colored, yellows after harvest CPB 15 Conic to catenulate, sometimes uplifted, broadly umbellate 11 (18) 32 Uniform rose color becoming rose in center fading to cream at margins Margins inrolled, becoming planar to uprolled; entire becoming undulating or eroded Dull and dry, wrinkled Cream colored, yellows after harvest CPB5 CPB9 Malpighiaceae Conic to catenulate, becoming slightly uplifted/convex; broadly umbellate 12 (14) 22 Pale crimson center to light rose or cream margin Margins inrolled, becoming planar to uprolled; entire becoming undulating or eroded Dull and dry, wrinkled Cream colored, yellows after harvest Conic to catenulate, broadly umbellate 7 (9) 10 Pigment present, but altered from preservation Margins inroled becoming planar; entire becoming undulating or eroded n/a Cream colored CPB12 CPB10 Liana, Malpighiaceae Conic to catenulate, broadly umbellate 6 (8) 10 Pigment bleached from preservation Margins inroled becoming planar; entire becoming undulating or eroded n/a Cream colored 81

PAGE 82

Table 2-5. Continued Stipe characteristics Accession Host Width Lengtha Color Shape Flesh Surface Base Cent er Apex 0.5 -3.0 0.5 2 5 (8) 10 Dark crimson base to cream at apex Round, slightly swollen base and apex Hollow striated CPB1 1. 0-2.0 0.5-1.0 0.5-1.0 7 (9) 10 Dark crimson base to cream at apex Round, slightly swollen base and apex Hollow striated TRDC74 0.9 0.5 0.5 5 (8) 12 Dark crimson base to cream at apex Round, slightly swollen base and apex Hollow striated CPB2 1.0-2.0 1.0-2.0 1.0-2.0 5 (7) 8 Dark crimson to rose at apex Equal Hollow striated CPB 15 1.0-1.5 1.5-4.0 6 (11) 14 Rose to crimson base to creampink at apex Tapered from apex to base Hollow striated CPB5 CPB9 Malpighiaceae 0.5-1. 8 0.5-1.8 0. 5-1.8 10 (13) 17 Dark crimson base to cream at apex Equal Hollow striated 1.0 1.0 1.0 3 (3) 4 Crimson base to cream at apex Equal Hollow striated CPB12 CPB10 Liana, Malpighiaceae 1.0 1.0 1.0 7 (only one stipe intact) Crimson base to ream at apex Equal Hollow striated Lamellae characteristics Accession Host Lengtha Color Attachment Spacing Margin Face Lamellulae 2 (3) 5 Cream Adenate Distant, 2 (4) 6 Smooth to undulating Powdery Present, 0-3 CPB1 3 (3 ) 4 Creamyyellow Adenate Distant, 1 (2) 4 Smooth to undulating Powdery Present, 0-1 TRDC74 Cream Adenate Distant Smooth to undulating Powdery Present, 0-3 CPB2 3 (4) 4 Cream Adenate Distant, 2 (3) 4 undulating Powdery Present, 0-3 CPB 15 Cream Adenate Distant, 2 (4) 9 Smooth to undulating Powdery Present, 0-3 CPB5 CPB9 Malpighiaceae 1 (2) 3 Cream Adenate Distant, 1 (3) 4 Smooth to undulating Powdery Present, 0-3 n/a Cream Adenate Distant, 2 (3) 3 Smooth to undulating n/a Present, 0-1 CPB12 CPB10 a Measurem inimum (mean) maximum, in m ents are given as mLiana, Malpighiaceae n/a Cream Ade nate Semi-distant 1 (1) 2 Smooth to undulating n/a Present, 0-1 82

PAGE 83

Table 2-6. Cystidia, basidia, and basidiospor e characters of s elect accessions of isolates. Cystidia characteristics Pileocystidia Cheilocystidia Host Shape Pigment Melzers Dimensionsa Shape Pigment Melzers Dimensions Accession 1b clavate, trichoderm, clamps abundant red pigment, cytoplasmic and incrusted, many hyaline dextrinoid incrustations 30 (39) 52 x 7(7) 8 pyriform, swollen, clamped at base hyaline none 17 (25) 33 x 12(16) 25 CPB1 2 ---c ----28 (44) 60 x 5 (7) 8 pyriform, lageniform (bottleshaped) to swollen elliptical ----17(27) 40 x 10 (12) 22 1 clavate, elliptical to obclavate to mucorate; clamps abundant elliptical cystidia darkly cytoplasmic ally pigmented; incrustations dextrinoid incrustations 25 (40) 52 x 5 (9) 17 clavate to swollen; scarce hyaline none 20 (27) 38 x 7 (12) 17 TRDC74 2 ------28 (39) 60 x 5 (9) 17 ------18 (30) 45 x 7 (10) 12 CPB2 1 clavate, obclavate to mucorate red pigment cytoplasmic, some hyaline, sometimes clamped; pigmented cystidia are thickwalled/ incrusted dextrinoid incrustations ; 1 case of dextrinoid cytoplasmic 20 (36) 48 x 10 (8) 15 clavate hyaline none 17 (25) 30 x 8 (10) 12 83

PAGE 84

Table 2-6. Continued. Host Shape Pigment Melzers Dimensionsa Shape Pigment Melzers Dimensions Accession 2 clavate, trichoderm cylindrical to obclavate, mucorate; clamps abundant red pigment cytoplasmic (thickwalled ) and incrusted (thin walled) dextrinoid cytoplasmic and incrustations 25 (54) 99 x 5 (6) 8 swollen elliptical to mucorate, lageniform ----17 (23) 32 x 8 (12) 13 1 clavate, trichoderm to obclavate to mucorate; clamps abundant Cytoplasmic pigmentation --19 (42) 69 x 5 (7) 13 Pyriform, lageniform Hyaline None 22 (28) 34 x 9 (11) 14 CPB 15 1 swollen clavate, often clamped, second cell swollen red pigment often incrusted, sometimes cytoplasmic; often darker in tip of cell dextrinoid in incrustations 18 (36) 57 x 7 (11) 15 clavate to obclavate to slightly mucorate, scarce hyaline none 23 (30) 42 x 8 (12) 17 CPB5 2 clavate, swollen, a few cylindrical; rarely mucorate; clamps abundant pigment darker at tip of cell; cytoplasmic and incrusted dextrinoid in incrustations 23 (40) 72 x 7 (9) 13 ------18 (26) 32 x 7 (10) 13 84

PAGE 85

Table 2-6. Continued. Host Shape Pigment Melzers Dimensions Shape Pigment Melzers Dimensions Accession CPB9 Malpighiaceae 1 clavate, elliptical to obclavate to mucorate; clamps abundant elliptical cystidia darkly cytoplasmic ally pigmented; incrustations dextrinoid cytoplasmic, incrustations 18 (38) 54 x 5 (8) 13 pyriform, napiform hyaline none 17 (23) 30 x 8 (10) 12 2 clavate, elliptical to obclavate to mucorate; clamps abundant elliptical cystidia darkly cytoplasmic ally pigmented; incrustations --28 (48) 93 x 5 (10) 12 pyriform, napiform, rarely mucronate hyaline none 23 (25) 29 x 6 (8) 9 CPB12 1 clavate, cylindrical, rarely obclavate mostly hyaline, very few with light cytoplasmic pigment --40 (77) 112 x 6 (12) 19 swollen clavate to pyriform hyaline none 23 (28) 33 x 8 (10) 13 CPB10 Liana, Malpighiaceae 1 Clavate, trichoderm, rarely obclavate Cytoplasmic pigment --43 (68) 99 x 5 (7) 12 Obclavate to pyriform, rare, subtle difference from basidia hyaline None 23 (27) 33 x 9 (11) 14 2 Clavate, trichoderm, rarely obclavate Cytoplasmic pigment --40 (78) 130 x 5 (8) 12 None observed hyaline None --85

PAGE 86

Table 2-6. Continued. Basidia characteristics Basidia Basidiospores Accession Host Shape # cells Dimensions Color Shape Surface Melzers Dimensions 1 clavate 2-3 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 8 (9) 10 x 3 (5) 7 CPB1 2 --4 17 (20) 27 x 7 (8) 10 hyaline ------7 (9) 10 x 3 (5) 7 1 clavate no sterigmata seen --hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 8 (10) 12 x 5 (5) 7 TRDC74 2 --no sterigmata seen 18 (23) 25 x 5 (7) 8 hyaline ------8 (9) 12 x 5 (5) 8 CPB2 1 clavate mostly 4 20 (23) 27 x 5 (7) 8 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 8 (9) 12 x 5 (5) 7 2 clavate mostly 4 18 (20) 22 x 7 hyaline ------8 (10) 12 x 3 (5) 5 1 Clavate no sterigmata seen 21 (24) 25 x 8 (8) 9 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 9 (10) 13 x 4 (5) 6 CPB 15 1 clavate 4 17 (20) 27 x 7 (8) 10 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 8 (10) 12 x 5 (6) 7 CPB5 2 clavate 4 18 (24) 28x 7 (8) 8 hyaline elliptical to amygdaliniform thinwalled, smooth none 8 (11) 12 x 5 (5) 7 Maphigeaceae1 clavate No sterigmata seen 20 (23) 27 hyaline elliptical to amygdaliniform thinwalled, smooth none 7 (9) 10 x 3 (5) 5 CPB9 2 clavate 2-4 celled 25 (28) 29 hyaline elliptical to amygdaliniform thinwalled, smooth none 6 (8) 10 x 4 (5) 5 CPB12 1 clavate 2-4 celled 25 (27) 30 x 6 (7) 8 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 9 (11) 13 x 5 (6) 6 86

PAGE 87

Table 2-6. Continued. Accession Host Shape # cells Dimensions Color Shape Surface Melzers Dimensions CPB12 2 clavate no sterigmata seen 23 (24) 28 x 5 (7) 8 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 9 (11) 13 x 5 (6) 6 CPB10 Liana, Malpighiaceae 1 clavate 4-celled 25 (29) 30 x 6 (7) 8 hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none 10 (11) 13 x 6 (6) 8 2 clavate 4-celled no clear measures hyaline elliptical to amygdaliniform, apiculate thinwalled, smooth none a Measurements are given as minimum (mean) maximum, in m b Two basidiocarps were examined wh en available. When possible, basidi ocarp 2 was older than basidiocarp 1 c Dashes denote identical de scription as basidiocarp 1 87

PAGE 88

88 Table 2-7. Dimensions of basidiospores of select accessions of Spore dimensions (m) Accession Clade Length Width Ratio TRDC74 1a 9.6.2 ba 5.2.1 cd 1.9.0 b CPB1 1a 8.9.1 c 5.0.1 d 1.8.0 b CPB2 1 9.5.2 b 5.1.1 d 1.9.0 b CPB15 1 9.9.2 b 5.9.1 b 1.9.0 b CPB5 2 10.6.1 a 5.4.1 bc 2.0.0 b CPB9 2/3 8.4.2 d 4.7.1 e 1.8.0 a CPB10 3 10.7.1 a 6.0.1 a 1.8.0 b 3 10.9.1 a 5.8.1 a 1.9.0 b CPB12 a Values within columns with the same letter are not significantly different based on Fishers LSD, =0.05

PAGE 89

4. 7. 31. TCP33-1 TCP44-5 CPB1 TCP90 TRDC74 TBOC3-2 17. 46PR 28. TRDC15 TCP3-2 73-31 73-62 92-10-7 CPB15 CPB2 TCP24-1 CPB22 CPB21 CPB20 CPB17 CPB14 CPB3 CPB4 CPB8 CPB5 TCP8 CPB7 CPB12 CPB10 CPB6 CPB9 TCP7 M. roreri Dis116e M. roreri var. gileri IMI 389649 M. roreri C21 Moniliopthora sp. MCA2500 C. liliputianus C61867 Outgroup 2 60/59 55 / -90/84 96/89 100/10 0 100/100Clade 2 Clade 3 Clade 1 Clade 1a 89/86 83 / 78 59/59 49/44 60 / -68/65 64 / -Clade 1b Figure 2-1. Phylogenetic relationships among 36 accessions of based on ITS sequence data. The phylogram represen ts one of 18 most parsimonious trees (391 steps, CI=0.903, RI=0.893). Support va lues are bootstrap values over 50% including indels coded as ch aracters (before slash), and data set treating indels as missing data (second slash). Isolates are code d for biotype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L) 89

PAGE 90

4. F igure 2-2. Phylogenetic relationships among accessions of resolv ed using ML analysis of ITS sequence da ta, ln score = -2472.65. Support values are bootstrap values over 50%. Isolates are code d for biotype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L) Clade 3 Clade 1 Clade 1a Clade 1b Clade 2 73-31 31. 73-62 TRDC74 TCP90 7. 17. 46PR TCP33-1 TCP44-5 CPB15 28. TRDC15 CPB1 TCP3-2 TBOC3-2 CPB2 CPB17 CPB22 CPB14 CPB4 CPB3 92-10-7 CPB20 CPB21 TCP24-1 CPB8 CPB5 TCP8 TCP7 CPB7 CPB10 CPB12 CPB6 CPB9 M. roreri C21 M. roreri IMI 389649 M. roreri Dis116e 2 Moniliopthora sp. MCA2500 C. liliputianus C61867 66 99 50 62 52 88 86 100 Outgroup 0.02 90

PAGE 91

CPB17 CPB20 CPB21 CPB9 CPB4 CPB7 TCP24-1 CPB15 92-10-7 TRDC74 TBOC3-2 4. TCP3-2 CPB14 CPB3 CPB22 7. CPB1 17. TCP90 73-62 TCP44-5 73-31 CPB2 31. 28. TCP33-1 46PR TRDC15 CPB10 CPB12 CPB6 CPB8 TCP8 CPB5 TCP7 M.roreri Dis116e M. roreri C21 Clade 1 95/86 53/69 63/78Clade 3 100/100 100/100 Clade 2 83/78 91/55 Outgroup 20 Figure 2-3. Phylogenetic relationships among 36 accessions of based on IGS sequence data. The phylogram represents one of 200 most parsimonious trees (475 steps, CI=0.931, RI=0.934). Support va lues are bootstrap values over 50% including indels coded as ch aracters (before slash), and data set treating indels as missing data (second slash). Isolates are code d for biotype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L). 91

PAGE 92

TCP3-2 CPB3 4. 17. 31. TRDC15 46PR TRDC74 CPB4 28. TCP24-1 TCP44-5 CPB2 73-31 73-62 TCP33-1 TCP90 CPB22 CPB7 92-10-7 CPB14 CPB1 CPB17 CPB20 CPB15 CPB21 CPB9 TBOC3-2 7. CPB8 TCP8 CPB5 CPB12 CPB6 CPB10 TCP7 M. roreri Dis116e M. roreri C21 0.00592 76Clade2 Clade 1 99Clade 3 Outgroup Figure 2-4. Phylogenetic relationships among accessions of \resolved using ML analysis of IGS sequence data, ln score = -3023.85. Support values are bootstrap values over 50%. Isolat es are coded for biot ype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L) 92

PAGE 93

CPB15 4. 46PR 17. 73-62 CPB1 CPB2 CPB5 TCP90 TRDC15 TRDC74 TCP44-5 TCP33-1 TCP3-2 TBOC32 73-31 31. 28. 7. TCP8 TCP7 CPB7 CPB4 92-10-7 CPB14 CPB17 CPB20 CPB22 TCP24-1 CPB3 CPB2 1 CPB8 CPB12 CPB9 CPB10 M. roreri C21 M.roreri Dis116e M. sp. MCA2500 C. liliputianus 63/61 Clade 1b Clade 1a Clade 1 64/64 65/63 87/8765/100 87/88 100/100/ 50 Clade 3 81/87100/100 61/60 Outgroup 3Figure 2-5. Phylogenetic relatio nships among 35 accessions of based on RPB1 sequence data. The phylogram repr esents one of 198 most parsimonious trees (230 steps, CI=0.948, RI=0.911). Support values are bootstrap values over 50% including indels coded as ch aracters (before slash), and data set treating indels as missing data (second slash). Isolates are code d for biotype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L) 93

PAGE 94

94 Clade 1b TCP33-1 TRDC74 CPB2 31. 17. TCP90 TCP44-5 73-31 73-62 TCP3-2 7. CPB1 CPB5 4. CPB15 46PR TBOC3-2 28. TRDC15 TCP7 TCP8 CPB7 TCP24-1 92-10-7 CPB17 CPB4 CPB22 CPB3 CPB21 CPB14 CPB20 CPB8 CPB9 CPB10 CPB12 M. roreri C21 M.roreri Dis116e M. sp. MCA2500 C. liliputianus 0.02 58 57 84 79 82 62Clade 1a Clade 3 Outgroup Figure 2-6. Phylogenetic relationships among accessions of resolv ed using ML analysis of RPB1 sequence da ta, ln score = -2124.22 Support values are bootstrap values over 50%. Isolates are code d for biotype as follows (open circle=C, filled circle=S, open square= and other Malpighiaceae, closed square=L)

PAGE 95

C D E F B A Figure 2-7. Microscopic characteristics of access ions of A) Pileocystidia of isolate TRDC74 showing long, clavate-el liptical cystidia, as well as shorter mucorate cystidia (arrow); B) Obcl avate pileocystidia of CPB5 with swollen second cell and dark pigmentation at tip ; C) Obclavate to slightly pyriform cheilocystidia of CPB10; D) Swollen, pyriform cheilocystid ia of CPB12; E) Lageniform (bottle-shaped) cheilocystidia of CPB15; F) Typical basidiospores (CPB10). Bars=10m. 95

PAGE 96

CHAPTER 3 USE OF PHYLOGENIES FROM M ULTI PLE DNA REGIONS TO ASSESS THE DIVERGENCE OF LINEAGES OF SENSU LATO THAT ARE ASSOCIATED WITH MA NGO ANTHRACNOSE IN SOUTH FLORIDA Mango ( ) is an important fruit crop in the tropics and subtropics. The FAO (Food and Agricultural Organi zation) estimated world production of the fruit in 2007 at over 33 million tons (FAOSTAT 2009 statistics, h ttp://faostat.fao.org/). Asian nations are the largest producers, incl uding India, China and Thailand. Mexi co is the largest producer in the Americas, with 2.05 million tons in 2007, and is also the worlds largest exporter. Although production in Florida is marginal in compar ison (2500 tons in 1997) (Mossler and Nesheim 2002), many commercially important cultivars were developed in the state, beginning with the release of Haden in 1912, which served as a fe male parent for Tommy Atkins, Lippens, Zill, and several other cultivars (Knight and Sc hnell 1994). A high level of crop diversity exists in south Florida due to the array of germplasm th at was introduced to the area (Schnell et al. 2006). In humid growing regions, anthracnose is the most prevalent th reat to mango production. Although anthracnose also affects leaves, it is most important on panicles, where it causes reduced fruit set, and on fruit, where postharvest losses can be severe if management measures are not implemented. Three taxa have been associated with these diseases: (worldwide), var. (Australia, Simmonds 1965), and (Australia, Fitzell 1979; Taiwan, Weng and Chuang 1995; and Homestead, Florida, Riveras-Vargas et al. 2006). The presence of a mangospecific population of was suggested by several researchers in the 1990s. Hodson et al. (1993) showed with restriction fragment length 96

PAGE 97

polymorphisms (RFLPs) of m itochondrial (mtDNA) and ribosomal (rDNA) DNA that isolates of from mango were the only ones from four tropical fruit species that could be grouped by host. Further studies with RFLPs and random amplification of polymorphic DNAs (RAPDs) identified a distinct mango population when anthracnose isolates fr om tropical fruits in Sri Lanka and Australia were compared (Alakah oon et al. 1994, Hayden et al. 1994). Inoculation studies have shown that isolates of usually cause symptoms on diverse fruit hosts, regardless of the original host, but that greater disease us ually develops on the original host (Alakahoon et al. 1995, Sanders and Korsten 2003). Prior work in south Florida investigated tissue specialization w ithin populations of on mango. Pectic zymograms of 63 isolates from leaves, inflorescences, and immature and mature fruit revealed 10 pectinas e profiles that were somewhat related to host tissue (Gantotti and Davis 1993). For example, 75% of the immature fruit isolates had one profile, and 44% and 53% of the mature fruit and leaf isolates had another profile. A follow-up study with RAPDs and pathogenicity tests on fruits and inflorescences also suggested tissue specificity among isolates from mango (Davis 1999). One population tended to come from inflorescences and was less virule nt on leaves and mature fruit th an inflorescences. However, no phenetic or cluster analysess were performed with these data. Since the 1990s, no subsequent work has been conducted to characterize mango-associated in south Florida or else where. Additional work is warranted to qualify and quantify phylogenetic and bi ological relationships for populations from mango and other hosts, as are c ontemporary investigations of the tissue-specificity of mangospecific populations. 97

PAGE 98

var. was first associated with mango anthracnose by Simmonds (1965), who distinguished it from as having slightly narrower conidia (3.7 m vs. 4.8 m), despite the considerable overlap he observed in conidium width between the two taxa. Cox a nd Irwin (1988) later reported that conidium widths for var. were 3.0-4.2 m compared to 4.5-5.5 m for Both taxa had unlobed or slightly lobed appressoria. Given these slight differences, var. is no longer recognized. Taxonomic classification within was traditionally based on morphology or host specificity. The last thor ough revision of the genus was co mpleted by von Arx (1957), who sought to condense the hundreds of species that had been erected based largely on host origin into morphologically defined species. For fungi, the morphological species concept can be highly artificial, and morphology has repeatedly been shown to do a poor job of organizing phylogenetically or biologically relevant ta xa, a generalization th at holds true for (Sutton 1992, Du et al. 2005). Crypt ic species, in which phylogeneti cally distinct taxa cannot be separated by morphology, are common and espe cially problematic (Hawksworth 2001, Kohn 2005). Currently, more meaningful species that fit the evolutiona ry species concept (Wiley 1978, Taylor et al. 2001), and which incorporate phylogenetic analys es of molecular data, host specialization, and the traditional biological species concept (i.e. interfertility within species), are sought by mycologists and plant pathologists. In von Arxs (1957) revision of approximately 600 previously described species were placed in synonymy with even though extensive morphological and host diversity were noted. sensu lato is currently considered a species complex that affects hundreds of plant hosts and contai ns several described species, 98

PAGE 99

including (cause of banana anthracnose), (strawberry crown rot), and (coffee berry disease) (von Arx 1957, Sutton 1980, Sutton 1992). Although slight morphological differences have been reported for these hostspecific species, phylogenetic analyses consistently p lace them in the sensu lato clade (Sreenivasaprad et al. 1996a, Munaut et al. 2002, Martinez-Culebras et al. 2003, Du et al 2005). The extent to which these taxa are separated phylogenetically and the emphasis given to host-specificity would determine whether they remain distinct species or are considered subspecific groups within the species complex. Although the previous ly described mango biotype appears to be another example of a host-specific lineage within senus lato, additional phylogenetic analysis is needed to determine if it represents a distinct taxon and, if so, the appropriate level of its classification. The objective of this study wa s to determine whether mango-specific groups exist within sensu lato. Phylogenetic analyses using DNA sequence data from several regions were used to characteri ze relationships among isolates of from different mango organs, as well as other host specie s. Morphological and pathological data were assessed for isolates that repres ented the different phylogenetic taxa that were identified in the study. The case for a mango-specific taxon in sensu lato is discussed, as are evolutionary patterns in A hierarchical sampling scheme was used to collect isolates During June-July 2007, symptomatic mango leaves, peduncles, and mature fr uit, as well as asymptomatic immature fruit and peduncles were collected in two groves at the University of Florida Tropical Research and Education Center (TREC), one of mixed cultivar s and the other of cv. Keitt. During March99

PAGE 100

August 2008, inflorescences, leaves, immature fr uit and mature fruit from mango, as well as symptomatic fruit from (avocado), (banana), (carambola), (guava), and (papaya) were collected in plantings at TREC and elsewhere in Miami-Dade County. Up to three symptomatic organs were chosen from each tree included in the study, from which up to three lesions were sampled, from wh ich three acervuli were sampled, resulting in up to 27 subsamples per organ from each tree. Lesions were excised from tissues and surface disinfested (10 s 70% EtOH, 2 mi n bleach, rinsed in sterile H2O) and placed in a moist chamber (plastic Petri plate with moist filter paper). Af ter 2-3 days, conidia were collected from acervuli with a sterile needle and placed on half-strength PDA. Isolated colonies were streaked onto fresh PDA plates and single, germinated spores pi cked off and plated on PDA. Up to three asymptomatic, immature fruit were sampled from each tree included in the study. Three pieces of tissue, approximately 5x5 mm were excised and su rface disinfested as described above. Tissues pieces were then placed in molten (ca. 50C) PD A to ensure maximum surface area contact with tissue and media. isolates growing out of tissue pieces were streaked onto fresh PDA, and single germinated spores were picked off and plated on PDA. A ll single-spore cultures were stored directly onto filter paper stored at 4C and in 10% glycerol stored at -80C. At least one isolate from each organ was stored for future analysis (Table 3-1). Molecular analyses were conducted with 51 isolates of spp. that represented all mango organs and isolates from TREC and other locations (Table 3-2). Also included were three isolates from avocado, four from banana, three from carambola, three from guava, two from papaya, one from passionfruit and one from and were used as outgroup taxa. 100

PAGE 101

Isolates were grown for 3-5 da ys in PDB at room temperatur e (23-25C) on a shaker at ca. 80 rpm Approximately 150 mg mycelia were harves ted and triple rinsed with sterile deionized water and dried on sterile filter paper. DNA wa s extracted using a DNA genomic preparation protocol from the University of Wisconsin Biotechnology Center. Briefl y, mycelia were ground in 500 l Shorty DNA Extraction Buffer (0.2M Tris-HCl, pH 9.0, 0.4M LiCl, 25mM EDTA, 1% SDS) and incubated at 68C for 10 minutes. Ti ssue was centrifuged at 14,000 rpm for 5 minutes, 400 l of the supernatant was transferred to ne w tube and the DNA was precipitated with 400 l 99% isopropanol. The tubes were centrifuged for 10 minutes at 14,000 rpm and the supernatant was decanted. The DNA pellets were air-dried for 5 minutes then resuspended in 400 l TE buffer (10 mM Tris-Cl, pH 7.5, 1 mM EDTA) for 30 minutes at room temperature. 2l DNA was used in PCR reactions. Three nuclear regions were used in phyloge netic analyses: the widely used ITS1-5.8SITS2 (ITS); a gene that encodes the high m obility group-box (HMG) region of the MAT1-2 mating locus; and CGTT5, a locus that was cloned from and exhibits homology to a hypothetical protein from and The MAT1-2 locus has previously been s hown to be more useful than ribosomal sequences or protein-coding gene introns in phylogenetic studies of closel y related taxa at the subgeneric level (Turgeon, 1998; Du et al., 2005) The CGTT5 marker was developed by cloning a 895-bp band from isolate Cg49 of that was generated with the (TCC)5 RAPD primer (Appendix A); it c ontained a higher number of phylogenetically informative sites than the ITS and MAT1-2 regions (Table 3-3). All PCR reactions were carried out in 50 l reactions that co ntained: 38.25 l of sterile distilled, deionized water; 6.5 l ThermoPol Re action Buffer (New England Biolabs, Ipswich, 101

PAGE 102

MA); 1 l 10 mM dNTP mi x (New Engla nd Biolabs, Ipswich, MA); 0.25 l Taq DNA polymerase (conc. 5,000 units/ml); 1 l each of 15 M primers; and 2 l DNA template. Standard cycling parameters with a 55C anneal ing temperature were used. The ITS region was amplified with the ITS1 (Gardes and Bruns 1993) and ITS4 (White et al. 1990) primers, the MAT locus was amplified with primers HMGg loF1 and HMGgloR1 (Du et al. 2005), and CGTT5 with primers pTT5F and pTT5R (Appe ndix A). Direct Sanger sequencing of both strands of the PCR products was performed at the University of Florida ICBR Facility with the above primers. For ambiguous bases, the base that agreed with the consensus sequence was retained. Because sequencing signal from direct sequencing was strong, cloning was not performed. Sequences were aligned with CLUSTALX (T hompson et al. 1994) in Mega4 and default parameters, and alignments were adjusted manually. Both maximum parsimony (MP, using PAUP* v.4.0b; Swofford 2000) and maximum likelihood (ML, using GARLI v.0.946) analyses were conducted for all data sets, and gaps were treated as missing data. Parsimony analyses were conducted using heuristic searches with 100 rand om addition replicates, with tree bisection reconnection (TBR) swapping, saving no more than tw o trees with tree scores greater than five per repetition. Consistency (CI) a nd retention indices (RI) values were calculated with PAUP*. To determine statistical support for groups in the phylogenies, the non-para metric bootstrap test (Felsenstein 1985) was performed using 1,000 repe titions and heuristic search criteria as described above except that 10 random addition replicates were used, saving no more than two trees with scores greater than five per repetition. ML analyse were with GARLI v.0.946, using defa ult parameters and a randomly generated starting topology, and were stopped after 5x106 generations or a 0.01 de crease in ML score. 102

PAGE 103

Three independent ML runs were performed a nd resulting topologies were com pared in TreeView v1.6.6. If similar topolog ies were generated, bootstrap an alysis was performed using the same parameters with 1,000 repetitions. Tree files from analyses were imported into MEGA 4.0 for visualization and to edit phylograms. For the ITS and MAT1-2 analyses, and were included as outgroups. Because CGTT5 sequences for taxa outside of sensu lato were not available, analysis of the CGTT5 data set was not rooted. In order to determine if combining data from the three loci was appropriate, levels of congruence were tested between pairs of the da ta sets and between topologies of phylogenetic trees (Seelanen et al. 1997, Johnson and Soltis 1 998). Overall incongruence between data sets was quantified using the Mickevich-Farris index (IMF) (Mickevich and Farris 1991) and the incongruence length difference (ILD) between loci (Farris 1994) using the partition homogeneity test in PAUP*. This test was run with 500 replicates of 10 random addition sequences, and < 0.05 indicated significant incongruence between data sets. The ILD test is known to be very sensitive (Yoder et al. 2001; Darl u and Lecointre 2002), so the decision as to whether to combine data sets was largely based on examina tion of topology and bootstrap support. To assess topological congruence, an approach similar to that used by Seelanen et al. (1997) was followed. First, consensus trees for each of the three regions were compared empirically to determine if instances of topol ogical incongruence were hard (terminology of Seelanen et al. 1997 for isolates that resolve in different we ll-supported clades in different topologies), or soft (due to w eak phylogenetic signal in one of th e data sets). To determine the statistical significance of topologica l incongruence, topological cons traints were applied to each DNA region corresponding to the 70% bootstrap consensus tree fo r the other two regions. The Wilcoxon signed-rank (WSR) test was applied to de termine if the number of steps that were 103

PAGE 104

gained or lost in the re sulti ng tree was significantly differe nt than the original trees ( =0.05,as described by Templeton 1983). In addition, strongly supported clades that were resolved in each single region analysis were used as single constraints for the other loci (Table 3-5). This approach allowed congruence of specific clades to be tested in each data set, and identified individual clades and taxa that caused incongruence between data sets. This approach helped determine if data sets should be combin ed, and provided phylogenetic information about evolutionary events that resulted in incongr uence, such as cryptic outcrossing, hybridization, introgression and horizontal gene transfer (Wendel and Doyle 1998). Loci that were determined to be largely congruent were combined and analyzed using parsimony and maximum likelihood analysis as described above. For these analys es, accessions with missing sequences were pruned from the combined data set. Thirteen isolates representing genetic clades that resolved in the phylogenetic analyses were characterized morphologica lly. The shapes and dimensions of conidia and hyphopodia are used to define species and were shown recently to be phylogenetically useful (Du et al. 2005). Conidia were harvested from 7-day-old PDA cultures. Hyphopodia (vegetative appressoria) were produced by grow ing isolates over sterile glass coverslips at the edge of PDA colonies. The shapes and dimensions of coni dia and hyphopodia were determined with a Leitz Laborlux 2 microscope and an ocular micrometer. For all pathogenicity tests, isolates were retrieved from -80C st orage and grown on PDA for 7-10 days at room temperature (23-25C) and ambient li ght. Spores were harvested by flooding plates with sterile deio nized water, scraping colonies, and straining suspensions through 104

PAGE 105

a layer of sterile cheesecloth. Inoculum was adjusted to 5x105 to 1x106 conidiaml-1 in sterile deionized water. Field inoculations were carried out at TREC in two experiment al groves: one of cv Keitt and one of mixed cultivars. Trees were spaced 4 m apart in rows 7 m apart. Keitt tree height was approximately 4 m; trees in the mixed cultiv ar block were approximately 6 m. Trees were chosen for experiments based on having suffici ent numbers of organs at the appropriate developmental stage, and 4-5 trees were incl uded in each experiment. At the end of each experiment, lesion margins were excised from l eaves, surface disinfested, and plated on PDA to confirm presence of the pathogen. In 2009, mouse-ear to green-colored stage in florescences (Schoeman et al. 1995) were inoculated in the field with 13 isolates; 106 conidiaml-1 of each isolate was applied with a handheld manual spray bottle. Treatments (isolate s and a water control) were replicated four (expt 3, cv Sensation) or five (expts 1 a nd 2, cv Keitt) times on single-inflorescence experimental units in a randomized complete block design (RCBD), where blocks were individual trees. Inflorescences were sprayed until runoff and th en covered with plastic bags inside brown paper bags for 48 hrs. Severity ratings were taken six times for four weeks after inoculation, using a synoptic key (key 1.5, Jame s 1971). Analysis of variance of the area under the disease progress curve (AUDPC) and ymax (highest disease severity) were performed for each experiment. Mean separations were performed using least significant differences (PROC GLM, SAS v.9.0, Cary, NC). During June-July, 2009, two experiments were conducted on cv. Keitt at TREC. Leaves on newly opened vegetative shoots were inoculated in the field as described for infloresences, 105

PAGE 106

except that plastic bags were om itted due to high summer temperatures. Ten (expt 2) or 14 (expt 1) treatments (nine or 13 isolates and a water contro l) were replicated four time in RCBDs, where buds and the subtending five leaves were experimental units and blocks were single trees. Disease severity was measured after 4, 7, 10 and 14 days with a synoptic key (key 2.1.2, James 1971). Experiments were terminated after 14 days due to severe defoliation in some treatments. Analyses of variance of AUDPCs and ymaxs were performed with PROC GLM in SAS. Least significant differences (PROC GLM, SAS) were used to separate means. Imported, heat-treated fruit of c v. Tommy Atkins were obtained from a local packinghouse (LimeCo, LLC, Naranja, FL). Treatments (seven isolates and a noninoculated cont rol) were replicated five ti mes in a completely randomized design (CRD). Depending on fruit size, each fruit was treated at two or three points, each of which was wounded with a sterile needle and covered subsequently with 15l droplets of sterile 0.3% water agar that did not c ontain (control) or contained 106 conidiaml-1 of a given isolate (i.e., 15,000 conidia per inoculated site). Fruit we re incubated in plastic sweater boxes on wire mesh over moistened paper towels at 25C in th e dark, and lesion diameters were measured in two directions at a right angl e after 4, 5, 6, and 7 days. The experiment was conducted four times. Analyses of variance of AUDPCs and fina l lesion diameters were performed with PROC GLM in SAS, and least significant differe nces were used to separate means. In 2009, single experiments were conducted in the field on attached fruit of three cvs, Haden (susceptibl e), Tommy Atkins (moderately resistant) and Van Dyke (moderately resistant). Immature fruit (ca 8 cm in dia) were surface disinfested (10 s 70% ethanol, 2 min 10% bl each, rinse with sterile H2O) and dried before they were treated 106

PAGE 107

individually with either one of seven isola tes or a water control; fruit were misted with a suspension of 5x105 conidiaml-1 sterile water or sterile water, and covered with brown paper bags. Bags were removed when fruit were harvested at maturity. Disease severity was rated four times over 10 days after harvest w ith a synoptic key (Corkidi et al. 2005). Single fruit were experimental units, treatments were replicated four times in RCBDs, and trees were blocks. Analyses of variance of AUDPCs and ymaxs were performed with PROC GLM in SAS, and least significant differences were used to separate means. CGTT5 was more informative than either the ITS or MAT regions (Table 3-3). Overall, 61 PICs were found at the CGTT5 locus vs 27 and 39 for, respectively, ITS and MAT (Table 3-3). Moreover, all CGTT5 PICs were found in ingrou p accessions (11.4% of all characters), whereas only 11 (2.0%) and 13 (6.0%) of th e respective ITS and MAT PICs were found in the ingroups used in those analyses. Most of the 65 accessions of sensu lato in the ITS analyses were from mango, and and were used as outgroup taxa (Table 3-2). The data set contained 557 characters and MP analyses resulted in 11 trees with 112 steps (CI=0.938, RI=0.949). A MPT that agrees with the majority-r ule consensus tree is shown in Figure 3-1, and the ML tree is shown in Figure 3-2 (ln score=-1071.9). In the ITS-based phylogeny, isolates were in a highly supported clade (93% MP, 98% ML) with equivocal placement either within (ML analysis) or sister to the rest of (MP analysis) the isolates. Several clades were resolved with low to moderate bootst rap support. Clade 1 (40% MP) contained all mango isolates from blighted panicl es and immature fruit, as well as most leaf isolates. Subclade 1a was well supported (86% MP, 61% ML) and c ontained four leaf isolates 107

PAGE 108

and three immature fruit isolates from various gr oves. Clade 2 (75% MP) c ontained most isolates from mature fruit, peduncles and leaves, as well as single isolates from banana, guava and passionfruit. The remaining isolates of did not resolve into a specific clade in the ingroup. The MAT1-2 locus was anal yzed in 58 accessions of sensu lato, most of which were from mango (Table 3-2). and were outgroup taxa. The data set included 216 ch aracters. MP analysis resulted in 28 MPT with a tree score of 146 (CI=0.966, RI=0.981), one of which is shown in Figure 3-3. ML analysis resulted in the tree in Figure 3-4 (ln score=-1166.49). Clade 1 in the MAT1-2 analysis contained the same isolates as Clade 1 in the ITS phylogeny, and resolved in a larger clade (92% MP) th at contained isolates of and isolate Gua 3. It was moderately suppor ted in MP analysis (72% bootstrap), but the ITS subclades 1a and b were not resolved. Al l other isolates did not resolve into separate clades, but were part of the general ingroup clade. The orga n sources for clades of mango isolates were similar in the MAT1-2 and ITS phylogenies. The unrooted analysis of the CGTT5 region included 58 accessions of sensu lato. The data set contained 535 characters and MP analysis resulted in two MPT with a treelength of 73 (CI=0.959, RI=996) (Figure 3-5). ML analysis resulted in a tree of similar topology with a ln score of 1166.49 (Figure 3-6). Isolates of and Gua3 resolved with those in Clade 1 in the MAT phylogeny (100%MP, 99% ML). However, isolates from ITS Clade 2 also resolved with these isolates. Subclade 1a, similar to that in the ITS phylogeny, was resolved with CGTT5 data (82% MP, 79% ML), but it contained several different isolates; only isolates Cg34, Cg141, Cg164, and Cg165 were comm on to subclade 1a in both the ITS and CGTT5 phylogenies. CGTT5 data resolved two additional clades. Clade 2 had high MP support 108

PAGE 109

(92%), and included a subset of eight isolates from the ITS Clade 2. Clade 3 contained the remaining isolates in th e analys is (86% MP, 87% ML), except for two from carambola, Cm1 and Cm3. In general, Clade 1 was resolved in all phy logenies and was well supported in the MAT1-2 and CGTT5+MAT analyses (Table 3-4). It was comprised of only mango isolates, including all from inflorescences and immature fruit, and 59 to 63% of the leaf isolates, depending on the region. However, a low percentage of isolates from mango fruit ( 27 to 33%) and peduncles (0 to 14%) resolved in this clade. The remaining isolates of resolved differently in the different analyses without clear location or host associations. The ITS Clade 2 was resolved with strong support within the larger general clade, and contained 21 isolates, including seven from peduncles (78% of that total) and six fr om mature fruit (55%). In the MAT1-2 phylogeny, about one-half the isolates were not resolved, and included six of th e mature fruit isolates (67%), all peduncle isolates, and isolates from other fruit hosts. Four of the peduncle isolates (57%) were resolved in the Clade 2 CGTT5 phylogeny. Less than half the isolates from any mango organ were found in any other clade/phylogeny. Data-set-based congruence measures suggested th at all sets were incongruent (Table 3-3). The MAT and ITS data were significantly heter ogenous, but with less significance than the ITS and CGTT5 data sets. Using th e 70% bootstrap trees as a limit, only the MAT1-2 and CGTT5 data sets did not display signifi cant levels of incongruence or in stances of hard incongruence (Table 3-5). Therefore, these two data sets were combined and analyzed. Individual clades (MAT1-2 Clade 1, MAT1-2 Clade 1+ +Gua3, ITS Clade 1-1a, ITS Clade 2, CGTT5 Clade 1-1a, CGTT5 Clade 2, and CGTT5 Clade 3) were also used as 109

PAGE 110

constraints to determine where cases of hard inco ngruence were present betw een data sets (Tab le 3-5). With the ITS topology constrained on the MAT data set, only Clade 2 produced barely significant incongruence. ITS Clade 2 was significantly incongruent when constrained on the CGTT5 data set. Clade 1 and subclade 1a were congruent over every data set. The CGTT5+MAT combined data set had 54 ingroup accessions with and as outgroup taxa. MP analysis resulted in 63 trees with a length of 224 (CI=0.91, RI=0.986), one of which is shown in Figure 3-7. The ML tree is shown in Figure 3-8 (ln score=2075.2). Clade 1 and subclade 1a were highly supported (99/97% and 87/85% MP/ML, respectively), and the monophyly of Clade 1+ +Gua3 was moderately supported (78% MP). Clades 2 and 3 were identical to those resolved in the CGTT5 phylogeny, and had moderate support (76% MP, 74% ML and 69% MP 68% ML, respectively). A new clade, which contained carambola isolates Cm1 and Cm3, reso lved with lower support (60% MP, 66% ML). Conidia of the 13 isolates that were examined were typical of : straight, cylindrical and sometimes tapered toward the base or constricted in the middle, and from10-20 x 3-6 m, with an average of 13-15 x 4 m (Table 3-6, Figure 3-9). Significant differences existed among isolates, but there was no relationship betw een conidium size and genetic clade (Table 37). Hyphopodia were irregular (lobe d) or clavate-shaped and smooth (Table 3-6, Figure 3-10). With the exception of isolate Cg136, isolates from Clade 1 had sm ooth hyphopodia, whereas those from other clades ha d irregular hyphopodia. Hyphopodia range d from 5-16 x 4-11 m in size with an average of 8-10 x 6-9 m, which was slightly shorter than that reported by Sutton (1992). Significant differences also existed in hyphopodia size, but without relationships between size and genetic clade (Table 3-7). 110

PAGE 111

During blossom blight experime nts, isolates had highly significant impacts on AUDPC and ymax ( <0.0001) (Table 3-8). Also, with the exception of Cg136, isolates in Clade 1 caused significantly greater dise ase than isolates in Clades 2 a nd 3. Less disease developed with Cg136 than for other Clade 1 isolates in Experiment 1 and 2, probably due to the limited numbers of conidia that this isolate produced during thes e experiments and the corresponding low volumes of inoculum and poor panicle cove rage that resulted. Isolates in Clades 2 and 3 did not produce significant disease compared to the water cont rol, and were considered nonpathogenic. In general, isolates in Clade 1 caused disease se verities above 50% 28 days after inoculation (ymax), whereas isolates in Clades 2 and 3 caused seve rities less than 20% (Tab le 3-8). All pathogenic isolates produced similar blossom blight symptoms (Figure 3-11). Numerous small lesions coalesced, resulting in large areas of necrosis. Necrotic panicle tips curled downward to produce a shepherds crook symptom. Florets often senesced, and when severi ties exceeded 75% the entire panicle usually senesced. Leaf inoculation experiments produced similar results: isolates in Clade 1 caused greater disease than isolates in Clades 2 and 3, although these differences were not as pronounced as in the blossom experiments. In both experiments, block and treatment were significant ( <0.0001) (Table 3-9). Except for isolate Cg138 in Experiment 1, all Clade 1 isolates caused more severe leaf anthracnose than the water control. All other isolates of were slightly to not pathogenic, and caused disease severities sim ilar to those in the wa ter control treatments. With the exceptions of isolates Cg134 and Cg129, is olates in Clade 1 cause d severities greater than 20%, whereas severities for all other is olates were less than 10% (Figure 3-12). For experiments with detached fruit, isolat es generally had significant impacts on AUDPC and lesion diameter (Table 3-10). Although all isolates of produced greater 111

PAGE 112

disease than the water control, the significance of and rank orders for disease developm ent were inconsistent for isolates in the different experiments (Figure 3-13). Average lesion diameters after 7 days were approximately 10-20 mm, but reached 30 mm. Results from attached fruit inoculations genera lly agreed with resistance levels that have been reported for the tested cultiv ars (Crane et al. 2003), in that ymax values for the susceptible Haden were higher (0.1-0.87) than for the mode rately resistant Tommy Atkins (0.1-0.34) and Van Dyke (0.05-0.26) (Table 311). On Haden, isolate tr eatments were significant for AUDPC and ymax (respectively, =0.0013 and 0.0018), but they were not significant for Tommy Atkins and were significant on Van Dyke for only AUDPC ( =0.0224). Isolate Cg141 (Clade 1) caused the greatest disease on all three cultivars, and isolat es in Clade 1 produced greater disease than other isolates, alt hough these differences were not alwa ys significant (Table 3-11). Although more disease developed on inoculated than on control fruit (Figure 3-13), these differences were usually not significant due to va riation in these data. This study examined the occurrence and iden tification of a mango-specific taxon within sensu lato. Alakahoon et al. (1994), Hayden et al. (1994) and Hodson et al. (1993) had suggested such a taxon, but relied on s cant data obtained with old techniques. Likewise, tissue specificity ha d been previously indicated am ong mango isolates with isozyme and RAPD profiles (Davis 1999, Gantotti and Davis 1993). Considering the importance of these research topics, a modern and comprehensive re -examination of previous work was warranted. The current study used both phylogenetic and biol ogical data to investig ate the presence of distinct evolutionary species associated with mango anthracnose. The ecological and epidemiological dynamics of the three clades that we re resolved in this work is discussed below. 112

PAGE 113

With ITS, MAT1-2, and CGTT5 sequence data a m ango-specific biot ype (Clade 1) was resolved, and was well supported in both the MAT1-2 and CGTT5+MAT phylogenies. Clade 1 contained only mango isolates, and isolates from mango inflorescences and asymptomatic immature fruit were found in no other clade. Is olates from ripe fruit and young leaves of mango were found in Clade 1, but were also found in Clad es 2 and 3; isolates from necrotic peduncles were restricted to the latter clades. Isolates out side Clade 1 could be considered generalists as they came from mango and several other hosts. The prevalence of isolates in Clades 2 and 3 on mature mango fruit and peduncles suggests that host-specific factors were no longer present when they were recovered from these tissues. Unli ke isolates in Clade 1, these isolates did not consistently resolve across all data sets, and hard incongruenc e was found among the clades that did resolve (Clade 2 between the ITS and CGTT5 data sets). While these incongruencies could be due to lineage sorting among recently diverg ed lineages, they could also indicate that recombination has occurred among these strains. The placement of relative to the other clades was equivocal in this study. In the ITS MP phylogeny, isolates formed a strong ly supported clade sister to the clade, but in the other phyl ogenies it resolved within Since the placement of within the larger, more general clade differed by locus, it is difficult to speculate on its re lationship to other clades. Carambola isolates Cm1 and Cm3 resolved as sister to all other isolates of although this placement was not always statistically supported. It is possible that isolates of causing anthracnose on cara mbola form another host113

PAGE 114

specialized lineage that has diverged from the host-general isolates of sensu lato. Further characterization of isolates from carambola is necessary to investigate this question. Pathogenicity experiments supported the separa tion of the mango-specific from the hostgeneral clades. Isolates from blossom blight an d leaf anthracnose samples all fell in Clade 1, whereas isolates from all three clades were associated with fruit anthracnose. Since isolates in Clade 1 also caused severe symptoms on inflor escences in pathogenicity studies, this hostspecific group appears to play a role in blosso m blight development in South Florida. This contradicts previous reports fr om South Africa that attributed blossom blight of mango to spp., and indicated that caused only small lesions on mango panicles (Darvas 1993, Lonsdale and Kotz 1993). Since the role of spp. in blossom blight development in South Florida has not been investigated, these fungi may also be involved with th is disease here. However, based on the slight symptoms that caused on mango panicles in South Africa, it is possible that host-general isolat es outside Clade 1 were used in that work. Although additional work would be needed in South Africa to confir m this hypothesis, results from the present study illustrate the importance of cryptic lineages in the mango anthracnose and blossom blight pathosystems: erroneous conclusi ons about the etiology and epidem iology of these diseases are clearly possible when pathogens and nonpathogens are obscured and experimental work is conducted with generalist rather than host-specific isolates. Arauz (2001) indicated that anthracnose on mango fruit is a monocyclic disease that begins with latent infection of immature fruit, similar to that described for avocado fruit (Prusky and Lichter 2007, 2008). On avocado, symptoms deve lop after fruit maturation in response to several pathogen and host factor s, including the production of th e enzymes laccase and pectate 114

PAGE 115

lyase (Yakoby et al.2001b, Guetsky et al.2005, Kramer-Haimovich et al. 2006), and the decrease of antifungal compounds in the fruit pericarp (Guetsky et al. 2005). In ma ngo, Droby et al.(1986, 1987) described preformed antifungal re sorcinols in the mango fruit pericarp that were associ ated with latent infections of cause of black spot Fungitoxic levels of two compounds, 512-cis-heptadecenyl resorcinol and 5pentadecenyl resorcinol, were present in the pe el of Haden prior to ripening, and as ripening commenced, resorcinol concentra tions declined and postharvest decay began (Prusky and Keen, 1993). To date, 15 resorcinols have been iden tified in mango peels (Kndler et al. 2007). Recently, Hassan et al. (2007) corr elated concentrations of thes e compounds and the levels of anthracnose that developed on fruit of several different mango cultivars. Because isolates in Clade 1 were the only ones that infected panicles, leaves and immature fruit, it is possible that they ha ve a higher tolerance to these antifungal compounds than other lineages. Although it appear s only isolates in Clade 1 laten tly infected immature fruit, isolates from all clades were recovered from, and caused anthracnose on, ripening fruit. Thus, anthracnose that is caused by isolates in Clades 2 and 3 does not appear to originate from latent infections. Work to confirm this observation with a larger set of isolates is warranted as it contradicts a long held view that anthracnos e on mango fruit always results from latent infections. The sensitivity of is olates in the different clades to resorcinols that are found in immature mango peels should also be determined, as it may significantly cl arify the etiology of this important disease. Furthermore, resorcinol concentrations in other mango organs (leaves and inflorescences) should be assessed to investigate whether they play a role in infection by isolates in Clades 1, 2 and 3 and the disease that they cause or do not cause on different mango tissues. For example, are Clade 1 isolates capable of cau sing blossom blight and leaf anthracnose due to 115

PAGE 116

their insensitivity to high resorcinol concentrat ions that occur in those tissues? A more thorough understanding of these relationships is needed. In addition to assessing resorcinol sensitivity in isolates in the different clades, other hostpathogen interactions should be investigated. Th e production of pectate lyase, alkalization of host tissue in the area of infec tion, and sugar and ammonia production have been reported as important factors in the infection of, and di sease development on, a vocado fruit (Yakoby et al.2001b, Guetsky 2005, Kramer-Haimovich et al.2006); whether they also play a role in the mango pathosystems should be examined. Whethe r isolates in Clade 1 produce a phytotoxin, such as that reported for the /yam pathosystem (Abang et al. 2009), should also be assessed, as it could explain the seve re symptoms that these isolates produce on mango panicles. By combining results from isolation, pathogenicity and phylogenetic studies, the etiology, ecology and epidemiology of the mango pathosystems were clarified in the present work, and areas of research that were needed were indentified. Pathogenic species of are considered hemibiotrophs, with a short period of biotrophic colonization after infection (with a longer biotrophic period when fruit are latently infected), followed by a switch to necrotrophic growth that coincided with symptom developm ent (Bailey et al. 1992). Additionally, pathogenic species of are often associated with endophyti c or saprotrophic host colonization of alternative hosts (Perez et al. 2005, Freeman et al. 2001, Photita et al. 2005), indicating that individuals in this genus can al ternate among endophytic, saprotrophic and pathogenic lifestyles. Direct evidence supporting this hypothesis ha s been presented by several studies. In a phylogenetic study of species isolated from there was no resolution separating pathogenic, saproptrophic, and endophytic strains, indicating that isolates 116

PAGE 117

from different clades could fill different ecological niches depending on environmental factors (Promphutta et al. 2007). Direct evidence for such flexibility was pr esented by Promphutta et al. (2007). In a phylogenetic study of spp. that were isolated from pathogenic, saproptrophic and endophytic strain s were not resolved, indicating that isolates from different clades could fill different ecological niches de pending on environmental factors. A strain of that was transformed with the GFP protei n colonized corn roots and systemically colonize aerial portions of th e plant (Sukno et al. 2008), suggesti ng that anthracnose can also result from systemic infections. The asymptomat ic root infection by this strain exhibited many similarities to infection by r ecognized root pathogens, sugges ting that many fungi, even those recognized as foliar pathogens, maintain the abil ity for root infection. Another hemibiotrophic foliar pathogen, also infected roots and systemically colonized rice plants. Many of the rice genes that were upregulated duri ng this process are similar to those that are upregulated during colonizati on by endophytic arbuscular mycorrhizal fungi (Sesma and Osborne 2004, Gimil et al. 2005). Freeman and Rodriguez (1993) produced a single-gene mutant of path-1, that was nonpathogenic but still able to colonize plants endophytically. Thus, the genetic differenes between endophytes and pathogens may be slight, and pathogenicity and host colonization are controlled by different genes. Based on the above studies, it appears that the distinction betw een endophyte and pathogen depends more on symptom development than on di fferences in host colonization. Indeed, Schulz et al. (1999) suggested that both endophyte-plant and pat hogen-plant interactions are characterized by mutual-antagonism, but in the case of pathogens, the interaction is imbalanced and results in disease development. They repor ted a greater defense re sponse in endophyte-host 117

PAGE 118

reactions, suggesting endophyt es are only able to overcome defens es to the point of infection and colonization, but not to the extreme of sym pto m development. Pathogens have an increased ability to overcome plant defenses (as in laccase production by in avocado to overcome antifungal dienes in peels for developmen t of fruit anthracnose; Guetsky et al. 2005). Better understanding is needed of the attributes and circum stances that distinguish the endophytic and pathogenic behavior of fungi. The present study supports the hypothesis that strains of are ecologically flexible. The recovery of an isolate in Clad e 1, En2, from an asymptomatic, developing mango leaf suggests that pathogenic individuals may colonize mango as endophytes. In addition, nonhost-specific individuals coloni zed necrotic mango tissue as saprotrophs. Not only were nonpathogenic isolates in th e generalist clade recovere d from leaf lesions, but also sporulated on necrotic leaf le sions that were caused by other pathogens (data not shown). The same pattern was reported on coffee, where avirulent was recovered along with from lesions of coffee berry disease (Beynon et al. 1995, Derso et al. 2003). Derso et al. (2003) suggested that had limited saprotrophic capabilities, based on its specialized carbon metabolism (Waller et al. 1993), a nd that it was displaced in lesions it caused on coffee berries by saprophytic strains of A similar situation may occur with the mango-specific Clade 1, as isolates from ot her clades were often r ecovered from old leaf lesions or ripening fruit, but not from immature fruit or panicles. The extent to which generalist, saprotrophic strains of are able to colonize diverse hosts and host tissues warrants further investigation as this information could dramatically increase our understanding of the ecology and epidemiology of pathogenic and non-pa thogenic interactions in this genus. 118

PAGE 119

By identifying host-specific and host-general populations of and the tissues that they colonize, the present st udy also sheds light on the development and epidemiology of i mportant anthracnose diseases on mango. Most studies on tropical fruits that have investigated host range indi cated that isolates were often most aggressive on the original host, but that generalists caused lesions on a wide range of hosts (Alahakoon et al. 1994, Hayden et al. 1994, Sanders and Korsten 2003). Although host range per se was not assessed in the present study, the phylogentic da ta corroborate the existence of a general, non host-specific population of that causes anthracnose symptoms on diverse fruit hosts. Findings in this study also have ramifications for resistance assessments and which isolates of the pathogen should be used in artificial inoculations. Although it appears that only host-specific populations of cause blossom blight and infe ct fruit latently, diverse populations are capable of causing an thracnose on mature and ripening fruit. Work is needed to assess the reservoirs and levels of the differe nt generalist and hostspecific populations of that occur on mango and the relative th reats that they pose to disease development on different organs. The three DNA regions used in this study repr esented ones that have traditionally exhibited the ability to resolve clades at different levels of classificat ion, as well as a third novel region chosen for its high number of polymorphic nucleo tide sites. The ITS1-5.8-ITS2 rDNA region is most commonly used in phylogenetic studies of fungi and is generally able to resolve clades at the species level (Hillis and Dixon 1991, Nilsson et al. 2008). While speci es can generally be resolved using the ITS region, intraspecific varia tion is very low, which should make the region a good indicator of cryptic speciation. Several studies suggested that sequence identity of 97% indicates an intraspecific rela tionship (Iwen et al. 2002, Ciardo et al. 2006, Nilsson et al. 2008). 119

PAGE 120

However, assigning a quantitative limit on sp ecies definitions for all fungi is an oversimplification, as the am ount of intraspecific variation at a specific locu s can vary among genera (Bridge 2002). Within a genus, the level of sequence identity at which two taxonomic units are considered to be distinct species there should be consiste nt. In this study, there were only 11 PIC within sensu lato in the ITS phylogeny. Isolates of differed from the generalists in Clades 2 and 3 by only three nucl eotides, and the mango-specific isolates in Clade 1 differed by only one to three nucl eotides. Similarly, isolates of from strawberry, cyclamen and date palm differed by three nucleot ides (MacKenzie et al. 2008), and isolates of differed by two nucleotides (Sreenivasapra sad et al. 1996a). Thus, the lack of polymorphic sites in the ITS region limits its util ity to resolve such closely related lineages. In contrast, ITS differentiated sensu lato from other species of species that were recovered from fruits in south Florida. For example, isolates of that were recovered from mango leav es (see Chapter 4) differed from by 48 nucleotide sites and 11 in del sites, isolates of that were recovered from papaya differed by 32 nucleotides and four indel sites, and isolates of from passionfruit and differed by 26 nucleotide substitutions and four indel sites (see Appendix B). A gr aphical representation is shown in Figure 3-14 of the numbers of polymorphisms that are presen t in the ITS regions between and these taxa. Mating type sequences have been used in phylogenetic analysis of many fungal taxa, including (ODonnell et al. 2004), (Du et al. 2005), (Barve et al. 2003), (Turgeon 1998), and and (Poggeler 1999). The former four studies have all used the HMG-box region of the MAT1-2 idiomorph, and indicated 120

PAGE 121

that this region contained more phylogenetically inform ative ch aracters among closely related taxa than the ITS region. They suggested that this locus could be used to build highly resolved phylogenies among closely related accessions. While between-species vari ation was relatively high, intraspecific variation was very low, en abling further resolving power among species. The MAT loci are unique in that instead of alleles, alternate forms of the locus are designated idiomorphs because the genes inhabi ting identical positions on the chromosome are not similar and in fact encode different prot eins (Moore 1998). The locus does not recombine, and its inheritance has been compared to th e human Y chromosome (Merino 1996). The pattern of high interspecific and low intraspecific va riability in mating genes is common in many eukaryotes, and rapid evolution in mating ge nes is thought to accompany speciation events (Ferris et al. 1997). Variation within a biological species is limited by recombination and selection mechanisms. Once speciation occurs and two populations no lo nger recombine, the MAT locus starts to diverge. Therefore, this re gion could be used to marry the phylogenetic and biological species concepts in classification. However, a disa dvantage of using mating type genes in heterothallic species is that only one of the idiomo rphs would be present in an individual and that the locus th at is utilized would not be am plified in all mating populations. A previous study evaluated the use of th e HMG box region for phylogenetic analysis among species (Du et al. 2005). They f ound that the MAT phylogenies followed previously reported patterns (hi gh interspecific and low intrasp ecific variation), had a stronger phylogenetic signal than ITS, and was better able to resolve relationships among closely related accessions. Since resolution of closely related strains of was desired in the present work, a highly polymorphic locus was needed. To that end, the CGTT5 locus was cloned and 121

PAGE 122

amplified as a sequence characte rized amplified region (SCAR) m a rker (McDermott et al. 2004). BLAST searches indicated that this region displays homology to sequences coding for hypothetical proteins in and The CGTT5 data set had 61 PIC among isolates, compared to 13 and 11, respectively, for the MAT and ITS data sets. Clad es resolved for the CG TT5 region also showed higher bootstrap support. In the resulting phylogeny, all but three isolates resolved into wellsupported clades, compared to the ITS and MAT re gions where large numbers of isolates did not resolve further into clades within the ingroup. This suggests that this region does provide a stronger phylogenetic signal, and this approach to developing lo ci is well-suited for analyses involving closely related accessions. Martinez-Culebras et al. (2002) Du et al. (2005), Bridge et al. (2008), and MacKenzie et al. (2008), and the present results indicated that populations of sensu lato have diverged several times to specialize on specific hosts or tissues; for example, on banana, on strawberry, on coffee and Clade 1 on mango. However, these lineages diverged relatively recently from and it is not entirely clear how and at what taxonomic levels they shoul d be distinguished from this species. Carbone and Kohn (2001) discussed a populati on-species interface that distinguishes between recent divergence of populations and ol der speciation events. Identifying the point on the evolutionary continuum at which populations become species will always be artificial. A consensus has not been reached on where species boundaries exist in Taylor et al. (2001) suggested that phylogenies could be suppl emented with biological data (ultrastructural characteristics, mating compatibil ity, pathogenicity) to erec t evolutionary species, defined by Wiley (1978) as single lineage of ancestral descendant popul ations of organisms 122

PAGE 123

123 which maintains its identity from other such lineages and which has its own evolutionary tendencies and historical fate. Biological characters may or may not be of evolutionary significance. For example, hos t-specificity could clearl y influence the ecology of phytopathogens, and probably focused the evoluti on of the existence of the mango-specific taxon that was identified in the presen t study. And although conidial dime nsions were not distinctive in the present work, hyphopodial morphology tended to distinguish isolates, in that those for isolates in Clade 1 were usually smooth and clavate-shaped vs the lobed appearance of those for isolates in Clades 2 and 3. This study used a multifaceted approach combining phylogenetic, pathogenicity and morphology data to characterize se veral lineages that were associ ated with diseases of mango that are caused by sensu lato in south Florida. The results provide new insight into, and suggest several new avenues for, research of the etiology and epidemiology of these diseases.

PAGE 124

Table 3-1. Num ber of trees, samples and single-spore isolates collected in the hier achical sampling scheme from various mango organs and tropical fruit hosts. Inflorescences Leaves Immature fruit Mature fruit Peduncles Avocado Banana Carambola Guava Papaya # trees sampled 5 47 16 36 27 4 6 2 2 7 # organs sampled 11 104 26 60 52 9 6 2 3 7 # isolates stored 25 204 33 101 59 26 18 4 8 30 124

PAGE 125

Table 3-2. Accessions included in study. Accession Tissue origina Recovery location Cladeb Genbank accession number ITS MAT CGTT5 En2 Leaf endophyte TREC, Keitt 1 GQ373201 GQ925061 GQ924960 Cg23 Leaf TREC, mixed --GQ373208 ----Cg52 Leaf TREC, mixed 1 GQ373210 GQ925017 GQ924968 Cg60 Leaf TREC, Keitt ----GQ925019 --Cg101 Leaf TREC, Keitt 3 GQ373215 GQ925023 GQ924974 Cg129 Leaf TREC, Keitt 2 GQ373218 GQ925026 GQ924977 Cg136 Leaf TREC, Keitt 1 GQ373225 GQ925032 GQ924982 Cg141 Leaf USDA 1 GQ373228 --GQ924985 Cg142 Leaf USDA 1 GQ373229 GQ925035 GQ924986 Cg145 Leaf TREC, Keitt 1 GQ373230 GQ925036 GQ924987 Cg151 Leaf TREC, mixed 3 GQ373234 GQ925042 GQ924991 Cg160 Leaf TREC, mixed 3 GQ373243 GQ925049 GQ925000 Cg161 Leaf TREC, mixed 3 GQ373244 GQ925050 GQ925001 Cg162 Leaf Commercial grove --GQ373245 GQ925051 --Cg163 Leaf USDA 1 GQ373246 GQ925052 GQ925002 Cg165 Leaf Commercial grove 1 GQ373248 GQ925054 GQ924004 Cg166 Leaf Commercial grove --GQ373249 ----Cg167 Leaf Commercial grove 3 GQ373250 GQ925055 GQ925005 Cg170 Leaf Commercial grove 1 GQ373252 GQ925057 GQ925007 Cg171 Leaf Commercial grove --GQ373253 ----Cg137 Inflorescence TREC, Keitt 1 GQ373226 GQ925033 GQ924983 Cg138 Inflorescence TREC, Keitt 1 GQ373227 GQ925034 GQ924984 Cg147 Inflorescence TREC, Keitt 1 GQ373231 GQ925037 GQ924988 Cg148 Inflorescence TREC, Keitt 1 GQ373232 GQ925038 GQ924989 Cg34 Immature fruit TREC, mixed 1 GQ373206 GQ925014 GQ924965 Cg35 Immature fruit TREC, mixed 1 GQ373209 GQ925015 GQ924966 Cg36 Mummified fruitlet TREC, m ixed 1 GQ373207 GQ925016 GQ924967 Cg66 Immature fruit TREC, Keitt 1 GQ373212 GQ925020 GQ924970 125

PAGE 126

Table 3-2. Continued. Accession Tissue origina Recovery location Clade Genbank accession number ITS MAT CGTT5 Cg75 Immature fruit TREC, mixed 1 GQ373214 GQ925022 GQ924973 Cg135 Immature fruit TREC, Keitt 1 GQ373224 GQ925031 GQ924981 Cg158 Immature fruit Commercial grove 1 GQ373241 GQ925041 GQ924998 Cg131 Mature fruit TREC, Keitt 1 GQ373223 GQ925029 GQ924979 Cg132 Mature fruit TREC, Keitt --GQ373219 ----Cg152 Mature fruit TREC, Keitt 1 GQ373235 GQ925040 GQ924992 Cg153 Mature fruit TREC, Keitt 2 GQ373236 GQ925043 GQ924993 Cg154 Mature fruit USDA 2 GQ373237 GQ925044 GQ924994 Cg155 Mature fruit USDA 3 GQ373238 GQ925045 GQ924996 Cg157 Mature fruit USDA 2 GQ373240 GQ925047 GQ924997 Cg169 Mature fruit TREC, mixed 1 GQ373251 GQ925056 GQ925006 Cg16 Peduncle TREC, mixed --GQ373204 ----Cg20 Peduncle TREC, Keitt 2 GQ373205 GQ925013 GQ924964 Cg55 Peduncle TREC, Keitt 2 GQ373211 GQ925018 GQ924969 Cg72 Peduncle TREC, mixed 1 ----GQ924971 Cg73 Peduncle TREC, Keitt 2 GQ373213 GQ925021 GQ924972 Cg127 Peduncle TREC, Keitt 3 GQ373216 GQ925025 GQ924975 Cg128 Healthy peduncle TREC, Keitt 3 GQ373217 GQ925027 GQ924976 Cg130 Peduncle TREC, Keitt 3 GQ373222 GQ925028 GQ924978 Cg133 Healthy peduncle TREC, Keitt --GQ373220 ----Cg134 Peduncle TREC, Keitt 2 GQ373221 GQ925030 GQ924980 Cg149 Peduncle TREC, Keitt 1 GQ373233 GQ925039 GQ924990 Cg156 Peduncle Commercial grove 3 GQ373239 GQ925046 GQ924995 Avo1 Avocado TREC 3 GQ373189 GQ925008 GQ924951 Avo2 Avocado TREC 3 GQ373190 GQ925009 GQ924952 Avo3 Avocado TREC --GQ373191 ----Ban1 Banana TREC GQ373192 GQ925010 GQ924953 Ban2 Banana TREC GQ373193 GQ925011 GQ924954 Ban3 Banana TREC GQ373194 --GQ924955 126

PAGE 127

Table 3-2. Continued. Accession Tissue origin Recovery location Clade Genbank accession number ITS MAT CGTT5 Ban4 Banana TREC 3 GQ373195 GQ925012 GQ924956 Cm1 Carambola TREC carambola GQ373196 GQ925058 GQ924957 Cm2 Carambola TREC 3 GQ373197 GQ925059 GQ924958 Cm3 Carambola TREC carambola GQ373198 GQ925060 GQ924959 Gua1 Guava TREC 3 GQ373198 GQ925062 GQ924961 Gua2 Guava TREC 3 GQ373199 GQ925063 GQ924962 Gua3 Guava TREC 1 GQ373200 GQ925064 GQ924963 Pap16 Papaya TREC ----GQ925065 --Pap17 Papaya TREC ----GQ925066 --Pas4 Passionfruit Farm --GQ373202 ----Farm --GQ373203 ----Piper3 (outgroup) ------DQ003120 DQ002843 --(outgroup) ------DQ003104 DQ002828 --a Unless another host is specified, tissue origin indicat es mango organs from which an isolate was recovered. b Clade designations follow the CGTT5+MAT1-2 maximum parsimony phylogeny 127

PAGE 128

Table 3-3. Details of phylogenetic analyses. Data Set ITS MAT CGTT5 ITS+MAT CGTT5+MAT ITS+CGGT5 # accessions 65 58 58 --54 --# characters 557 216 535 --751 --# PICa 27 39 61 --84 --# PIC in ingroupb 11 13 61 ------# trees 11 28 2 --63 --# steps 112 146 73 --224 --CI 0.938 0.966 0.959 --0.951 --RI 0.949 0.981 0.996 --0.986 --ILD -valuec ------0.008 0.002 0.002 IMF d ------0.38 0.36 0.52 Clade Bootstrap supporte 1 49/* 72/* */* --99/97 --1a 91/61 */* 97/85 --87/85 --2 50/* */* 96/* --76/74 --3 */* */* 86/87 --69/68 --aParsimony informative characters b Number of parsimony informative characters among accessions of sensu lato (excluding outgroup taxa) Incongruence length difference (ILD) d Mickevich-Farris index (IMF) e Bootstrap support values for given clades, MP analysis/ML analysis, *<40% 128

PAGE 129

Table 3-4. Num ber of isolates by ti ssue type and phylogenetic clade. ITS data set Tissue Clade 1 Clade 2 CG Total 1 1a Clade 1 total Leaf 8 (39)a 4 (20) 12 (59) 5 (26) 2 (11) 19 Inflorescence 4 (100) 0 (0) 4 (100) 0 (0) 0 (0) 4 Immature Fruit 4 (57) 3 (43) 7 (100) 0 (0) 0 (0) 7 Mature Fruit 3 (27) 0 (0) 3 (27) 6 (55) 2 (18) 11 Peduncle 1 (11) 0 (0) 1 (11) 7 (78) 1 (11) 9 Mature fruit (other hosts) 0 (0 ) 0 (0) 0 (0) 3 (30) 7 (70) 10 Total 19 7 26 21 12 59 MAT1-2 data set Tissue Clade 1 CG --Total Leaf 10 (63) 6 (37) --16 Inflorescence 4 (100) 0 (0) --4 Immature Fruit 7 (100) 0 (0) --7 Mature Fruit 3 (33) 6 (67) --9 Peduncle 0 (0) 8 (100) --8 Mature fruit (other hosts) 0 (0) 8 (100) --8 Total 24 28 --52 CGTT5 data set Tissue Clade1 Clade2 Clade 3 Total 1 1a Clade 1 total Leaf 4 (27) 5 (33) 9 (60) 1 (7) 5 (33) 15 Inflorescence 2 (50) 2 (50) 4 (100) 0 (0) 0 (0) 4 Immature Fruit 5 (71) 2 (29) 7 (100) 0 (0) 0 (0) 7 Mature Fruit 2 (20) 1 (10) 3 (30) 3 (30) 4 (40) 10 Peduncle 1 (14) 0 (0) 1 (14) 4 (57) 2 (29) 7 Mature fruit (other hosts) 0 (0) 0 (0) 0 (0) 0 (0) 6 (100) 6 Total 15 9 24 8 17 49 aNumber of isolates followed by percentage of total isolates from that tissue type 129

PAGE 130

Table 3-5. Topography-based congruence analyses Tree length Gain Loss Net a Data sets with constraints 1. ITS 112 124120120.0023 124120120.0023 1154130.2568 C. musae 124120120.0023 123110110.0384 123110110.0384 1154130.2568 1154130.2568 112000---b 1154130.2568 2. MAT1-2 146 1494040.0455 1494040.0455 145000--145000--1494040.0455 1516060.0339 1483030.1797 145000--145000--145000--1472020.1573 3. CGTT5 73 741010.3173 741010.3173 73000--C. musae 775140.1025 130

PAGE 131

Table 3-5. Continued. Tree length Gain Loss Net a Data sets with constraints 89182160.0003 89182160.0003 752020.1573 752020.1573 87162140.0010 a Wilcoxon Rank Sum test performed on # steps gained/lost to determine instances of topological incongruence; <0.05 is significant Wilcoxon Rank Sum test not performe d when gain/loss is equal to 0 131

PAGE 132

Table 3-6. Morphological de scription of conidia and hyphopodia of 13 isolates. Isolate Clade a Conidia shape Conidia dimensions Hyphopodia shape Hyphopodia dimensions # clavate # irregular Cg131 1 Cylindrical, sometimes constricted toward base or constricted 11 (14) 20 x 4 (4) 5 Scarce; clavate, smooth to rarely irregular, lobed 6 (10) 16 x 6 (7) 9 196 Cg135 1 Cylindrical, sometimes constricted toward base or constricted 13 (15) 18 x 4 (4) 5 Clavate, smooth to rarely irregular, lobed 5 (10) 14 x 5 (7) 13 23 2 Cg136 1 Cylindrical, sometimes constricted toward base or constricted 10 (13) 18 x 4 (4) 4 Abundant; irregular, lobed to rarely clavate, smooth 6 (9) 13 x 5 (7) 11 124 22 520 223 322 Cg138 1 Cylindrical, sometimes constricted toward base or constricted 11 (14) 20 x 3 (4) 5 Scarce; clavate, smooth to sometimes irregular, lobed 6 (8) 11 x 5 (6) 8 15 10 Cg141 1 Cylindrical, sometimes constricted toward base or constricted 13 (14) 21 x 4 (4) 6 Scarce; clavate, smooth to rarely irregular, lobed 8 (10) 13 x 5 (6) 8 19 2 Cg164 1 Cylindrical, sometimes constricted toward base or constricted 11 (14) 18 x 4 (5) 5 Very rare; clavate, smooth to irregular, lobed 6 (9) 11 x 5 (6) 6 Cg129 2 Cylindrical, sometimes constricted toward base or constricted 13 (14) 16 x 4 (4) 5 Abundant; irregular, lobed to rarely clavate, smooth 6 (8) 11 x 5 (6) 10 Cg134 2 Cylindrical, sometimes constricted toward base or constricted 10 (14) 18 x 4 (4) 5 Abundant; irregular, lobed to rarely clavate, smooth 6 (9) 14 x 4 (6) 8 Cg157 2 Cylindrical, sometimes constricted toward base or constricted 10 (13) 16 x 3 (4) 5 Abundant; irregular, lobed to rarely clavate, smooth 6 (9) 14 x 5 (7) 9 Cg128 3 Cylindrical, sometimes constricted toward base or constricted 10 (14) 19 x 3 (4) 5 Abundant; irregular, lobed to sometimes clavate, smooth 6 (9) 16 x 6 (9) 16 10 15 132

PAGE 133

Table 3-6. Continued. Isolate Clade Conidia shape Conidia dimensions Hyphopodia shape Hyphopodia dimensions # clavate # irregular Avo1 3 Cylindrical, sometimes constricted toward base or constricted 11 (14) 16 x 4 (4) 5 Abundant; irregular, lobed to slightly irregular 6 (8) 13 x 5 (7) 10 025 Cg156 3 Cylindrical, sometimes constricted toward base or constricted 11 (13) 15 x 3 (4) 5 Scarce; irregular, lobed to sometimes clavate, smooth 6 (8) 13 x 4 (7) 10 312 124 Cg161 3 Cylindrical, sometimes constricted toward base or constricted 11 (14) 20 x 3 (4) 5 Abundant; irregular, lobed 8 (9) 14 x 5 (7) 11 a Clade designations follow the CGTT 5+MAT1-2 maximum parsimony phylogeny 133

PAGE 134

Table 3-7. Statistical comparison of conidia and hyphopodia dim ensions. Isolate Clade a Conidia Hyphopodia length width dimensions length width dimensions 1 14.05 0.23 bcdb 4.30 0.09 b 3.32 0.07 de 10.11 0.40 ab 6.94 0.26 bc 1.47 0.05 abc Cg131 Cg135 1 15.05 0.19 a 4.03 0.06 cde 3.77 0.06 ab 9.90 0.48 abc 7.40 0.39 b 1.40 0.08 abc Cg136 1 12.73 0.16 g 3.86 .03 ef 3.30 0.05 de 9.40 0.32 abcd 7.00 0.32 bc 1.40 0.07 abc Cg138 1 14.05 0.23 bcd 3.90 0.06 ef 3.63 0.07 bc 8.42 0.59 de 7.08 0.45 d 1.29 0.16 abc Cg141 1 14.20 0.22 bc 4.22 0.09 bc 3.41 0.07 d 10.36 0.32 a 6.67 0.16 bcd 1.58 0.07 a Cg164 1 13.60 0.18 def 4.56 0.08 a 3.04 0.08 f 9.06 1.07 bcde 5.94 0.31 d 1.55 0.22 a Cg129 2 13.98 0.15 bcde 4.23 0.08 bc 3.36 0.07 de 8.10 0.29 e 6.40 0.25 cd 1.28 0.04 bcd Cg134 2 14.05 0.17 bcd 4.15 .07 bcd 3.43 0.07 cd 9.00 0.36 bcde 6.60 0.17 bcd 1.42 0.11 abc Cg157 2 13.45 0.24 ef 3.94 0.09 def 3.49 0.10 cd 8.70 0.40 cde 6.55 0.25 bcd 1.37 0.09 abc Cg128 3 13.73 0.20 cde 4.00 0.09 de 3.51 0.09 cd 9.49 0.43 abcd 8.87 0.43 a 1.11 0.06 d Avo1 3 13.90 0.15 bcde 4.25 0.08 b 3.34 0.08 de 8.40 0.27 de 7.00 0.28 bc 1.26 0.08 cd Cg156 3 13.08 0.14 fg 4.24 0.09 bc 3.16 0.08 ef 8.42 0.59 de 7.08 0.45 bc 1.29 0.16 bcd Cg161 3 14.39 0.23 b 3.74 0.07 f 3.90 .09 a 9.80 0.37 abc 6.60 0.24 bcd 1.52 0.07 ab aClade designations follow the CGTT 5+MAT1-2 maximum parsimony phylogeny b Means and standard errors are reported; means were compared using Fishers LSD test; values in co lumns with same letter are not significantly different, <0.05 134

PAGE 135

Table 3-8. Area under the disease progress curve (AUDPC) and ymax values for blossom blight experiments, 2009. Exp. 1 Exp. 2 Exp. 3 Isolate Cladea AUDPC ymax b AUDPC ymax AUDPC ymax 1 8.89 2.49 abc 0.53 0.17 a 12.62 2.35 ab 0.67 0.10 ab 13.71 2.02 a 0.80 0.10 a Cg131 Cg135 1 11.22 1.91 a 0.64 0.12 a 14.64 2.57 a 0.82 0.09 a 13.28 3.75 a 0.85 0.11 a Cg136 1 6.37 0.12 b 0.18 0.07 b 9.55 1.13 b 0.55 0.09 b 9.04 2.15 ab 0.91 0.08 a Cg138 1 8.29 0.99 ab 0.45 0.06 a 12.31 2.68 ab 0.80 0.20 a 11.50 2.63 a 0.74 0.08 a Cg141 1 8.34 0.74 ab 0.51 0.07 a 10.98 1.75 ab 0.73 0.13 ab 9.50 1.90 ab 0.69 0.17 a Cg164 1 10.39 1.68 a 0.53 0.09 a 10.23 2.96 ab 0.48 0.13 b 6.51 1.23 b 0.75 0.25 a Avo1 3 2.57 0.63 c 0.13 0.04 b 0.87 0.17 c 0.06 0.01 c 0.85 0.70 c 0.07 0.06 b Cg156 3 2.61 2.02 c 0.12 0.09 b 2.06 0.85 c 0.14 0.08 c 0.89 0.47 c 0.08 0.05 b Cg161 3 1.09 0.51 c 0.05 0.03 b 1.00 0.17 c 0.06 0.01 c 1.30 0.38 c 0.09 0.03 b Cg128 3 2.09 0.79 c 0.13 0.06 b 0.51 0.14 c 0.02 0.01 c 0.42 0.16 c 0.02 0.01 b Cg129 2 1.55 0.78 c 0.08 0.04 b 1.04 0.42 c 0.10 0.05 c 0.12 0.05 c 0.01 0.01 b Cg134 2 2.58 1.41 c 0.20 0.14 b 0.72 0.20 c 0.02 0.00 c 0.18 0.07 c 0.01 0.01 b Cg157 2 1.28 .08 c 0.11 0.10 b 2.09 0.14 c 0.10 0.07 c 1.29 1.04 c 0.11 0.07 b Control --1.00 0.49 c 0.04 0.02 b 0.36 0.28 c 0.03 0.02 c 0.31 0.10 c 0.02 0.01 b LSD 3.53 0.231 4.66 0.249 4.91 0.234 = <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 a Clade designations follow the CGTT 5+MAT1-2 maximum parsimony phylogeny b ymax is the severity value for panicles 28 days after inoculation c Values in columns with same letter are not significantly different based on Fishers LSD, =0.05. 135

PAGE 136

Table 3-9. Area under the disease progress curve (AUDPC) and ymax values for leaf anthracnose experiments, 2009. Exp 1 Exp 2 Isolate Clade AUDPC ymax b AUDPC ymax 1 2.18 0.31 ac 0.31 0.05 a 2.91 0.22 a 0.33 0.03 a Cg131 Cg135 1 1.46 0.39 ab 0.21 0.04 ab 1.97 0.25 b 0.21 0.03 b Cg136 1 1.61 0.22 b 0.23 0.03 ab Cg138 1 0.71 0.16 c 0.08 0.02 d 1.28 0.27 c 0.16 0.03 bc Cg141 1 1.52 0.17 a 0.20 0.03 abc 2.39 0.24 b 0.29 0.03 a Avo1 3 0.42 0 .07 c 0.07 0.01 d 0.21 0.06 d 0.04 0.01 d Cg156 3 0.18 0.03 c 0.03 0.00 d 0.31 0.06 d 0.04 0.01 d Cg161 3 0.35 0.09 c 0.06 0.02 d Cg128 3 0.48 0.14 c 0.06 0.02 d 0.14 0.03 d 0.02 0.00 d Cg129 2 0.27 0.06 c 0.04 0.01 d 1.10 0.15 c 0.13 0.02 c Cg134 2 0.75 0.21 bc 0.12 0.04 bcd 0.97 0.21 c 0.13 0.03 c Cg157 2 0.54 0.11 c 0.08 0.02 d Control --0.57 0.10 c 0.09 0.02 cd 0.30 0.05 d 0.03 0.01 d LSD 0.73 0.11 0.46 0.06 = <0.0001 <0.0001 <0.0001 <0.0001 a Clade designations follow the CGTT 5+MAT1-2 maximum parsimony phylogeny b ymax is the severity value for leaves 14 days after inoculation c Values in columns with same letter are not significantly different ba sed on Fishers LSD, =0.05. 136

PAGE 137

Table 3-10. Anthracnose area under the diseas e progress curve (AUDPC) and lesion diam eters on detached Tommy Atkinsfruit Exp 1 Exp 2 Isolate Clade a AUDPC diameter (mm)b AUDPC diameter (mm) 1 23.38 4.25 bc 12.38 2.16 ab 19.06 8.16 10.75 4.25 a Cg131 Cg135 1 11.63 2.70 c 7.75 1.45 bc 33.50 9.82 14.88 3.06 a Cg141 1 35.63 5.61 a 17.38 2.33 a 31.81 3.81 16.75 1.65 a Cg129 2 21.44 1.99 bc 10.50 1.04 bc 26.38 3.32 12.13 2.19 a Cg157 2 20.00 4.74 bc 10.13 2.35 bc 33.50 10.87 16.63 4.52 a Cg128 3 14.25 3.70 bc 8.67 3.09 bc 27.75 10.81 15.13 5.21 a Cg161 3 10.81 3.47 cd 5.75 1.48 c 36.13 11.09 18.38 4.51 a Control --0.00 0.00 d 0.00 0.00 d 1.50 0.87 0.25 0.25 b LSD 11.39 5.73 24.14 10.46 = 0.0002 0.0004 0.1178 0.0390 Exp 3 Exp 4 Isolate Clade AUDPC Diameter (mm) AUDPC Diameter (mm) Cg131 1 14.94 5.60 9.00 2.89 bc 22.13 .18 abc 10.25 2.37 ab Cg135 1 16.25 6.65 10.00 3.28 ab 23.56 6.04 ab 11.63 2.92 ab Cg141 1 19.88 5.11 15.00 2.02 a 28.94 1.60 ab 15.75 0.72 a Cg129 2 16.94 2.67 8.38 1.11 bc 27.83 5.65 ab 12.33 2.68 ab Cg157 2 8.69 2.95 4.33 1.09 cd 30.94 3.93 a 13.25 0.92 ab Cg128 3 13.00 1.40 7.13 0.83 bc 29.44 7.40 ab 13. 38 3.79 ab Cg161 3 13.81 2.54 7.13 0.72 bc 14.75 6.61 bc 6.88 3.27 bc Control --3.38 3.38 1.50 1.50 d 0.50 0.35 c 0.50 0.29 c LSD 11.05 5.50 15.09 7.18 aClade designations follow the CGTT 5+MAT1-2 maximuony phylogeny eter measurem14 days after inoculation c Values in columns with same letter are not significantly different ba sed on Fishers LSD, =0.05. m parsimb Diam ents from = 0.1942 0.0067 0.0049 0.0056 137

PAGE 138

138 Table 3-11. Area under the disease progress curve (AUDPC) and ymax values for attached fruit anthracnose experiments, 2009. Haden Tommy Atkins Van Dyke Isolate Clade AUDPC ymax b AUDPC ymax AUDPC ymax Cg131 1 1.22 0.29 bcc 0.38 0.11 bcd 0.61 0.28 0.19 0.06 0.42 0.25 b 0.14 0.05 Cg135 1 1.99 0.49 b 0.49 0.11 bc 1.08 0.23 0.26 0.05 0.28 0.18 b 0.08 0.05 Cg141 1 4.18 1.01 a 0.87 0.02 a 1.16 0.55 0.34 0.13 1.06 0.47 a 0.26 0.13 Cg129 2 0.76 0.20 bc 0.23 0.05 cd 0.51 .24 0.16 0.06 0.24 0.08 b 0.07 0.02 Cg134 2 1.98 0.47 b 0.59 0.12 ab 0.67 0.09 0.24 0.04 ----Cg157 2 --------0.25 0.15 b 0.10 0.05 Cg128 3 1.03 0.40 bc 0.43 0.14 bc 0.48 0.17 0.18 0.06 0.05 0.03 b 0.03 0.01 Cg161 3 0.87 0.11 bc 0.30 0.05 bcd 0.30 0.11 0.13 0.04 0.18 0.05 b 0.06 0.02 Control --0.36 0.29 c 0.10 0.07 d 0.35 0.16 0.10 0.03 0.10 0.04 b 0.05 0.02 LSD 1.42 0.29 0.67 0.16 0.46 0.14 = 0.0013 0.0018 0.0896 0.1118 0.0224 0.0983 a Clade designations follow the CGTT 5+MAT1-2 maximum parsimony phylogeny b ymax is the severity value for fruit 14 days after harvest c Values in columns with same letter are not significantly different based on Fishers LSD, =0.05.

PAGE 139

Cg23 En2 Cg35 Cg52 Cg66 Cg131 Cg135 Cg136 Cg137 Cg138 Cg145 Cg147 Cg148 Cg149 Cg152 Cg158 Cg162 Cg163 Cg169 Cg170 Cg171 Cg165 Cg164 Cg142 Cg141 Cg75 Cg36 Cg34 Avo2 Avo1 Avo3 Cm2 Gua1 Gua2 Piper3 Cg16 Cg101 Cg130 Cg156 Cg161 Cg151 Ban4 Gua3 Pas4 Cg20 Cg55 Cg73 Cg127 Cg128 Cg129 Cg132 Cg133 Cg134 Cg153 Cg154 Cg155 Cg157 Cg159 Cg160 Cg166 Cg167 Cm1 Ban3 Ban1 Ban2 G. acutata 5.7.52 G. magna L2.5 5C. musae 1a 4 0 86 41 4 0 7 5 93 1 CG 2 Outgroup Figure 3-1. Phylogenetic relationships among 65 accessions of based on IT S sequence data. The phylogram represents one of 11 most parsimonious trees (112 steps, CI=0.938, RI=0.949). Support values are bootstrap values over 40%. Isolates are coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fr uit, closed circle=peduncles, closed square=mature fruit). 139

PAGE 140

Pas4 Cg133 Cg73 Cg132 Gua3 Cg128 Cg157 Cg129 Cg127 Cg153 Cg55 Cg134 Cg160 Ban4 Cg155 Cg20 Cg159 Cg166 Cg154 Cg167 Cg151 Cg16 Cm1 Piper3 Cg161 Gua1 Avo3 Cg130 Figure 3-2. Phylogenetic relationships among accessions of using ML analysis of ITS sequence da ta, ln score = -1071.9. Support values are bootstrap values over 50%. Isolates are coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit). Ban1 Ban2 Ban3 Cg156 Gua2 Avo1 Cg101 Avo2 Cm2 Cg148 Cg152 Cg169 Cg170 Cg135 Cg136 Cg145 Cg158 En2 Cg23 Cg137 Cg52 Cg162 Cg131 Cg163 Cg66 Cg35 Cg138 Cg147 Cg149 Cg171 Cg36 Cg34 Cg75 Cg141 Cg164 Cg142 Cg165 G. magna L2.5 G. acutata 5.7.52 C. musae 98 1a 61 97 Outgroup 0.03 140

PAGE 141

Cg34 Cg35 Cg36 Cg52 Cg60 Cg66 Cg75 Cg131 Cg135 Cg136 Cg137 Cg138 Cg142 Cg145 Cg148 Cg147 Cg152 Cg158 Cg162 Cg163 Cg164 Cg165 Cg169 Cg170 En2 Figure 3-3. Phylogenetic relatio nships among 58 accessions of based on MAT1-2 sequence data. The phyl ogram represents one of 14 most parsimonious trees (216 steps, CI=0.966, RI=0.981). Support values are bootstrap values over 50%. Isolates are coded for hos t tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit). Ban1 Ba n2 Gua3 Cg167 Cg160 Cg159 Cg153 Cg127 Cg129 Pap17 Pap16 Gua2 Gua1 Cm2 Cg161 Cg157 Cg156 Cg155 Cg154 Cg151 Cg134 Cg130 Cg128 Cg126 Cg101 Cg73 Cg55 Cg20 Ban4 Avo1 Avo2 Cm1 Cm3 G. acutata 5.7.52 G. magna L2.5 Outgroup 1 C. musae 72 92 97 CG 100 10 141

PAGE 142

Cg137 En2 Cg169 Cg164 Cg131 Cg145 Cg148 Cg135 Cg170 Cg36 Cg142 Cg165 Cg35 Cg163 Cg152 Cg34 Cg60 Cg138 Cg147 Cg75 Cg52 Cg162 Cg149 Cg66 Cg158 Cg136 Ban1 Figure 3-4. Phylogenetic relationships among accessions of using ML analysis of MAT1-2 sequence data, ln score = -821.25. Support values are bootstrap values over 50%. Isolates are coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit). Ban2 Gua3 Cg127 Cg159 Cg129 Cg167 Cg153 Cg160 Pap17 Avo1 Cg161 Cg55 Cm2 Cg130 Cg154 Cg73 Cg155 Cg128 Gua1 Cg151 Cg134 Avo2 Cg126 Cg157 Cg20 Ban4 Pa p16 Cg101 Gua2 Cg156 Cm1 Cm3 G. acutata 5.7.52 G. magna L2.5 1 C. musae 97 90 Outgroup 0.05 142

PAGE 143

Figure 3-5. Phylogenetic relatio nships among 58 accessions of based on CGTT5 sequence data. The phylogr am represents one of two most parsimonious trees (73 steps, CI=0.959, RI=0.995). Support values are bootstrap values over 50%. Isolates are coded for hos t tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit).57 86 100 98 63 91 97 En2 Cg35 Cg36 Cg52 Cg75 Cg138 Cg142 Cg148 Cg149 Cg158 Cg169 Cg170 Cg66 Cg131 Cg136 Cg165 Cg164 Cg163 Cg152 Cg147 Cg145 Cg141 Cg137 Cg34 Cg135 Gua3 Cg72 Cg20 Cg129 Cg157 Cg154 Cg153 Cg134 Cg55 Cg73 Ba n3 Ba n1 Ban2 Cm1 Cm3 Cg127 Gua1 Cg130 Cg128 Cg101 Cg167 Cg161 Cg160 Cg159 Cg155 Cg156 Cg151 Gua2 Cm2 Ban4 Avo1 Avo2 1 1a 100 2 C. musae 100carambola 3 2 143

PAGE 144

69 87 100 100 99 64 85 Cg34 Cg141 Cg164 Cg152 Cg145 Cg165 Cg147 Cg135 Cg137 Cg163 En2 Cg142 Cg148 Cg35 Cg158 Cg131 Cg138 Cg136 Cg149 Cg170 Cg52 Cg169 Cg36 Cg66 Cg75 Gua3 Cg72 Cg55 Cg20 Cg129 Cg73 Cg134 Cg153 Cg154 Cg157 Ban1 Ban2 Ban3 Cg127 Avo1 Cg101 Avo2 Cg128 Cg160 Cg161 Cg156 Cg155 Ban4 Gua1 Gua2 Cg130 Cg167 Cm2 Cg151 Cg159 Cm3 Cm1 0.0053 1 carambola C. musae 2 1a Figure 3-6. Phylogenetic relationships among accessions of using ML analysis of CGTT5 sequence data ln score = -1166.49. Support values are bootstrap values over 50%. Isolates are coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit). 144

PAGE 145

En2 Cg35 Cg36 Cg52 Cg66 Cg75 Cg131 Cg136 Cg138 Cg142 Cg148 Cg149 Cg158 Cg169 Cg170 Cg165 Cg164 Cg163 Cg152 Cg147 Cg145 Cg137 Cg34 Cg135 Gua3 Figure 3-7. Phylogenetic relatio nships among 54 accessions of based on combined MAT1-2+CGTT5 sequenc e data set. The phylogram represents one of 63 most parsimonious trees ( 224 steps, CI=0.951, RI=0.986). Support values are bootstrap values over 50%. Isolates ar e coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit). Ban1 Ba n2 Cg129 Cg153 Cg157 Cg154 Cg134 Cg73 Cg20 Cg55 Cg127 Cg159 Cg160 Cg167 Cg161 Cg156 Cg155 Cg151 Cg130 Cg128 Cg101 Gu a2 Gua1 Cm2 Ban4 Avo1 Avo2 Cm 1 Cm3 G. magna L2.5 G. acutata 5.7.52 99 1 78 1a 87 78 52 C. musae 100 76 2 77 54 3 69 60 carambola Outgroup 10 145

PAGE 146

Figure 3-8. Phylogenetic relationships among accessions of using ML analysis of MAT1-2+CGTT5 da ta set, ln score = -2075.22. Support values are bootstrap values over 50%. Isolates ar e coded for host tissue as follows: open triangle=leaves, open circle=panicles, open square=immature fruit, closed circle=peduncles, closed square=mature fruit).99 Cg36 Cg66 Cg136 Cg148 Cg75 Cg158 Cg149 Cg169 Cg142 Cg52 Cg138 Cg170 En2 Cg131 Cg35 Cg165 Cg145 Cg152 Cg137 Cg163 Cg147 Cg164 Cg34 Cg135 Gua3 Ban1 Ban2 Cm1 Cm3 Cg127 Cg159 Cg167 Cg160 Cg101 Cg155 Cg156 Cm2 Gua2 Cg151 Avo2 Gua1 Avo1 Cg128 Ban4 Cg130 Cg161 Cg129 Cg153 Cg154 Cg134 Cg73 Cg20 Cg55 Cg157 G. magna L2.5 G. acutata 5.7.52 97 152 1a 63 66 68 69 85 C. musae carambola 3 Outgroup 0. 01 146

PAGE 147

147 C g 131 C g 135 C g 136 C g 138 C g 141 C g 164 C g 129 C g 134 C g 157 C g 128 Avo1 C g 156 C g 161 Figure 3-9. Micrographs of conidi a of indicated isolates of Bars=10 m.

PAGE 148

148 Cg131 Cg135 Cg136 Cg138 Cg141 Cg164 Cg129 Cg134 Cg157 Cg128 Avo1 Cg156 Cg161 Figure 3-10. Micrographs of hyphopodia of indicated isolates of Bars=10m.

PAGE 149

A B C E G F F E D C B A Figure 3-11. Blossom blight 28 days after i noculation. Severe symp tom s were induced by isolates in Clade 1: A) Cg131, B) Cg164, C) Cg135 and D) Cg138. In contrast, minor or no symptoms were caused by isolates outside Clade 1: E) Cg129 and F) Cg161. 149

PAGE 150

D C B A Figure 3-12. Leaf anthracnose 10 days after inoc ulation. Symptoms caused by isolates in Clade 1, A) Cg131 and B) Cg141, showing nume rous lesions as well as large areas of necrosis and leaf distortion. Leaves inoc ulated with non-Clade 1 isolates, C) Cg128 and D) Avo1, showing few small lesions. 150

PAGE 151

151 A Cg 141 Control Cg 128 A D B C Figure 3-13. Lesion developm ent on artificially inoc ulated fruit: A) 7 days after inoculation of a detached fruit, with isolates Cg128 and Cg141 compared to the water control; and attached fruit 14 days after harvest, B) wa ter control, C) isolate Cg141 and D) isolate Cg161.

PAGE 152

152 # polymorphisms 010203040506070 C. musae (4) C. boninense C. capsici C. acutatum 25 36 59 C. gloeosporioides sensu lato C. fragariae (3 ) C. kahawae, Mango clade 1 (2) Figure 3-14. Graphical representation of the num ber of polymorphisms (i ncluding nucleotide and indel sites) among four taxa within sensu lato and three distantly related species, and Numbers are shown in parentheses or in white text w ithin bubbles. Bubble size is proportional to the numbers of polymorphisms.

PAGE 153

CHAPTER 4 INVESTIGATING TH E ROLE PLAYED BY IN THE DEVELOPMENT OF BLOSSOM BLIGHT AND LEAF AND FRUIT ANTHRACNOSE OF MANGO IN SOUTHERN FLORIDA Anthracnose is the most impoortant disease of mango in wet climates. Three organs of the host are affected, resulting in blossom blight leaf anthracnose, and fruit anthracnose. var. and have been associated with these diseases, but is most prevalent. has only been reported on mango in Aust ralia (Fitzell 1979), Taiwan (Weng and Chuang 1995) and Homestead, Florida (Riveras-Var ga et al. 2006). In so uth Florida, RiverasVargas et al. (2006) reported that comprised only 13% of th e isolates that they recovered from mango, the rest of which were identified as Although they have been confused due to their similar morphol ogies, molecular markers are now available that identify sensu lato and sensu lato (Mills et al. 1992, Sreenivasaprasad et al. 1996b). and are distinct but diverse taxa that are now recognized as species complexes (Sutt on 1980, Sreenivasaprasad and Talhinhas 2005). Sreenivasaprasad and Talhinhas (2005) reported eight distinct lineages in sensu lato based on phylogenetic analysis of the ITS1-5.8S-ITS2 region of 109 accessions. Several host specific lineages were iden tified, including A8 from and A1 from ( ). The type specimen from resolved in group A5. The species and resolve within sensu lato (Nirenberg et al. 2002, Farr et al. 2006). 153

PAGE 154

In 2007, isolations were made from s ymptomatic leaves, blossoms and fruit from a cv.Keitt mango grove at the Tropical Resear ch and Education Center (Homestead, FL). Although was recovered from all blossom a nd fruit lesions, most of the leaf lesions yielded (Tarnowski and Ploetz 2008). Given its prevalence on Keitt leaves and the infrequent reports of it as a mango pat hogen, the present study sought to define and contrast the roles played by and on mango in southern Florida. Although the work focused primarily on leaf anthr acnose, blossoms and fruit were also studied as host organs. Between July 2007 to May 2008, nine mango groves were surveyed, six of which were commercial plantings in the Redlands agricultura l area, and three of which were experimental (University of Florida, Tropi cal Research and Education Center (TREC), Homestead, FL; USDA-ARS station, Miami, FL). Three of the commercial properties were exclusively of cv. Keitt, one was exclusively of cv. Tommy Atkins, and two contained blocks/rows of several cultivars, from which trees of Tommy Atkins were sampled. Two plantings were studied at the TREC location, a solid block of cv. Keitt and a mixed germplasm collection. The USDA location was a mixed germplasm collection. Lesions were sampled hierarchically and only from new, bright green and supple le aves. From past observation, is most often recovered from new, actively growing lesi ons on young leaves, and becomes increasingly difficult to isolate in old lesions (data not shown). During th e survey, the morphology of all lesions from which was recovered was documented with digital photography. At each location, three to 17 trees were sampled depending on the abundance of leaf anthracnose symptoms. One to three symptomatic leaves were chosen from each tree, and one to three lesions 154

PAGE 155

were chosen from each leaf. Lesions were cut fro m the leaf, surface disinfested (submersion for 60s in 10% bleach and rinsed with sterile water), and incubated in a moist chamber for 3 to 4 days. Conidia were then lifted from the acervuli that developed with a ster ile needle and plated on PDA. The resulting colonies were identified as or based on colony and conidium morphology (Figure 4-1). Colonies of colonies were fast growing (6.72 0.11 mmday-1), white to grey, fluffy with orange sporulation and produced straight and cylindrical conidia. Colonies of were slow growing (3.33 0.31 mmday-1), grey to pink with pink reverse, and powdery, with bright salmon sporulation, and produced fusiform conidia. Isolates were single-s pored and stored for future use in 10% glycerol at -80C. The above phenotypic identifica tions were corroborated with molecular diagnoses using and specific primers. Isolates were grown for 3-5 days in PDB at room temperature (23-25C) on a shaker at ca. 80 rpm. Approximately 150 mg mycelia were harvested and triple rinsed with sterile deioni zed water and dried on sterile filter paper. DNA was extracted using a DNA genomic preparation protocol from the University of Wisconsin Biotechnology Center. Briefly, mycelia were gr ound in 500 l Shorty DNA Extraction Buffer (0.2M Tris-HCl, pH 9.0, 0.4M LiCl, 25mM EDTA, 1% SDS) and incubated at 68C for 10 minutes. Tissue was centrifuged at 14,000 rpm fo r 5 minutes, 400 l of the supernatant was transferred to new tube and the DNA was precipi tated with 400 l 99% isopropanol. The tubes were centrifuged for 10 minutes at 14,000 rpm and the supernatant was decanted. The DNA pellets were air-dried for 5 mi nutes then resuspended in 400 l TE buffer (10 mM Tris-Cl, pH 7.5, 1 mM EDTA) for 30 minutes at room temperature. 2l DNA was used in PCR reactions. 155

PAGE 156

The ITS1-5.8S-ITS2 ribosomal region was am plified via PCR in 50 l reactions that contained 38.25 l of sterile di stilled, deionized water, 6.5 l ThermoPol Reaction Buffer (New England Biolabs, Ipswich, MA), 1 l 10 mM dN TP mix (New England Biolabs, Ipswich, MA), 0.25 l Taq DNA polymerase (conc. 5,000 units/ml), 1 l each of 15 M primers and 2 l DNA template using the following primer s: CaInt (Sreenivasaprasad et al. 1996b) and CgInt (Mills et al. 1992) for and respectively, and ITS4 (White et al. 1990). Standard cycling parameters used a 55 C annealing temperature. The identity of five isolates of sensu lato from this study was confirmed by comparing their ITS sequences with those from previous studies. The entire ITS1-5.8S-ITS2 region of Ca3, Ca13, Ca26, Ca33, and Ca60 was amplified with primers ITS1 (Gardes and Bruns 1993) and ITS4, using a 55C annealing temperature, and amplified products were sequenced by the Interdisciplinary Center for Biotechnology Research, Gainesville, FL. Sequences were compared with those of isolates of and in Genbank from studies by Freeman et al. (2001), Nirenberg et al (2002), Moriwaki et al (2002), Saha et al. (2002), Vinnere et al. (2002), Af anador-Kafuri et al. (2003), Ma rtinez-Culebras et al. (2003), Lubbe et al. (2004), Talhinhas et al. (2005), Than et al. (2006), Farr et al. (2006), and MacKenzie et al. (2009) (Table 4-1). Th e sequences represented differe nt genetic groups in the species complex, as described in Sreenivasa prasad and Talhinas (2005) and including and (Farr et al. 2006), and different hosts of in Florida. Sequences were aligned with CLUSTALX (Thompson et al. 1994) and adjusted manually in Mega 4.0 using default parameters. Maximum parsimony analysis was performed in Mega 4.0 using closest neighbor interchange (CNI) searches with 100 random taxon additions and gaps were treated as missing 156

PAGE 157

data. Bootstrap analysis according to Felsenst ein (1985) was performed using the above search criteria, with 500 repetitions. Maximum like lihood analysis was performed using GARLI v.0.946, with default parameters and a randomly generated starting topology. Analysis was stopped after 5x106 generations or a 0.01 decrease in ML score. Three independent ML runs were performed and the resulting topologies were compared in TreeView v1.6.6. If similar topologies were generated, bootstra p analysis was performed using the same parameters with 500 repetitions. The isolate Cg131 ( Chapter 3), isolate L2.5, and M1.001 were used as outgroup taxa. Inoculum was prepared by growing isolates retrieved from -80C storage on PDA for 7-10 days (Table 4-2). Conidia were harvested by flooding plates with st erile deionized water, scraping colonies with a sterile glass, and straining the suspen sion through a layer of sterile cheesecloth. Inoculum suspensions were adjusted to a concentration of 106 conidiaml-1. Attached panicles were inoculated in the field in January and February, 2008 and 2009. Seven treatments (three isolates of three isolates of and a water control) were applied to runoff with a handheld manual spray bottle to inflorescences at the mouse-ear to green-colored stage (florets extend ed but no pigment development, as described by Schoeman et al. 1995); conidium concentrations were 106 conidiaml-1 sterile deionized water. Inflorescences were then covere d for 48 hrs with a plastic bag in serted in a Kraft brown paper bag. A randomized complete block design (RCBD) was used, where experimental units were single inflorescences, treatments were replicated five times, and individu al trees were blocks. The experiment was conducted three times in the same grove of cv. Keitt at TREC. 157

PAGE 158

Disease severity was rated six times for four weeks after inoculati on, using a synoptic key (key 1.5; Jam es, 1971). At the end of the experi ment, lesion margins were excised from leaves, surface disinfested, and plated on PDA to confirm presence of the pathogen. Area under the disease progress curve (AUDPC) was calcula ted for each experimental unit, and ymax was disease severity at 28 days after inoculati on. Analyses of variance of AUDPC and ymax were performed for each experiment, and mean separations were performed using least significant differences (PROC GLM, SAS). During the summers of 2008 a nd 2009, three attached leaf ex periments were conducted in the same grove of cv. Keitt that was used in the above blossom bli ght experiments. Seven treatments (three isolates of three isolates of and a water control) were applied to runoff with a handheld manual spray bottle to newly opened vegetative shoots. Conidium concentrations were 106 conidiaml-1 sterile deionized water, and treated shoots were covered with Kraft paper bags for 48 hours. Treatments were replicated four times in a RCBD and trees were blocks. Experimental un its were single shoots, and the five youngest leaves in a shoot were subsam ples in disease analyses. Disease severity was measured after 4, 7, 10, and 14 days using key 2.1.2 of James (1971). The experiments were stopped after 14 days due to severe defoliation on some shoots. Lesion margins were excised from leaves, surface disinfested, and plated on PDA to confirm presence of the pathogen. Analyses of variance of AUDPC and ymax were performed using PROC GLM in SAS. Mean separations were performed using l east significant differences (PROC GLM, SAS). Mature fruit of cvs. Florigon Haden, Saigon and Turpen tine were harvested from a cultivar collection at TREC in July 2008. On each cv, five treatments (two isolates of 158

PAGE 159

, two of and a control) were im posed in a completely randomized design (CRD). Fifteen l droplets of 0.3% water agar, without (c ontrol) or with 106 conidiaml-1, were applied to sites on the fru it surface that had been wounded with a sterile needle. Each fruit was treated in two or three locations, depending on fr uit size, and treatments were replicated five times. Fruit were incubated in plastic sweater boxes on wire mesh over moistened paper towels at 25C in the dark, and disease was assessed af ter 4, 5 and 6 days. Lesion diameters were measured twice at right angles, and mean diameters were used in statistical analyses. After 6 days, lesion margins were excised from fruit, surface disinfested, and plated on PDA to confirm presence of the pathogen. The experiment was conducted three times. Anal yses of variance of AUDPC and ymax were performed with PROC GLM in SA S. Mean separations were performed using least significant diffe rences (PROC GLM, SAS). spp. were not recovered from high proportions of lesions in the grove survey (4-22% recovery across locations), due to their age or other coloni zing fungi; only lesions from which was recovered are reported (Table 4-3). was only recovered from leaves of Keitt (thr ee locations) and from an unknown cultivar in a mixed orchard; it was prevalen t (>70% of all isolates of spp. that were recovered) in only two locations. Fewer trees were sampled in Tommy Atkins groves (3-9 trees) vs Keitt groves (12-17 trees), due to lower inci dences of leaf anthracnose, and only was recovered from leaf lesions on Tommy Atkins. The leaf symptoms associated with and caused by began as small pinpricks that often coalesced to cover large pa rts of the lamina (Figure 4-2A). In contrast, 159

PAGE 160

caused fewer, angular lesions that grew to approximately 5-7 mm in dia meter (Figure 4-2B and data not shown). Although they were larger than indivi dual lesions caused by lesions caused by usually did not coalesce. Leaves affected by became puckered as healthy l eaf tissue around lesions continued to grow. Necrotic areas in lesions eventually se nesced and fell from the leaf surface resulting in a shot-hole appearance. In phylogenetic analyses with isolates representing the known diversity of as well as the closely related species and mango isolates from south Florida fell in Sreenivasaprasad and Talhinhass ( 2005) group A2 with 53% bootstrap support (MP analysis) (Figure 4-3). MP analysis of the 44 accessions resulted in 5,412 most parsimonious trees (length 65, CI=0.778, RI=0.966). ML analysis resulted in a tree (ln score=-1199.699) of similar topology to the MP majority consensus tree, except that some of the clades had lower bootstrap support (Figure 4-3). The same clades, (A4, A6, A7), (A1), A3, and A8, had bootstrap support >65% for both analyses. C. acutatum C. gloeosporioides There was a significant isolat e effect for AUDPC and ymax ( <0.05) for each of three experiments (Table 4-4); in each, all isolates of caused significant blossom blight, compared with the water controls, whereas isolates of did so inconsistently. With few exceptions, isolates of caused significantly higher AUDPC and ymax values than those of Although isolates of and caused comparable levels of anthracnose on artificially i noculated leaves, disease deve lopment was inconsistent among 160

PAGE 161

isolates and the three experiments (Table 4-5). Overall tre atment effects were insignificant in experiment 1 ( =0.0640 and 0.0565 for AUDPC and ymax, respectively), and only one of the isolates, Ca2, caused AUDPC and ymax values significantly higher than the water control in all three experiments. Both species caused symptoms similar to those observed on naturally affected leaves in the field (Figure 4-2) and, in general, caused more severe disease than the water control. All isolates produced significant lesions on wounde d fruit of each of the four cvs that were tested (Table 4-6). On each cultivar, generally produced smaller lesions and lower AUDPC values than but these differences were often not significant. Fitzell (1979) first reported as a cause of mango anthracnose in New South Wales. He isolated from leaves and fruit, and re produced anthracnose symptoms on leaves, panicles and fruit. has also been reported on mango in Taiwan (Weng and Chang 1995) and Homestead, FL (Riveras-Vargas et al. 2006). Riveras-Vargas et al. (2006) isolated only from mango flowers, peduncles and immature fruit. In contrast, in the present study was found most often on leaves, once on a panicle, and never on fr uit (data not shown). Why such different results were obtained in generally the same production area is not clea r. The previous study in cluded a single location from Homestead, and the cultivars sa mpled were not reported. It is possible that the leaf isolates obtained from this location were not sampled from lesions caused by Additionally, a population of other than group A2, as defined by Sreenivasaprasad and Talhinhas (2005), could have been involved in the previ ous study, different groups of cultivars may have been examined, or production practices or other conditions may have differed. 161

PAGE 162

Regardless of the reason(s) for the se differenc es, there was substantial agreement between results from the field survey and the artifici al inoculations in the present study. Although complete data is available for only for Keitt, it appears that may be an important pathogen only on mango leaves, on which it causes symptoms that are distinct from those that are caused by As reported in previous st udies (Fitzell 1979, Freeman and Shabi 1996, Peres et al. 2002), also caused lesions on wounded fruit. However, during field surveys was recovered only once from na turally blighted inflorescences and it was never recovered from fruit lesions, ev en in locations where it was most prevalent on leaves. If plays a significant role in the development of blossom blight and fruit anthracnose on mango, it occurs on different cultivars or under different environmental conditions than were examined in the present study. causes blights and anthracnose diseases on leaves, fruit and inflorescences of other hosts. It is not a genera l necrotroph, and different hostand tissue-specific populations exist (Peres et al. 2005). Fo r example, on citrus one lineage of causes key lime anthracnose and a ffects all young tissues of and another causes post-bloom fruit drop on many citrus species a nd affects only young devel oping fruit (Agostini et al. 1992, Timmer et al. 1994, Peres et al. 2008). Therefore, it is not surprising that the mango population would show some tissue specificity. sensu lato is a species complex (Sreenivasaprasad and Talhinhas 2005). Although phylospecies within the taxon have not been thor oughly characterized, and have been described and resolved within sensu lato (Nirenberg et al. 2002, Farr et al. 2006). Mango isolates from the present study had identical ITS sequences and fell in Sreenivasaprasad and Talhinhass (2005) A2 group. A2 is diverse and contains isolates 162

PAGE 163

from , and It is common in North American and Europe, and found in Florida on strawberry. Additional work is needed on the phylogeny and biology of the mango population of that was identified in this study and to determine its relationship to those in other locations. Results from the leaf anthracnose survey suggest that the and are not uniformly distributed in south Florida. In each of the examined groves, one of the species prevailed (Table 4-3). was recovered in only four four locations, and was preval ent in two. In addition, it was never recovered from Tommy Atkins leaves. Anthocyanins, the end-products of the fl avonoid biosynthesis pathway, have been associated with disease resistance in other pathosystems (Hammerschmidt and Nicholson 1977, Kraft 1977, Wegulo et al. 1998). In mangos, anthocyanin content varies among different cultivars (based on color of emerging shoots) and as vegetative shoots develop, and increase as leaves emerge but decline as th ey expand (Ali et al. 1999). Whethe r the differential responses of Tommy Atkins and Keitt leaves to that were observed in this study are associated with anthocyanins should be investig ated. If a relationship between anthocyanins and mango anthracnose exists, it would probably be complex. For example, the wide range in anthracnose responses among red fruited mango cu ltivars (e.g. fruit of Tommy Atkins are tolerant but those of Sensation are susceptible) suggests that different flavonoids would either be responsible for these interac tions on fruit or that flavonoids are involved in anthracnose responses on some, but not all organs. By comparing the geographic distribution of the two species reported in the USDA-ARS Fungal Database ( http://nt.ars-grin.gov/fungaldatabases/index.cfm ), it is apparent that 163

PAGE 164

164 is most important in temperate regions and is more common in the tropics. Thus, the subtropics may be a transition zone for the two species, and may be the only location where would be expected to be found on a tropical crop such as mango. To date, has only been found on mango in the subtropics (Fitzell 1979, Weng and Chang 1995, Riveras-Vargas et al. 2006). and cause anthracnose on peach, apple, pecan, grape, almond, strawberry and other hosts, often in the same location (Kummuang et al. 1984, Smith and Black 1990, Bernstein et al. 1995, Shi et al. 1996, Freeman et al. 1998, Gonzalez and Sutton 2004, Hong et al. 2008). On some of these hosts, such as strawberry and apple (Smith and Black 1990, Urea-Padilla et al. 2002, Gonzalez and Sutton 2004), the two species cause distinct symptoms similar to the situation on Keitt leaves in the present study. This study sheds new light on the role played by on mango. affects newly emerged leaves of Keitt, but additional work is needed to fully understand its impact on other cultivars a nd organs of this important fruit crop.

PAGE 165

Table 4-1. ITS sequences includ ed in phylogenetic analysis Species Host Origin Genbank accession # Group designationa Publishedb Accession ATCC 56816c Australia DQ286132 --Farr et al. 2006 sp. Mexico DQ286130 --Farr et al. 2006 MEP1534 USA DQ286121 --Farr et al. 2006 ATCC MYA-662 New Zealand DQ286124 --Farr et al. 2006 MEP1323 USA DQ286119 --Farr et al. 2006 AR2826 USA DQ286117 --Farr et al. 2006 AR2820 South Africa DQ286138 --Farr et al. 2006 AR3410 England DQ286144 --Farr et al. 2006 CBS198.35 South Africa DQ286146 --Farr et al. 2006 AR3787 England DQ286142 --Farr et al. 2006 CBS199.35 New Zealand DQ286140 --Farr et al. 2006 MEP1334 England DQ286142 --Farr et al. 2006 CBS199.35 d Germany AJ301931 A4 Nirenberg et al. 2002 BBA 67435 --AJ301924 A5 Nirenberg et al. 2002 BBA 62124 Canada AJ301959 A1 Nirenberg et al. 2002 BBA 71249 Germany AJ301933 A1 Nirenberg et al. 2002 BBA 70358 --AJ301910 A3 Nirenberg et al. 2002 BBA 70338 South Africa AY376507 A2 Lubbe et al. 2004 STE-U 4459 South Africa AY376498 A5 Lubbe et al. 2004 STE-U 164 USA AY376509 A3 Lubbe et al. 2004 STE-U 5287 Spain AJ536209 A4 Martinez-Culebras et al. 2003 IMI 345026 New Zealand AJ536213 A7 Martinez-Culebras et al. 2003 IMI 345585 France AJ536199 A2 Martinez-Culebras et al. 2003 IMI 345027 Japan AJ536205 A2 Martinez-Culebras et al. 2003 IMI 3600866 Florida EU647311 --MacKenzie et al. 2009 05-200 Florida EU647307 --MacKenzie et al. 2009 Ss Florida EU647306 --MacKenzie et al. 2009 STF-FTP-10 Florida EU647305 --MacKenzie et al. 2009 OCO-ARC-4 Florida EU647302 --MacKenzie et al. 2009 02-163 Netherlands AJ749675 A2 Talhinas et al. 2004, unpublished PD85-694 165

PAGE 166

166 Table 4-1. Continued. Isolate name Species Host Origin Genbank accession # Group designation Publisheda CBS193.32 Italy AJ749688 A4 Talhinhas et al. 2005 PT250 Portugal AJ749700 A6 Talhinhas et al. 2005 S8 Sweden AF411731 A4 Vinnere et al. 2002 IMI 117619e Australia AF411701 A5 Vinnere et al. 2002 IMI 223120 Australia AF272783 A2 Freeman et al. 2001 PCN5 USA AF272786 A3 Freeman et al. 2001 TOM-21 Colombia AF521196 A8 Afanador-Kafuri et al. 2003 TOM-9 Colombia AF521205 A8 Afanador-Kafuri et al. 2003 S2 Thailand DQ454018 --Than et al. 2006 IMI 383015 India AF488778 A2 Saha et al. 2002 Cooley2 --AF489558 A3 Denoyes-Rothan et al. 2003 MAFF 306282 Japan AB042300 A3 Moriwaki et al. 2002 Ca3 Florida GU045506 ----Ca13 Florida GU045507 ----Ca26 Florida GU045508 ----Ca33 Florida GU045509 ----Ca60 Florida GU045510 ----Cg131f Florida GQ373223 ----L2.5 USA DQ003103 --Du et al. 2005 M1.001 Missouri DQ003110 --Du et al. 2005 a Group designation based on ITS analysis publi shed in Sreenivasaprasad and Talhinhas 2005 b Publication in which ITS sequence was published. c Type speciment for d Species diagnosis in Genbank is questionable e Paratype specimen for f Outgroup taxon, recovered from in Florida (see Chapter 3)

PAGE 167

Table 4-2. Isolate, species and ma ngo origin for isolates that we re used in pathogenicity tests Isolate Species Tissue origin Organ on which isolate tested for pathogenicity Leaf Blossom, Leaf, Fruit Ca1 Leaf Blossom, Leaf, Fruit Ca2 Ca3 Leaf Blossom, Leaf Cg135 Immature fruit Blossom, Leaf, Fruit Cg136 Leaf Blossom, Leaf Cg131 Mature Fruit Blossom, Leaf, Fruit 167

PAGE 168

Table 4-3. Results for leaf anthracnose surveys Locationa Cultivar(s) # trees sampled # lesions sampledb c Grove A Keitt 12 24 22 (92) 2 (8) Grove B Keitt 12 24 7 (29) 17 (70) Grove C Mixed 3 1 1 (100) 0 (0) Grove D Tommy Atkins 3 5 5 (100) 0 (0) Grove E Tommy Atkins 9 11 11 (100) 0 (0) Grove F Tommy Atkins 6 7 7 (100) 0 (0) TREC A Keitt 17 48 8 (17) 40 (83) TREC B Mixed 9 9 8 (89) 1 (11) USDA Mixed 17 18 18 (100) 0 (0) a Groves A-F are commercial groves, TREC groves are located at the University Floridas Tropical Research a nd Education Center in Hometead, and the USDA grove is located at the USDA station in Miami. b The number of lesions from which was recovered. c The numbers (percentages) of lesi ons that yielded a given species. 168

PAGE 169

Table 4-4. Areas under the diseas e progress curve (AUDPC) and ymax values for blossom blight experiments, 2008-2009 Exp1 Exp2 Exp3 AUDPC ymax a AUDPC ymax AUDPC ymax Isolate 6.92 2.24 abb 0.37 0.09 ab 3.36 0.86 b 0.18 0.04 b 3.00 0.84 bc 0.20 0.07 cd Ca1 Ca2 3.62 1.21 bc 0.20 0.04 bc 3.69 0.42 b 0.24 0.04 b 3.14 1.96 bc 0.21 0.15 cd Ca3 2.86 1.11 c 0.20 0.07 bc 2.53 0.68 b 0.15 0.03 b 5.76 1.69 b 0.40 0.10 bc Cg131 --------11.38 2.31 a 0.74 0.11 a Cg136 8.92 1.89 a 0.56 0.12 a 7.66 1.72 a 0.49 0.09 a 3.39 0.29 bc 0.18 0.03cd Cg135 7.75 2.22 a 0.51 0.16a 8.43 0.39 a 0.64 a 12.33 1.53 a 0.69 0.10a Control 1.05 0.28c 0.10 0.02c 1.94 0.52 b 0.15 0.04 b 0.98 0.27 c 0.06 0.01 d LSD 3.85 0.22 2.52 0.18 4.65 0.32 -value 0.0001 <0.0001 0.0047 0.0032 <0.0001 0.0006 a ymax is the severity value for pani cles 28 days after inoculation b Values in columns with same letter are not significantly different based on Fishers LSD, =0.05. 169

PAGE 170

Table 4-5. Areas under the disease progress curves (AUDPC) and ymax values for leaf anthracnose experiments, 2008-2009 Exp1 Exp2 Exp3 AUDPC ymax a AUDPC ymax AUDPC ymax Isolate 0.73 0.20 abb 0.10 0.02 b 0.67 0.17 a 0.11 0.04 ab 0.87 0.20 bc 0.13 0.03 cd Ca1 Ca2 1.59 0.67 a 0.34 0.14 a 0.51 0.16 abc 0.08 0.03 abc 1.22 0.18 b 0.21 0.03 bc Ca3 1.09 0.41 a 0.15 0.06 ab 0.59 0.15 ab 0.12 0.04 a ----Cg131 0.92 0.13 ab 0.13 0.02 b 0.30 0.10 bcd 0.04 0.01 cd 2.10 0.32 a 0.32 0.05 a Cg136 ----0.25 0.04cd 0.04 0.01 cd 1.48 0.21 ab 0.23 0.03 ab Cg135 1.47 0.45 a 0.19 0.05 ab 0.37 0.07 abc 0.07 0.02 bcd1.34 0.24 b 0.20 0.04 bc Control 0.05 0.02 b 0.01 0.00 b 0.03 0.02 d 0.01 0.00 d 0.53 0.09 c 0.10 0.02 d LSD 1.01 0.21 0.31 0.06 0.62 0.10 0.0640 0.0565 0.0050 0.0039 <0.0001 0.0002 value a ymax is the severity value for pani cles 28 days after inoculation b Values in columns with same letter are not significantly different based on Fishers LSD, P=0.05. 170

PAGE 171

171 Table 4-6. Areas under the diseas e progress curve (AUDPC) and ymax (lesion diameter) values for detached fruit experiments, 20082009 Florigon Haden Isolate AUDPC Lesion diameter (mm)a AUDPC Lesion diameter (mm) Ca1 11.67 1.48 abb 9.67 1.12 a 11.88 4.72 ab 8.25 2.56 a Ca2 10.5 1.88 b 8.20 1.52 a 10.75 3.38 ab 9.25 2.82 a Cg135 15.90 1.97 a 12.80 1.08 a 13.50 1.77 a 10.00 0.54 a Cg131 15.00 .77 ab 11.20 2.20 a 19.69 4.25 a 11.88 2.37 a Control 4.04 0.04 c 2.08 0.08 b 2.13 0.13 b 1.00 0.00 b LSD 5.19 6.09 10.00 6.09 -value 0.0006 <0.0001 0.0293 0.0167 Saigon Turpentine AUDPC Lesion diameter (mm) AUDPC Lesion diameter (mm) Ca1 19.50 2.01 ab 13.75 0.83 ab 14.63 3.61 b 12.00 3.03 a Ca2 15.19 2.97 b 11.88 1.23 b 14.19 3.17 b 14.13 2.22 a Cg135 17.63 1.13 b 13.75 0.72 ab 25.63 4.88 a 17.75 3.25 a Cg131 24.17 1.74 a 16.33 0.89 a 25.19 2.25 a 17.63 1.12 a Control 2.75 0.43 c 1.50 0.29 c 2.00 0.00 c 1.00 0.00 b LSD 3.75 3.75 3.33 3.33 -value <0.0001 <0.0001 0.0004 0.0004 a Lesion diameters 14 days after inoculation b Values in columns with same letter are not significantly differe nt based on Fishers LSD, P=0.05.

PAGE 172

B A Figure 4-1. Conidia of A) were slightly fusiform, whereas those of B) had cy lindrical tips. Bars=10m 172

PAGE 173

B A D C F E Figure 4-2. Anthracnose lesions co llected during the grove survey (A-D) and lesions that were produced during pathogenicity tests (E-F). Lesions in A) and B) were associated with and displayed an angular morphology and shot-holes. Lesions in C) and D) were associated with and were initially small and numerous, but coalescenced to encomp ass large areas. After inoculation, with lesions were caused by E) and F) Note similar appearances of natural and artifi cially induced lesions caused by and 173

PAGE 174

174 Figure 4-3. Phylogeny for 44 accessions of based on ITS sequence data. The phylogram represents one of 5,412 MP trees (length 65, CI=0.778, RI=0.966). Bootstrap support values are shown at nodes. Nodes with asterisks were supported in maximum likelkihood analysis with >65% bootstrap support. Groupings defined by Sreenivasaprasad and Talhinha s (2005) are indicated w ith brackets. Type (ATCC 56816) and paratype (IMI 117619) strains of are marked with asterisks. *C. phormii C. lupini Ca60 Ca3 Ca13 Ca33 Ca26 Mango G. acutata IMI 360086 G. acutata PD85-694 G. acutata IMI 383015 G. acutata IMI 345027 G. acutata IMI223120 G. acutata S2 G. acutata 02-163 FL strawberry G. acutata STE-U 4459 G. acutata MEP1534 A2 G. acutata STF-FTP-10 FL sweet orange G. acutata OCO-ARC-4 FL sweet orange G. acutata 05-200 FL leatherleaf fern G. acutata IMI 117619 G. acutata STE-U 164 G. acutata ATCC 56816 G. acutata BBA 62124 C. lupini BBA 71249 C. lupini AR2820 C. lupini BBA 70358 C. lupini AR2826 A1 G. acutata Ss FL key lime G. acutata TOM-9 G. acutata TOM-21 A8 G63* ac utata ATCC MYA-662 G. acutata STE-U 5287 G. acutata PCN5 G. acutata MAFF 306282 G acutata BBA 70338 G. acutata Cooley2 G. acutata MEP1323 A3 C. phormii CBS198.35 G acutata IMI 345026 G. acutata S8 G. acutata CBS193.32 G. cingulata BBA 67435 C. phormii AR3410 A6 G. acutata PT250 C phormii MEP1334 A7. acutata IMI 345585 G C. phormii AR3787 C. phormii CBS199.35 A4, A6, A7 G. magna L2.5 C. gloeosporioides Cg131 G. graminicola M1.001 65 67 70 66 25 28 44 36 99 16 22 53 65* 80* 77* 61* Outgroup 5

PAGE 175

CHAPTER 5 CONCLUSIONS This study attem pted to answer hypotheses about host and tissue specificity of two tropical plant pathogens, and using phylogenetic analysis of DNA sequence data. Thes e questions were addressed with varying levels of success, and the importance of locus and taxon selection is illustrated in both phylogenetic studies. In the first study, the utility of the hos t-based biotype classification used for was evaluated. This biotype classification has traditionally been used to describe diversity within the species (de Arruda et al 2005), but a complete phylogenetic analysis investigating relationships among biotypes has not been completed. Additionally, several new hosts have recently been identified for whic h no biotype has been designated (R. Barreto, personal communication). The hypothe sis that previously describe d biotypes represent distinct evolutionary lineages that may comprise a species complex was investigated. In phylogenies of the ITS, IGS, and RPB1 regions indicated a level of host-specificity, and the presence of several distinct clades. For this taxon, all regions provided sufficient phylogenetic signal to separate the isol ates into several clades, although the RPB1 and IGS provided the strongest signal and the highest bootstrap suppor t for the clades. The IGS was especially rich in indel sites that, when used as characters in the data set, increased support for for the major clades. Most publis hed studies treat indels as mi ssing data, potentially ignoring important data when constructing phylogenies. The importance of taxon sampling is illustrate d in this study, which would have benefited from the inclusion of more accessions of some biotypes. For example, many isolates from cacao and were included, but only one isolate from three from other 175

PAGE 176

malpighiaceous hosts, and two from bignonaceo us lianas were included. The limited numbers of these isolates hindered an examination of a dist inct host-based lineage for this group. Future isolate collection in the pathoge ns geographic range from differe nt host species would enable a more robust analysis of the lineages that were identified in the present study. The second study dealt with rela tionships among populations of that are associated with mango. Pr evious studies indicated that a mango-specific population of exists (Hodson et al.1993, Alakahoon et al.1994, Hayden et al.1994), as well as a level of tissue specificity on mango (Gantotti and Davis 1993, Davis 1999). These previous studies relied on a small number of isolates and outdated molecular data, and were not accompanied by robust statistical analyses. The pres ent study re-investigated the lineages of this pathogen that are associated w ith anthracnose and blossom bli ght on mango with the specific hypotheses that: a) isolates from mango represent a distinct lineage within sensu lato that differ from isolat es that affect other tropical fru its; and b) these isolates display some level of tissue specificity. In the study, three regions were used to determine relationships among isolates. Three clades were re solved that supported the above hypotheses. Clade 1 contained only isolates from mango, and included is olates that caused bl ossom blight, leaf anthracnose, and fruit anthracnose. Clades 2 and 3 c ontained most fruit and peduncle isolates from mango and fruit isolates from other tropical hosts. They caused fruit anthracnose, but not leaf anthracnose and blossom blight, on mango. The three DNA regions provided varying levels of resolution. The ITS region was not very informative, with onl y 11 ingroup informativ e characters and low bootstrap support values for the clades. The MAT1-2 region, with 13 ingroup informative characters, resolved the mango clade (Clade 1) but not Clades 2 and 3. The CGTT5 region was 176

PAGE 177

more informative, with 61 informative sites and resolution of the th ree clades. These clades were also resolved by combining the congruent MAT1-2 and CGTT5 data sets. The varying degree of resolution achieved by the differe nt regions illustrated the impor tance of region selection during phylogenetic analyses. Cryptic species that we re found in the present studies are widespread and often associated with a specific host or ecological niche. As was demonstrated, molecular data can be especially useful for organisms with few morphological features. In there was evidence for and malpighiaceous-specific lineages although these associations were not absolute. In a mango-associated lineage was revealed. Whereas Clade 1 contained only mango isolates, Clades 2 and 3 included generalists from many hosts. Tissue specificity was apparent in the latter study in that the mango clade contained most of the isolates from green portions of the tree (leaves, panicles, peduncles) and immature fruit, whereas isolates from mature fruit usually fell into other clades. Pa thogenicity tests indicate d that isolates in the mango clade cause anthracnose on all of the above organs, but that those from the general clades also cause anthracnose on ripe fruit. Investigatio ns are warranted to determine why only isolates in Clade 1 are able to infect the above host tissues. Tissue specificity is exhibited by other taxa and lineages of On strawberry, is the major cause of fruit anthracnose, whereas and cause crown rot (Urea-Padilla et al. 2002). And on citrus, two lineages of cause either key lime anthracnose, which affects all young tissues of key lime, or postbloom fruit drop, which affects only young developing fruit of this and other spp. (Agostini et al. 1992, Timmer et al. 1994, Peres et al. 2008). 177

PAGE 178

According to the phylogenetic species concep t, phylogenetically distinct groups may be considered distinct species ev en if they do not possess defini ng morphological characters. As more molecular data becomes available and the pow er increases to resolve ever smaller groups, it will become increasingly important to esta blish the criteria for which species would be identified and what other charac ters might be considered when a new species is described. Adopting the evolutionary species concepts enable the incorporation of additional biological data to support the descript ion of new species (Wiley 1978, Tayl or et al. 2000). Agreeing upon and implementing such criteria would have tremendous academic and practical importance. There is a continous range of diversif ication between the individual and species. Carbone and Kohn (2001) discussed a population-species interf ace, wherein the threshold between populations and species can be somewhat arbitrary. In a ddition to monophyly, other criteria must be incorporated into a working species concept. For fungi, these include: morphological, ultrastuctural or biochemical characters; eco logical niche; host range; substrate; mating population; virulence; and, as shown for th e mango clade in Chapter 3, host and tissue specificity. With the exception of the L-biotype repor ted from Ecuador (Griffith and Hedger 1994a,b), utilizes a homothallic sexual re productive strategy. In contrast, appears to reproduce in the field only asexually. Whereas heterothallism enables adaptability via outcrossing and meitotic recomb ination, homothallism enables the fixation of genotypes that are well-s uited to hosts and environments in clonal populations (Milgroom 1996). Entire genotypes rather than indi vidual genes are propagated in clonal populations (Milgroom 1996), thus leading to the phylogene tic patterns observed in the pr esent studies: closely related, host-associate lineages with low diversity within clades. 178

PAGE 179

Homothallis m confers the following advantag es to a species. Because homothallic species are self-fertile, there is no need for different mating types and sexual reproduction can take place even in small populations. Homotha llic species may become highly evolved in specific environments, and outbreeding may actua lly dilute these specific adaptations. In monocultures that prevail in agriculture, hom othallism enables reproduction even when small numbers of individuals survive control measur es; these populations can develop into highly adapted and stable genetic lines (Wheel er 1954, Anderson et al. 1992, Milgroom 1996). Asexual pathogens, such as possess the same attributes. Given the prevalence of anamorphic plant-pathogenic genera (e.g. and ), one might ask how asexual and homoth allic sexual taxa differ and whether one reproductive strategy is superior to the other. Although both should result in identical offspring, there is some evidence that homothallic species can outcross occasionally, resulting in variation not seen in clonal species. Examples include many oomycetes (Cooke et al. 2005, Francis et al. 1994, Forster et al. 1994, Francis and St. Clair 1993), (Johnson et al.2004), and, based on incongruence noted for isolates CPB7 and CPB9 in Chapter 2, possibly Retention of a sexual cycle has also been shown to reduce the accumulation of deleterious mutations in populations (Bruggeman et al 2003). Asexual and sexual strains of were grown in culture for 38 genera tions and the growth rate of the 39th generation was used as a measurement of fitness. Although both type s of strains lost fitness, the asexual strains were significantly less fit. Phylogenomics will provide larg er character sets for phyloge netic analyses, and enable genome evolution to be tracked between closely related species. For example, over 90% of the genomes of and a closely related pathogen of wild grass were 179

PAGE 180

similar, but considerable rearrangeme nt ha d occurred in their respective dispensable chromosomes (Stuckenbrock et al. 2009). A similar study has been undertaken comparing and (Meinhardt et al. 2008a). They report that 70% of the gene sequences in and had at least 80% sequence identity and that the species shared several pathogenicity factors. Genomic comparisons of biotypes and homotha llic and heterothallic lineages of would provide insight into the evolution of this pathogen, and could help define the role that host jumps, as opposed to coevolution, played in the development of different host associations and specificity in the species. Additional work is needed to define and the species it contains. The genus now contains just six species (A ime and Phillips-Mora 2005, Kerekes and Desjardin 2009). species in section Iopodinae, such as should be reexamined and included in molecular phylogenies to determine if they shou ld be transferred to Additional collections in South America may identify new hosts of and host relationships in the species and genus. Three lineages of sensu lato occur on mango in South Florida, one of which, Clade 1, could be considered a distinct species similar to and Sequence data for the regions that were used in th e present study should be obtained for the type specimen, IMI 356878 (Cannon et al.2008), to determine if any of any these lineages are sensu stricto. To fully understand variation in sensu lato and the sp ecies that it contains, isolates from additional geographic regions and hosts should also be examined to dete rmine their relationships to one another and to sensu stricto. Currently, only the ITS sequence is available for the type. Although it was not included in the ITS analysis in the present study, when compared to 180

PAGE 181

sequences for Clade1, 2, and 3 isolates, the t ype sequence differs by 5 bases (1.0%), 3 bases (0.6%), and 4 bases (0.8%), re spectively (data not shown). Since th ese differences are comparable to those that separate and the mango clade may represent a new species in the sensu lato complex. Studies like the one presented here are first steps in clarifying the complexity of sensu lato. Once phylogenetic lineages are identified, additional host range, substrate, morphology, geographic and ecologica l characters are needed to understand the biological and evolutionary importa nce of these distinctions. One of the strengths of this study is that the identification of the mango and general clades is accompanied by pathogenicity and morphology data. Future investigations involving isolates belonging to these clades will add further understanding of the biology and ecology of evolutionary species of sensu lato on mango Work is needed on the infection strategies of the different clades on mango and on the physiology of the host-pathogen in teraction. The nature of quies cent infections of fruit by has been well characterized on avocado fruits, and include alkalization of host tissue, activation of p ectate lyase, and producti on of sugars (Prusky and Lichter 2007, KramerHaimovich et al. 2006). Similar topics should be evaluated for the different clades that occur on mango. Microscopic study of the inf ection processes of isolates fr om different clades on leaves and fruits could clarify the physical nature of infection on different host tissues, and whether those from all lineages are able to infect and ulti mately cause disease. The tolerance of isolates to antifungal resorcinols found in mang o peels (Hassan et al. 2007, Kndler et al. 2007), as well as the levels of these compounds that occur in leaves and inflorescences should also be 181

PAGE 182

182 characterized to determine the roles that reso rcinols play in the development of disease on different organs. These types of studies are a logical progr ession from phylogenetic st udies that simply identify the existence of distinct clades. The next step is to understand the roles that the different lineages play in the cacao and mango agroecosy stems, and to inform future ecological, physiological, pathological research.

PAGE 183

APPENDIX A IDENTIFICATION AND CHARACTERIZAT ION OF FIVE RANDOMLY AMPLIFIED DNA REGIONS FLANKING MI CROSATELLITE REPEATS FOR USE IN PHYLOGENETIC ANALYSIS OF is a pathogen of hundreds of hosts worldwide. Tropical fruit, such as mango, avocado, papaya and bana na, are among the most important hosts and can suffer immense losses to this pathogen. Informati on on the population struct ure and diversity in would help understand the epidemiology of the important diseases it causes and could be used to improve management methods. Various molecular markers have been used to study population stru cture and phylogenetics in Most common have been random amplified polymorphic DNAs (RAPDs) primed with commercially available RAPD primer s. However, other markers have been obtained with arbitrarily primed-PCR (a p-PCR), random amplified microsatellite (RAMS), microsatelliteprimed PCR (mp-PCR), and internal transcribe d spacer (ITS) sequences (Freeman et al. 2000, Martinez-Culebras et al. 2002, Urea-Padilla et al. 2002, Afanador-K afuri et al. 2003, Mahuku and Riascos 2004, Xiao et al. 2004, Abang et al. 2005, Talhinas et al. 2005, Munaut et al. 2002, Denoyes-Rothan et al. 2003, Martin ez-Culebras et al. 2003). While these studies have suggested high levels of diversity within the species, the development of better markers is needed to fully understand the extent and im portance of variation in sensu lato. For example, due to low numbers of informative sites, ITS se quence data do not give high resolution within sensu lato, which includes and RAPD markers also have well-documented problems of repeatability and the equivocal identity of comigrating bands (Harris 1999). 183

PAGE 184

More reliable and meaningful ma rkers, such as simple sequence repeats (SSRs), single nucleotide polymorphisms (SNPs) and polymorphic DNA sequences, have been applied to a wide range of biological questions (Agarwal et al. 2008). With re gard to the DNA sequence data, polymorphic loci were identified and developed in the present study to construct phylogenies in sensu lato. Bands that were gene rated with mp-PCR in isolates of from mango were cloned and sequenced with the sequence characterized amplified region (SCAR) technique (McDermott et al.1994). Five new loci were developed, and the extent of polymorphism at each was determined in closely related strains. RAPD analyses were conducted with 24 isolates of from mango leaves and fruit. Isolates were grown for 3-5 days in potato dextrose broth (PDB) at room temperature on a shaker at approx. 80 rpm. Mycelia were harves ted, triple rinsed with sterile deionized water, dried on sterile filter paper, and DNA was extr acted using a DNA genomic preparation protocol from the University of Wisconsin Biotechnology Center. Briefly, mycelia were ground in 500 l Shorty DNA Extraction Buffer (0.2M Tris-HCl pH 9.0, 0.4M LiCl, 25mM EDTA, 1% SDS) and incubated at 68C for 10 minutes. Tissue was centrifuged at 14,000 rpm for 5 minutes, 400 l of the supernatant was tr ansferred to new tube and the DNA was precipitated with 400 l 99% isopropanol. The tubes were centrifuged for 10 minutes at 14,000 rpm and the supernatant was decanted. The DNA pellets were air-dried for 5 minutes then resuspended in 400 l TE buffer (10 mM Tris-Cl, pH 7.5, 1 mM EDTA) for 30 minutes at room temperature. 2l DNA was used in PCR reactions. Three simple sequence repeat primers were used to produce RAPD profiles: (TCC)5, (GACA)4 and (ACTG)4. PCRs were conducted in 50 l reac tions that contained: 39.25 l of sterile distilled, deionized wate r; 6.5 l of ThermoPol Reaction Buffer (New England Biolabs, 184

PAGE 185

Ipswich, MA); 1 l of 10 mM dN TP m ix (New England Biolabs, Ipswich, MA); 0.25 l of Taq DNA polymerase (conc. 5,000 units/ml); 1 l of 15 M primer; and 2 l of DNA template. Standard cycling parameters with a 55C anneal ing temperature were used. PCR products were separated on a 1.5% agarose gel at 60v for 60 mi n. RAPD band patterns for each primer x isolate combination were compared, and three bands that were present in all isolates were cloned and sequenced. Each of the bands from three differe nt isolates (Cg43, Cg49, and Cg73) was cut from gels with a razor blade, and DNA was extracte d with QIAQuick Gel Extraction Kit (Qiagen) following manufacturer instructions Bands were cloned using pGEM-T Easy Vector System (Promega). Plasmid DNA was extracted from th e cloning vectors with QIAprep Spin Miniprep Kit (Qiagen), and DNA inserts for one or two isol ates from each region were sequenced at the University of Florida ICBR. The homology of the resulting sequences was compared to others in GenBank with BLASTn. SCAR primers were designed using Primer3 and tested on isolates of , and of from avocado, guava, carambola and mango. In addition, 11 mango isolates (Cg20, Cg66, Cg101, Cg127, Cg128, Cg129, Cg130, Cg131, Cg134, Cg135, Cg136) and one guava isolate (G ua1) were chosen for sequencing of the CGTT-3 to CGTT-6 (CGTT= cloned by T T arnowski) loci to determine how polymorphic these loci were in closely related isolates of RAPD banding patterns were variable among isolates with all three primers. From the (GACA)4 profile, 950bp and 1,200bp bands were cloned, and from the (TCC)5 profile a 1,000bp band was cloned. Two to three cloned inserts we re sequenced for each band from which five distinct regions were identifie d (Table A-1). Clearly, comigra ting RAPD bands can represent different DNA sequences. With the exception of the CGTT-1 primers, which produced a faint 185

PAGE 186

186 additional 1500 bp band for some isolates, each primer pair generated clean, single-band products (Fig. A-1). The CGTT-3 to CGTT-6 primers were used to directly sequence 12 isolates. The size and BLAST homologies of amplified products, as well as the number of polymorphic sites at each locus are recorded in Table A-2. All lo ci were amplified for isolates of from avocado, as well as For other species, only the CGTT-6 region was successfully amplified (Figure A-2). Thus, CGTT-6 appeared to be conserved in several species, whereas the other loci were only present in sensu lato. Only the CGTT-5 locus was homologous to other sequ ences of hypothetical proteins in and in GenBank (Table A-2). Based on the large number of polymorphisms that were present, CGTT-3 to CGTT-6 should be useful regions for popul ation and phylogenetic studies of sensu lato. The number of polymorphisms among 12 isolat es in this study ranged from 11-38, whereas the ITS and MAT1-2 regions yielded, respectivel y, only 11 and 13 informative characters in over 50 accessions in Chapter 3. In the present study th e CGTT-5 locus discriminated lineages in this species complex, and increased re solution and internal support wh en used with data from other DNA regions (Chapter 3). Because is a species complex, these regions would be useful to sort out closely re lated natural lineages in the co mplex. These regions can also be used in conjunction with already established loci to increase th e resolution and internal support of phylogenies. The CGTT regions de scribed have advantages over the RAPD type markers most often used in studies in that they are universal (they can be used by any lab and are reproducible).

PAGE 187

Table A-1. Origin of, and st atistics for, CGTT loci. RAPD bands cloned (GACA)4 (GACA)4 (TCC)5 Size 950bp 1,200bp 1,000bp # clone inserts sequenced 3 (2 from Cg49; 1 from Cg43) 3 (2 from Cg73; 1 from Cg49) 2 (from Cg49) # loci sequenced Three comigrating RAPD bands were cloned, an s fro m each were sequenced. Five distinct loci were identified d two to three insert 2 CGTT3 (860bp) CGTT4 (852bp) 1 CGTT1 (873 bp); 3 end of region not successfully sequenced) 2 CGTT5 (895bp) CGTT6 (891bp) 187

PAGE 188

188 Table A-2. PCR primers, product si ze, homologies, and the extent of polymorphism of CGTT regions Locus Primers PCR product length (bp) Homologya # polymorphism (% polymorphic)b CGTT-1 Forward: 5-AGGTGAGCCAACCTGTCAGT-3 Reverse: 5-AGCAGTCACAGGCACACATC-3 662 none n/a CGTT-3 Forward: 5-CGATGACGATGATAAGGTG-3 Reverse: 5-AGCAGTCACAGGCACACATC-3 669 none 38 (5.6%) CGTT-4 Forward: 5-AGACTGATGAAATGCGATGC-3 Reverse: 5-CTTACATGCCCCTGTTCCAT-3 813 none 38 (4.7%) CGTT-5 Forward: 5-CCCTCAGATTTCAGCCAAAG-3 Reverse: 5-GCTCTCTCGCTGCTTCATCT-3 564 hypothetical protein 38 (6.7%) CGTT-6 Forward: 5-CGGGGCACACTAACGTAAAA-3 Reverse: 5-TCATGGGGCTTCCTATTCAG-3 560 none 11 (2.0%) a Homology with sequences in Ge nbank using BLASTn algorithm. b Extent of polymorphism in a given locus for 12 isolates of from mango and guava, followed by percent of polymorphic sites in parantheses

PAGE 189

C 1 2 B 3 1 4 2 5 3 6 4 7 5 8 6 9 7 10 8 1 9 2 10 3 A 4 1 5 2 6 3 7 4 8 5 9 6 10 7 9 10 8 D Figure A-1. Amplification of A) CGTT-1, B) CGTT-3, C) CGTT-4, D) CGTT-5, a nd E) CGTT6 loci with primers described in Ta ble 1. Lanes 1-10: 1=DNA ladder, 2=Cg20, 3=Cg66, 4=Cg101, 5=Cg127, 6=Cg127, 7=Cg128, 8=Cg129, 9=Cg131, 10=Cg134. 6 4 5 3 2 1 E 189

PAGE 190

190 A 9 B 11 10 8 1 5 2 7 6 3 3 4 1 4 2 1 5 10 6 6 7 4 8 2 7 5 3 9 11 10 8 11 9 C 1 2 3 4 5 6 7 8 9 10 11 Figure A-2. Amplification of A) CGTT-3, B) CGTT-4, C) CGTT-5, and D) CGTT-6 loci in isolates Avo1 and Avo2, lanes 2 and 3; isolates Ban1 and Ban2, lanes 4 and 5; isolates Ca3 and Ca13, lanes 6 and 7; isolates Pap14, Pas5, lanes 8 and 9; isolates Pas1, Piper, lanes 10 and 11. Lane 1 =DNA ladder. D

PAGE 191

APPENDIX B FIRST REPORT OF CAUSING POSTHARVEST ANTHRACNOSE ON PAPAYA IN SOUTH FLORIDA Postharvest anthracnose of papaya is an important disease in most production areas worldwide (Persley and Ploetz 2003). Two types of symptoms have been associated with : 1) circular, sunken lesions with pink sporulation; and 2) sharply defined, reddish brown and sunke n lesions, described as cho colate spot (Dickman 1994, Persley and Ploetz 2003). Several / species were recently recovered from anthracnose lesions on papaya fruit from th e University of Florid a Tropical Research and Education Center (TREC), Homestead, FL in December of 2007, and from fruit imported to a local packinghouse from Belize in March of 2008. Eight lesions on four fruit from TREC and 19 lesions on eight fruit from Belize we re examined and the incidence of and a sp. were calculated (Table B-1). Le sions were variable and either: 1) sunken with darkly pigmented centers and masses of pink sporulation; or 2) initially brown and circular, becoming sunken with copious dark sporulation (Fig. B-1A-B). Fungi were isolated from lesions by streak ing sporulating acervuli on PDA and plating individual germinating spores. Species were identified using the mo rphology of colonies on PDA, conidium shape, whether acervuli were setose or glabrous, and whether a teleomorph formed in culture. Molecular identities were al so obtained for representa tives of the different morphological groups with ITS sequences that we re aligned with sequen ces of several other species in Genbank (Table B-2). Maximum parsimony analysis was performed in Mega 4.0 (Tamura et al. 2007) using closest ne ighbor interchange (CNI) searches with 100 random taxon additions and gaps were treated as missing data. Bootstrap analysis according to Felsenstein (1985) was performed using the above search criteria, with 500 repetitions. 191

PAGE 192

Three taxa were identif ied: and sp. (Table B1). Colonies of were white to grey, fluffy with orange sporulation and straight and cylindrical conidia, and those of had sparse, fluffy, white mycelia with setose acervuli and falcate conidia (Fig. B-2A-B). Colonies of sp. were darkly pigmented and produced fertile perithecia after 7-10 days. The ITS-based phylogeny (Fig. B-3) identified setose, falcate-spored is olates from papaya as (Pap14, Pap18), and isolates that formed a teleomorph as a nondescribed species (Pap3-6, 8). was most often associated with lesion type 3. Mature fruit (cv. Car ibbean Red) obtained from Brooks Tropicals Inc. (Homestead, FL) were wounded with a sterile needle and inoculat ed with a 15l drop of 0.3% water agar that contained 105 conidia ml-1of a given isolate (Table B-3). Inocul ated fruit were incubated in moist chambers at 25C in the dark in a randomized complete block experiment, wherein four (exp1), five (exp2), or seven treatments (exp3) (two isolates of each taxon recovered, and a mockinoculated control) were replicated six times on single fruit experimental units (=blocks). Experiments compared the pathogenicity of: 1) isolates of to sp.; 2) isolates of to ; and 3) all three taxa. Th e diameters of developing lesions were measured after 7 days. Lesion margin s were excised from fru it, surface disinfested and plated on PDA to confirm the presence of th e inoculated isolate. Analysis of variance was performed on lesion diameter values using PROC GLM in SAS v9.1.3 (SAS Institute Inc., Cary, NC) and means were separated with Fishers LSD. In all experiments, and both produced lesions significantly larger than the water control and sp. ( <0.05; Table B-3). sp. does not appear to cause anthracnose on pa paya, and it is not clear what role, if any, it played in the 192

PAGE 193

development of the lesions from whi ch it was re covered. The taxon has been reported most often as an endophyte of se veral hosts, including and (Farr et al. 2006), and causes anthracnose lesions on passionfr uit (Appendix C). It is possible that the isolates of sp. that were recovered from papaya lesions had colonized lesions that were caused by another pathogen. produced sunken lesions with dark grey centers and pink/grey sporulati on, which are typical of and match previous descriptions for anthracnose on papaya (Fig. B-3A; Dickman 1994, Persley and Ploetz 2003). In contrast, produced lesions with da rk sporulation, due to copious setae in acervuli (Fig. B-3B). A similar type of lesion on papaya, caused by has been reported from the Yucatan Peninsula (Tapia -Tussell et al. 2008). This is the first report of as a cause of anthracnose on papaya in Florida, and perhaps Belize (fruit from which this pathogen was recovered may have been latently infected prior to importation). The only othe r known report of a falcate-spored species on papaya in the United States is of which caused a fruit rot in Texas (Anonymous 1960). has been reported as a causal agent of papaya anthracnose in Japan (Yagushi et al. 1998), Malaysia (Rahman et al 2008), and in the Yucatan Peninsula, Mexico (Tapia-Tussell et al. 2008). 193

PAGE 194

194 Table B-1. Number of lesions from which species were recovered from fruit from two locations. # lesionsa Species TREC Imported 0 4 (21.1) 2 (25) 7 (36.8) sp. 2 (25) 0 a Eight lesions were sampled from from fruit fromTREC, and 19 were from imported fruit from Belize. Total numbers of lesions are followed pa renthetically by percenta ges of lesions from which a species was recovered.

PAGE 195

Table B-2. ITS sequences includ ed in phylogenetic analysis. Accession Species Host Origin Genbank accession # Publisheda Florida GQ373192 Chapter 3 Ban1 Florida GQ373193 Chapter 3 Ban2 Florida GQ373194 Chapter 3 Ban3 Florida GQ373215 Chapter 3 Cg101 Florida GQ373218 Chapter 3 Cg129 Florida GQ373223 Chapter 3 Cg131 IMI 356878b Italy EU371022 Cannon et al. 2009 Florida GU045514 --Pap19 Florida GU045511 --Papleaf Florida GU045512 --Pap14 Florida GU045513 --Pap18 Florida GU045515 --Pas5 Tanzania AY376526 Lubbe et al. 2005 STE-U 5304 USA DQ286156 Farr et al. 2006 AR4028 CBS120709c India EF683602 Shenoy et al. 2007, unpublished Japan AB196301 Moriwaki et al. 2002 MAFF 238714 Mexico DQ286154 Farr et al. 2006 AR3563 Mexico DQ286221 Farr et al. 2006 AR2930 Netherlands DQ286219 Farr et al. 2006 CBS318.79 Florida GU045516 Pas1 MAFF 305972d Japan AB051400 Moriwaki et al. 2003 Taiwan EF503672 Chen et al. 2007, unpublished CB0616 New Zealand DQ286172 Farr et al. 2006 CBS102667 China DQ286170 Farr et al. 2006 AR3751 sp. Florida GU045517 --Pas3 sp. Florida GU045522 --Pap6 sp. Florida GU045520 --Pap4 sp. Florida GU045521 --Pap5 sp. Florida GU045518 --Pap8 Piper5 sp. Florida GU045521 --195

PAGE 196

Table B-2. Continued. Accession Species Host Origin Genbank accession # Published sp. Mexico AY566308 Espinoza-Ortega et al. 2004, unpublished AC4-MMexico sp. Panama EF423535 Gilbert et al. 2007 P060 sp. Mexico AY841136 Villanuev a-Arce et al. 2004, unpublished MICH-3-345 sp. Costa Rica DQ286217 Farr et al. 2006 MCA2773 sp. Thailand DQ286215 Farr et al. 2006 AR3750 sp. Thailand DQ286213 Farr et al. 2006 AR3749 sp. Japan AB042319 Moriwaki et al. 2002 MAFF 305974 Florida GU045506 Chapter 4 Ca3 Florida GU045507 Chapter 4 Ca13 Florida GU045508 Chapter 4 Ca26 ATCC 56816e Australia DQ286132 Farr et al. 2006 England DQ286144 Farr et al. 2006 CBS198.35 Outgroup --GU060637 Bollig et al. 2009, unpublished TomIGZ --DQ825977 Zare et al. 2007 CBS 130.51 --GQ258661 Carlucci et al. 2009 Vt536 a Publication in which ITS sequence was published. b Type speciment for Type speciment for d Type specimen for e Type specimen for 196

PAGE 197

197 Table B-3. Lesion diameters from pathogenicity experiments. Exp1 Exp2 Exp3 Isolate Species Lesion diameter (mm)a Lesion diameter (mm) Lesion diameter (mm) Pap8 sp. 0.63 0.38 b --0.60 0.60 c Pap9 sp. 0.25 0.25 b --0.33 0.33 c Pap10 --20.75 0.75 bc 12.92 1.80 b Pap14 --18.00 3.66 c 10.00 3.58 b Pap11 18.25 5.02 a ----Pap16 --25.67 3.72 ab 22.42 3.77 a Pap17 --29.25 3.67 a 12.4 6.00 b Control --0.00 0.00 b 0.00 0.00 d 0.00 0.00 c LSD 2.262 8.12 6.79 -value 0.0017 <0.0001 <0.0001 a Lesion diameters 7 days after inoculation b Values in columns with same letter are not significantly different based on Fishers LSD, P=0.05.

PAGE 198

B A Figure B-1. Lesions associated with began as A) small brown and slightly sunken areas that became B) increasingly s unken and eventually covered with dark sporulation. 198

PAGE 199

A BFigure B-2. Morphological characteristics of recovered from papaya: A) acervu lus with setae, and B) falcate conidia. 199

PAGE 200

Cg129 Cg131 IMI 356878 EU371022* Cg101 Ban3 Ban1 Ban2 C. musae C. gloeosporiodes Papleaf CBS318.79 DQ286219 AR3920 DQ286221 C. agave CBS120709 EF683602* STE-U 5304 AY376526 Pap19 Pas5 Pap14 Pap18 AR4028 DQ286156 C. capsici Pas1 CBO616 EF503672 AR3751 DQ286170 MAFF 305972 AB051400* CBS102667 DQ286172 C. boninense MAFF 238714 AB196301 AR3563 DQ286154 C. dematium Ca3 Ca26 Ca13 ATCC 56816 DQ286132 C. phormii CBS198.35 DQ286144 C. acutatum MCA2773 DQ286217 Piper5 MICH-3-345 AY841136 AC4-M-Mexico AY566308 AR3750 DQ286215 AR3749 DQ286213 Pap4 P060 EF423535 Pap5 Pap8 MAFF 305974 AB042319 Pas3 Pap6 Glomerella sp. V. albo-atrum CBS 130.51 DQ825977 V. tricorpus Vt536 GQ258661 V. dahliae TomIGZ GU60637 Outgroup 70 73 91 77 71 68 68 99 71 69 63 99 33 21 38 55 100 63 63 77 34 5 Figure B-3. One of 2,255 most parsimonious IT S-based trees (101 steps, CI=0.796, RI=0.967) for taxa recovered from papaya and other hos ts. The data se t c ontained 38 accessions with 322 characters, 53 of which were pars imony informative. Bootstrap values over 40% are listed at nodes. Isolat es with an asterisk are types for a given species, and papaya isolates are design ated with the prefix Pap, passionfruit with prefix Pas. 200

PAGE 201

201 A BFigure B-4. Lesion produced after a rtificial inoculation with A) and B)

PAGE 202

APPENDIX C FIRST REPORT OF AND A SP. CAUSING POSTHARVEST ANTHRACNOSE ON PASSIONFRUIT ( SPP.) IN FLORIDA. Anthracnose, caused by several species (Alfieri et al. 1994, Kobayashi and Okamoto 2003, Manicom et al. 2003, Moriwaki et al. 2003), is one of the most important diseases on leaves, flowers and fruit of passi onfruit (Manicom et al. 2003). In October 2008, yellow and purple passion fruit ( f. and respectively) with postharvest anthracnose were examined in Mi ami-Dade County, FL. Lesions began as light brown areas that expanded to cover half of the fr uit surface in 3-4 days. The skin became papery, and some lesions developed a dark brown pigment. Pink to dark sporulation developed first in lesion centers, and then spread toward the margins. Fungi were isolated from lesions by st reaking sporulating acervuli on PDA and recovering individual germinating spores. Isolates were identi fied morphologically, based on colony appearance on PDA, conidium and ascospor e shape, whether acervuli were setose or glabrous, and whether a teleomorph was formed in culture. Molecular identities were also obtained for single representative s of each of three morphological groups with ITS sequences. Sequences were aligned with those for several other species in Genbank, and included the type strains of (ATCC 56816), (IMI 356878), (MAFF 305972) and (CBS 120709). Maximum parsimony analysis was performed in Mega 4.0 (Tamura et al. 2007) usin g closest neighbor interchange (CNI) searches with 100 random taxon additions and gaps were treated as missing data. Bootstrap analysis according to Felsenstein (1985) was performed using the above search criteria, with 500 repetitions. 202

PAGE 203

Four taxa were identified: and sp. Colonies of were white to grey, fluffy with orange sporulation and straight and cylindrical conidia; those of colonies were cream to orange and felted, with cream-colored sporulation and cy lindrical conidia; those of had sparse, fluffy, white mycelia with setose acervuli and falcate conidia; and those of sp. were darkly pigmented and produced fertile perithecia after 7-10 days (Fig. C-2A-D). In the ITS-based phylogeny, a setose falcate-spored is olate, Pas5, was identified as a straight-spored isolate with cream-colored colonies was Pas1, and an isolate that formed a teleomorph in culture was an undescribed species (Pas3) (Appe ndix B, Figure B-3). Inoculation studies were conducted with several isolates from each taxon, including several isolates of sp. that were recovered from other hosts (Table C-1). Mature yellow passionfruit were wounded with a sterile needle and inocul ated with a 15l drop of 0.3% water agar that contained 105 conidia ml-1of a given isolate. Inoculated fruit were incubated in moist chambers at 25C in the dark. A random ized completely block design (RCBD) was used, where experimental units were si ngle fruits inoculated with one of seven (Exp. 1) or 11 (Exp. 2) isolate treatments (Table C-1). Each isolate x treatment combination was replicated four times, and the two repetitions of the experiment were treated as blocks. Incidences of lesion development, which coincided with fruit softening, were recorded for 21 days after inoculation, where no lesion development was scored an d lesion development was scored Lesion diameters were not recorded because lesions developed at different times throughout the experiment, depending on maturity of individual fruit. Lesion ma rgins were excised from fruit, surface disinfested and plated on PDA to confirm pr esence of the inoculated isolate. Analysis of 203

PAGE 204

variance was performed for inci dence values using PROC GL M in SAS v9.1.3 (SAS Institute Inc., Cary, NC) and m eans were se parated with Fishers LSD. In both experiments, all accessions except those of and Pas2 of in Exp. 2 caused at leas t a 50% lesion incidence (Table C-1). The inoculated species were successfully recovered from lesions. Lesi ons that developed from inoculations with sp. produced dark brown pigment in the cen ter of the lesion, whereas lesions caused by other isolates were darkened due to extensive sporula tion (Fig C-1A-C). In the present study, three new species were id entified as causes of fruit anthracnose of passionfruit in Florida. has been associated with leaf anthracnose of passionfruit in Florida and Ja pan (Alfieri et al. 1994, Kobayashi and Okamoto 2003), but has apparently not been associated previously with fruit anthracnose. is a relatively recently erec ted species (Moriwaki et al. 2003), and has been associated with passionfruit anthracnose in Japan (Moriwaki et al. 2003) and Colombia (Freeman, personal communication). The undescribed sp. caused anthracnose on passi onfruit, but in other work caused no disease on papaya (Appendix B) and e ugenia (Ploetz et al.2009), two other hosts in South Florida from which it has been recovered. This fungus produces the teleomorph readily in culture, as well as directly on sporulating lesi ons on passionfruit (Figs. C-2A-D). Almeida and Colho (2007) characterized a sp. that caused anthracnose on passionfruit in Brazil, but DNA sequences of these isolates are not avai lable. Additional work to characterize this apparently new species is underway. 204

PAGE 205

Table C-1. Incidence of symptom s from two inoculation studies Incidencea Isolate Species Host Exp 1 Passionfruit, 0.13 0.13 bcb Pas4 Papaya, cv. Caribbean Red 0.00 0.00 c Pap11 sp. Papaya, cv. Caribbean Red 0.75 0.25 a Pap8 sp. Papaya, cv. Caribbean Red 0.63 0.18 ab Pap9 sp. Passionfruit, 0.75 0.16 a Pas3 sp. endophyte 0.50 0.29 abc 07-598 sp. Passionfruit, 0.50 0.29 abc pGlom Passionfruit, 0.75 0.16 a Pas2 Passionfruit, 0.75 0.25 a Pas1 Passionfruit, 0.75 0.25 a Pas7 Passionfruit, 0.75 0.25 a Pas5 Passionfruit, 0.50 0.29 abc Pas6 Control ----0.00 0.00 c LSD 0.5664 = 0.0065 aIncidence was rated for each experimental unit as for no lesion development, and if a lesion developed for b Values in columns are means from two experiments that, when followed by the same letter, are not significantly different based on Fishers LSD, P=0.05.. 205

PAGE 206

B C A Figure C-1. Anthracnose lesion development on inoc ulated yellow passionfruit. A) young lesions start as circul ar light brown areas and quick ly enlarge to B) large sporulating lesions developed on fruit inoculated with and C) large sporulating lesions with dark brown pigment developed on fruit inoculated with sp. 206

PAGE 207

207 B A D CFigure C-2. Morphology of sp. A) perithecia on fruit su rface, B) cross-section of perithecium, C) asci erupting from perithecia, and D) ascospores

PAGE 208

LIST OF REFERENCES Abang MM, Abraham WR, Asiedu R, Hoffman P, Wolf G, Winter S. 2009. Secondary metabolite profile and phytotoxic activity of genetically distinct forms of from yam ( spp.). Mycol Res 113:130-140. Fagbola O, Smalla K, Winter S. 2005. Tw o genetically distinct populations of Penz. causing anthracnose disease of yam ( spp.). J Phytopathol 153:137-142. Winter S, Green KR, Hoffman P, Mignouna HD, Wo lf GA. 2002. Molecular identification of causing yam anthracnose in Nigeria. Plant Pathol 51:63-71. Afanador-Kafuri L, Minz D, Matmon M, and Freeman S. 2003. Characterization of isolates from tamarillo, passiflora, and mango in Colombia and identification of a unique species from the genus. Phytopathology 93:579-587. Agarwal M, Shrivastava N, and Padh H. 2008. A dvances in molecular marker techniques and their applications in plant sc iences. Plant Cell Rep 27:627-631. Agostini JP, Timmer LW, and Mitchell DJ. 1992. Mo rphological and pathol ogical characteristics of strains of from citrus. Phytopathology 82:1377-1382. Aguileta G, Marthey S, Chiapello H, Lebr un M-H, Rudolphe F, Fournier E, GendraultJacquemard A, Giraud T. 2008. Assessing the performance of single-copy genes for recovering robust phylogenie s. Syst Biol 57:613-627. Aime MC. 2001. Biosystematic studies in Crepid otus and the Crepidot aceae (Basidiomycetes, Agaricales). Ph.D. Dissertation, Vi rginia Tech, Blacksburg, VA. 194 pp. Matheny PB, Henk DA, Frieders EM, Nilsson RH, Piepenbring M, McLaughlin DJ, Szabo LJ, Begerow D, Sampaio JP, Bauer R, Wei, Oberwinkler F, Hibbett D. 2006. An overview of the higher level classification of Pucciniomycotina based on combined analyses of nuclear large and small subunit rDNA sequences. Mycologia 98:896-905. Phillips-Mora W. 2005. The causal agent of w itches broom and frosty pod rot of cacao (chocolate, ), form a new lineage of Marasmiaceae. Mycologia 97:10121022. Alakahoon PW, Brown AE, Sreenivasaprasa d S. 1995. Genetic characterization of isolates obtained from ma ngo. Int J Pest Manage 40:225229. , 1994. Cross-infection potential of genetic groups of on tropical fruits. Ph ysiol Mol Plant P 44:93-103. 208

PAGE 209

Alexopolous CJ, Mims CW, M Blackwell. 1996. In troductory Mycology. New York: John Wiley & Sons, Inc. 880 p. Alfieri Jr. SA, Langdon KR, Kimbrough JW, El-Gholl NE, Wehlburg C. 1994. Diseases and disorders of plants in Flor ida. Bulletin 14. Florida Department of Agriculture and Consumer Sciences, Department of Plant Industy. Gainesville, FL. Ali K, Koeda K, Nii N. 1999. Changes in anatomical features, pigment content and photosynthetic activity related to age of Irwin mango leaves. J Japan Soc Hort Sci 68:1090-1098. Almeida LCC, Colho RSB. 2007. Caracteriza o de agressividade de isolados de de maracuj amarelo com marcadores bioqumico, fisiolgico e molecular. Fitopatol Bras 32:318-328. lvarez I, Wendel JF. 2003. Ribosomal ITS sequen ces and plant phylogenetic inference. Mol Phylogenet Evol 29:417-434. Anderbrhan T, Furtek DB. 1994. Random amplified polymorphic DNA (RAPD) analysis of isolates from different hosts. Plant Pathol 43:1020-1027. Anderson JB, Kohn LM, Leslie JF. 1992. Genetic M echanisms in Fungal Adaptation. In: Carrol GC, Wicklow DT, eds. The Fungal Commun ity: its Organization and Role in the Ecosystem. New York, NY: Marcel Dekker, Inc. 929 p. Andrews JH. 1995. Fungi and the evolution of growth form. Can J Bot 73:S1206-S1212. Aneja M, Gianfanga TJ, Hebbar PK. 2005. produces nonanoic acid, an inhibitor of spore germination and mycelia gr owth of two cacao pathogens. Physiolog Mol Plant P 67:304-307. Anonymous. 1960. Index of Plant Diseases in th e United States. U.S. Dept. of Agric. Handbook. No. 165. Washington, D.C. 531 pp. Anonymous. 2009. FRAC Code List*: Fungicides so rted by mode of action (including FRAC Code numbering). Fungicide Resistance Ac tion Committee, CropLife International. http://www.frac.info/frac/publi cation/anhang/FRAC_ CODE_LIST.pdf Accessed 11/06/09. Appel D, and Gordon TR. 1995. Intraspeci fic variation within populations of based on RFLP analysis of the interg enic spacer region of the rDNA. Exp Mycol 19:120-128. Appiah AA, Flood J, Archer SA, Brisge PD. 2004. Molecular analysis of the major species on cocoa. Pl ant Pathol 53:209-219. 209

PAGE 210

Arauz, LF. 2000. Mango anthracnose: economic impact and current options for integrated m anagement. Plant Dis. 84:600-611. Avila-Quezada GD, Silva-Rojas HV, Te liz-Ortiz D, Zelaya-Molina LX. 2007. spp. Associated to anthracnose on avocado fruits in Mexico. Phytopathology 97:S6. Baayen RP, ODonnell K, Bonants PJM, Cigeln ik E, Kroon LPNM, Roebroeck EJA, Waalwijk C. 2000. Gene genealogies and AFLP analyses in the complex identify monophyletic and nonmonophyletic causing wilt and rot disease. Phytopathology 90:891-900. Baker CJ, Harrington TC, Krauss U, Alfenas AC. 2003. Genetic variability and host specialization in the La tin American clade of Phytopathology 93:1274. Baker RED, Holliday P. 1957. Witches broom disease of cacao ( Stahel). Phytopathol Papers 2:1-42. Bailey JA, OConnell J, Pring RJ, Nash C. 1992. Infection strategies of species. In: Bailey JA, Jeger MJ, eds. : biology, pathology and control. Wallingford, UK: CAB International. 402 p. Bailey BA, Bae H, Strem MD, Crozier J, Thomas SE, Samuels GJ, Vinyard BT, Holmes KA. 2008. Antibiosis, mycoparasitism, and colonization success for endophytic isolates with biological control potential in Biol Control 46:24-35. , Roberts DP, Thomas SE, Crozier J, Samuels GJ, Choi IY, Holmes KA. 2006. Fungal and plant gene expression during colonization of cacao seedlings by endophytis isolates of four species. Planta 224:1449-1464. Ballard JWO, Rand DM. 2005. The populatio n biology of mitochondrial DNA and its phylogenetic implications. Annu Rev Ecol Evol S 36:621-642. Barroso G, Labarre J. 1997. Genetic evidence for nonrandom sorting of mitochondria in the basidiomycete Appl Environ Microbiol 63:4686-4691. Bartley BGD. 2005. The Genetic Diversity of C acao and its Utilization. Cambridge, MA: CABI Publishing. 341p. Barve MP, Arie T, Salimath SS, Muehlbauer FJ, Peever TL. 2003.Cloning and characterization of the mating type ( ) locus from (teleomorph: ) and a phylogeny of legume-associated spp. Fungal Genet Biol 39:151-167. Bastos CN. 2005. Produo de enzimas extracelulares por Fitopatol Bras 30: 286-288. 210

PAGE 211

, Anderbrhan T. 1986. Urucu ( ): Nova especie hospedera da vassoura de bruxa ( ) do cacueiro. Fitopatol Bras 11: 963-965. Evans HC. 1985. A new pathotype of (witches broom disease) on solanaceous hosts. Plant Pathol 304: 306-312. , Samson RA. 1981. A new hyperparasitic fungus, with potential for control of fungal pathogens of cocoa. T Brit Mycol Soc 77:273-278. Fonseca SEA, Melo WF. 1998. cf. cip nativo da Amaznia Brasileiro, hospedeiro de Fitopatol Bras 23 :504. Beno-Moualem D, Prusky D. 2000. Early events during quiescent infection development by in unripe avocado fru its. Phytopathology 90:553-559. Bergstrom GC, Nicholson RL. 2000. The biology of and maize anthracnose. In: Prusky D, Freeman S, and Dickman MB, eds. Host specificity, pathology and host pathogen interaction of St. Paul, MN: APS Press. 393 p. Bernstein B, Zehr EI, Dean RA, Shabi E. 1995. Characteristics of from peach, apple, pecan, and other hosts. Plant Dis 79:478-482. Beynon SM, Coddington A, Lewis BG, Varzea V. 1995. Genetic variation in the coffee berry disease pathogen, Physiol Mol Plant Path 46:457-470. Bowers JH, Bailey BA, Hebbar PK, Sanogo S, Lumsden RD. 2001. The Impact of Plant Diseases on World Chocolate Production. On line. Plant Health Progress doi:10.1094/PHP2001-0709-01-RV. Brent KJ, Hollomon DW. 2007. Fungicide resist ance in crop pathogens: How can it be managed? Brussels, Belgium: Fungicide React ion Committee, Croplife International. 60 p. Bridge P. 2002. The history and application of molecular mycology. Mycologist 16:90-99. Bridge PD, Spooner BM, Roberts PJ. 2005. The impact of molecular data in fungal systematics. Adv Bot Res 32:33-67. Waller JM, Davies D, Buddie AG. 2008. Variability of in relation to other species from tropical perennial crops and the development of diagnostic techniques. J Phytopathology 156:274-280. Britz H, Coutinho TA, Wingfield MJ, Marasas WFO, Gordon TR, Leslie JF. 1999. f. sp. represents a distinct mating population in the species complex. Appl E nviron Microbiol 65:1198-1201. 211

PAGE 212

Brown JS, Schnell RJ, Motomayor JC, Lopes U, Kuhn DN, Borrone JW 2005. Resistance gene mapping for witches broom disease in L. in an F2 population using SSR markers and candidate genes. J Am Soc Hort Sci 130:366-373. Bruggeman J, Debets AJM, Hoekst ra. RF. 2004. Selection arena in Fungal Genet Biol 41:181-188. Bugante RD, Lizada MCC. 1996. Disease control in Philippine Carabao mango with preharvest bagging and postharvest hot water tr eatment. Acta Hortic 455:797-804. Burdon JJ, Thrall PH. 2008. Pathogen evolutio n across the agro-eco logical interface: implications for disease management. Evol Appl 1:57. Cannon PF, Buddie AG, Bridge PD. 2008. The typification of Mycotaxon 104: 189. Carbone I, Kohn LM. 2001. A microbial populationspecies interface: ne sted cladistic and coalescent inference with multilocus data. Mol Ecol 10: 947-964. Carlucci A, Lops F, and Frisullo S. 2009. Tracheomycosis caused by on subsp. in Italy. J. Plant Pathol. In press. Ceita G de O, Macdo JNA, Sant os TB, Alemanno L, Gramacho KP, Silva D da C, Meinhardt L, Mazzafera P, Pereira GAG, Cascardo JC de M. 2007. Involvement of calcium oxalate degradation during programmed cell death in tissues triggered by the hemibiotrophic fungus Plant Sci 173:106-117. Celio GJ, Padamsee M, Dentinger BTM, Baue r R, McLaghlin DJ. 2006. Assembling the Fungal Tree of Life: constructing the bioche mical database. Mycologia 98:850-859. Ciferri R, Parodi E. 1933. Descrizione del fungo che causa la Moniliasi del cacao. Phytopathol Z 6:539-542. Coe SD, Coe MD. 1996. The True History of Choc olate. New York: Thames and Hudson Inc. 288p. Crozier J, Thomas SE, Aime MC, Evans HC, Holmes KA. 2006. Molecular characterization of fungal e ndophytic morphospecies isolated from stems and pods of Plant Pathol 55:783-791. 212

PAGE 213

Coleman JJ, Rounsley SD, Rodriguez-Carres M, Kuo A, W asmann CC, Grimwood J, Shmutz J, Taga M, White GJ, Zhou S, Schwartz DC, Fr eitag M, Ma L-J, Danc hin EG, Henrissat B, Coutinho PM, Nelson DR, Straney D, Napoli CA, Barker BM, Gribskov M, Rep M, Kroken S, Molnr I, Rensing C, Kennell JC, Zamora J, Farman ML, Selker EU, Salamov A, Shapiro H, Pangilinan J, Lindquist E, Lamers C, Grigoriev IV, Geiser DM, Covert SF, Temporini E, VanEtten HD. 2009. The genome of : contribution of supernumerary chromosomes to gene expansion. PLoS Genetics 5:e1000618 doi:10.1371/journal.pgen.1000618. Cooke DEL, Jung T, Williams NA, Schubert R, Osswald W, Duncan JM. 2005. Genetic diversity of European populations of the oak fine-root pathogen Forest Pathology 35:57-70. Corda ACI. 1837. Icones Fungorum Hucusque Cognitorum 1, 1-39. Nrnberg, Germany. Corkidi G, Balderas-Ruz KA, Taboada B, Serrano-Carren L, Galindo E. 2006. Assessing mango anthracnose using a new three-dimensi onal image-analysis t echniques to quantify lesions on fruit. Plant Pathol 55:250-257. Couch BC, Fudal I, Lebrun MH, Tharreau D, Valent B, van Kim P, Nottghem JL, Kohn LM. 2005. Origins of host-specific popul ation of the blast pathogen in crop domestication with subsequent expansi on of pandemic clones on rice and weeds of rice. Genetics 170:613-630. Cox ML, Irwin AG.1988. Conidium and appressorium variation in Australian isolates of the group and closely related species. Aust Syst Bot 1:139149. Crane JH, Balerdi CF, Maguire I. 2003. Mango growing in the Florida home landscaping. Publication HS2, Florida Cooperative Extension Service, IFAS, University of Florida. Crouch JA, Clarke BB, Hillman BI. 2006. Unraveling evolutionary relationships among the divergent lineages of causing anthracnose diseas e in turfgrass and corn. Phytopathology 96:46-60. , 2009a.What is the value of ITS sequence data in systematics and species diagnosis? A case study using the falcate-spored graminicolous group. Mycologia 101:648-656. Tredway LP, Clarke BB, Hilman BI. 2009b. Phylogenetic and population genetic divergence correspond with habitat for the pathogen and allied taxa across diverse grass communities. Mol Ecol 18:123-135. Darlu P, LeCointre G. 2002. When does incongruen ce length difference test fail? Mol Biol Evol 19:432-437. 213

PAGE 214

Darvas JM. 1993. an important mango pathogen in South Africa. Acta Hort 341:321-328. Davis MJ. 1999. Genetic and pathological dive rsity of the mango anthracnose pathogen in Florida. Proc Fl State Hort Soc 112:197-200. de Almeida AAF, Valle RR. 2007. Ecophysiology of the cacao tree. Braz J Plant Physiol 19:425448. de Arruda MCC, Ferreira MASV, Miller RNG, Resende MLV, Felipe MSS. 2003a. Nuclear and mitochondrial rDNA variability in perniciosa from different geographic origins and hosts. Mycol Res 107:25-37. Miller RNG, Ferreira MASV, Felipe MSS. 2003b. Comparison of isolates from Brazil by ERIC repetiti ve element sequence-based PCR genomic fingerprinting. Plan t Pathol 52:236-244. Sepulveda GF, Miller RNG, Santiago DVR, Resende MLV, Dianese JC, Felipe MSS. 2005. a new species based on mor phological and molecular data. Mycologia 97:1348-1361. de la Bastide PY, Horgen PA. 2003. Mitochondrial inheri tance and the detecti on of non-parentla mitochondrial DNA haplotypes in crosses of homokaryons. Fungal Genet Biol 38:333-342. Delsuc F, Brinkmann H, Philippe H. 2005. Phyloge nomics and the reconstruction of the tree of life. Nat Genet 6:361-375. de Marco J, Felix CR. 2007. Purifi cation and charac terization of a -Glucanase produced by showing biocontrol potential. Braz Arch Biol Techn 50:21-29. de Meyer EM, de Beer ZW, Summerbell RC, Moharram AM, de Hoog GS, Vismer HF, Wingfield MJ. 2008. Taxonomy and phylogeny of new woodand soil-inhabiting species in the complex. Mycologia 100:647-661. Denoyes-Rothan B, Gurin G, Dlye C, Smith B, Minz D, Maymon M, Freeman S. 2003. Genetic Diversity and pathogenic variability among isolates of species from strawberry. Phytopathology 93:219-228. Derso E, Waller JM. 2003. Variation among isolates from diseased coffee berries in Ethiopia. Crop Protection 22:561-565. Desrosiers R and von Buchwald A. 1949. Report of a trip to the Napo river. Annex to the Annual Report 1949, 1-11. Estacin Experimental Agaricola, Pichilingue. Ecudaor. 214

PAGE 215

Dickman M B. 1994. Anthracnose. In: Ploetz RC, Zentmyer GA, Nishijima WT, Rohrbach KG, Ohr HD, eds. Compendium of Tropical Fruit Di seases. St. Paul, MN: APS Press. 118 p. Dinh SQ, Chongwungse J, Pongam P, Sangchote S. 2003. Fruit infection by and anthracnose resistance of some mango cultivars in Thailand. Australas Plant Path 32:533-538. Dodd JC, Bugante R, Koomen I, Jeffries P, Jege r MJ. 1991a. Preand post -harvest control of mango anthracnose in the Philippines. Plant Pathol 40:576-583. Estrada AB, Matcham J, Jeffries P, Jege r MJ. 1991b. The effect of climactic factors on causal agent of mango anthr acnose, in the Philippines. Plant Pathol 40:568-575. Prusky D, Jeffries P. 1997. Fruit Diseas es. In: Litz RE, ed. The mango: botany, production and uses. New York: CAB International. p 257-280. Droby S, Prusky D, Jacoby B, Goldman A. 1986. Presence of antifungal compounds in the peel of mango fruits and their rela tion to latent infections by Physiol Mol Plant Pathol 29:173-183. Droby S, Prusky D, Jacoby B, Goldman A. 1987. I nduction of antifungal resorcinols in mesocarp of unripe mango fruits and its re lation to latent infection by Physiol Mol Plant Pathol 30:285-292. Du M, Schardl CL, Nuckles EM, Vaillancourt LJ. 2005. Using mating-type gene sequenced for improved phylogenetic resolution of species complexes. Mycologia 97:641-658. Duchesne LC, Anderson JB. 1990. Location and direc tion of transcription of the 5S rRNA gene in Mycolog Res 94:266-269. Dyer PS, Hansen J, Delaney A, Lucas JA. 2000. Ge netic control of resistance to the sterol 14 demthylase inhibitor fungicide prochloraz in the cereal eyespot pathogen Appl Env Microbiol 66:4599-4604. Edwards SV. 2009. Natural selection and phyloge netic analysis. Proc Natl Acad Sci 106:87998800. Estrada AB, Jeffries P, Dodd JC. 1996. Field ev aluation of a predictiv e model to control anthracnose disease of mango in the Philippines. Plant Pathol 45:294-301. Dodd JC, Jeffries P. 2000. Effect of humidity and temperature on conidial germination and appressorium development of two Phili ppine isolates of the mango anthracnose pathogen Plant Pathol 49:608-618. 215

PAGE 216

Evans HC. 1978. Witches br oom disease of cocoa ( ) in Ecuador. I. The fungus. Ann Appl Biol 89:193-200. .1980. Pleomorphism in a causal agent of witches broom disease of cocoa. Trans Br Mycol Soc 74:515-523. 1981a. Pod rot of cacao caused by ( ) Phytopath Pap 24:1-44. 1981b. Witches broom disease---a case study. Cocoa Growers Bull 32: 5-19. 2006. Cacao diseases-the trilogy revisited. Phytopathology 97:1640-1643. Holmes KA, Phillips W, Wilkinson MJ. 2002. Whats in a name: the final resting place for the frosty pod rot pat hogen of cocoa? Mycologist 16: 148-152. Krauss U, Rios RR, Secevich TA, Arevalo-Gardini E. 1998. Cocoa in Peru. Cocoa Growers Bull 51:7-22. Faleiro FG, Queiroz VT, Lopes UV, Guimares CT Pires JL, Yamada MM, Arajo IS, Pereira MS, Schnell R, Filho GA de S, Ferreira CF, Barros EG, Moreira MA. 2006. Mapping QTLs for witches broom ( ) resistance in cacao ( L.). Euphytica 149:227-235. Farr DF, Aime MC, Rossman AY, Palm ME. 2006. Species of on Mycolog Res 110:1395-1408. Farris JS, Kllersj M, Kuge AG, Bult C. 1994. Tes ting significance of incongruence. Cladistics 10:315-319. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. 39: 783-791. Ferris PJ, Pavlovic C, Fabry S, Goodenough UW. 1997. Rapid Evolution of sex-related genes in Proc Natl Acad Sci 94:8634-8639. Fitzell RD. 1979. as a cause of anthracnose of mango in New South Wales. Plant Dis Rep 63:1067-1070. Peak CM. 1984. The epidemiology of anthrac nose disease of mango: inoculum sources, spore production and dispersal. Ann Appl Biol 104:53-59. Flaishman MA, Kolattukudy PE. 1994. Timing of fungal invasion using hosts ripening hormone as a signal. Proc Natl Acad Sci 91:6579-6583. 216

PAGE 217

Fournier E, Giraud T, Albertini C, Brygoo Y. 2005. Partition of the complex in France using multiple gene genealogies. Mycologia 97:1251-1267. Forster H, Tyler BM, Coffey MD. 1994. races have arisen by clonal evolution and by rare outcrosses. Mol Plant Microbe In 7:780-791. Francis DM, Gehlen MF, St. Clair DA. 1994.Gen etic variation on homothallic and hyphal swelling isolates of var. and var. Mol Plant Microbe In 7:766-775. Francis DM, St. Clair DA. Outcrossing in the homothallic oomycete, detected with molecular markers. Curr Genet 24:100-106. Freeman S, Katan T, Shabi E. 1998. Characterization of species responsible for anthracnose diseases of vari ous fruit. Plant Dis 82:596-605. Freeman S, MinzD, Maynon M, Zveibi l A. 2001. Genetic Diversity within sensu Simmonds. Phytopathology 91:586-592. , Jurkevitch E, Maymon M, Shab i E. 2000. Molecular analyses of species from almond and othe r fruits. Phytopathology 90:608-614. Shabi E. 1996. Cross-infection of s ubtropical and temperate fruits by species from various hosts. Physiol Mol Plant P 49:395-404. Friesen TL, Stukenbrock EH, Liu Z, Meinhardt S, Ling H, Faris JD, Rasmussen JB, Solomon PS, McDonald BA, Oliver RP. 2006. Emerge nce of a new disease as a result of interspecific virulence gene transfer. Nat Genet 38:953-956. Fulton RH. 1989. The cacao disease trilogy: Blac k pod, monilia pod rot, and witches broom. Plant Dis 73:601-603. Galagan JE, Calvo SE, Cuomo C, Ma L-J, Wortma n JR, Batzoglou S, Lee S-I. Bastrkman M, Spevak CC, Clutterback J, Kapitonov V, Ju rka J, Scazzochio C, Farman M, Butler J, Purcell S, Harris S, Braus GH, Draht O, Busc h S, DEnfert C, Bouchier C, Goldman GH, Bell-Pedersen D, Griffith-Jones S, Doonan JH, Yu J, Vienken K, Pain A, Freitag M, Selker EU, Archer DB, Pealva M, Oakley BR, Mo many M, Tanaka T, Kumagai T, Asai K, Machida M, Nierman WC, Denning DW, Caddick M, Hynes M, Paoletti M, Fischer R, Miller B, Dyer P, Sachs MS, Osma ni SA, Birren BW. 2005. Sequencing of and comparative analysis with and Nature 438:11051115. Gams W. 1995. How natural should a nd anamorph be? Can J Bot 73:S747-S753. Gantotti BV, Davis MJ. 1993. Pectic zymogram anal ysis for characterizing genetic diversity of the mango anthracnose pathogen. Acta Hort 341:353-359. 217

PAGE 218

Garcia O, Macedo JAN, Tibrcio R, Zaparoli G, Rincones J, Bittencou rt LM, Ceita GO, Micheli F, Gesteira A, Mariano AC, Schiavinato MA, Medrano FJ, Meinha rdt LW, Pereira GAG, Cascardo JCM. 2007. characterization of necrosis and ethylene-inducing proteins (NEP) in the basidiomycete the causal agent of witches; broom in Mycol Res 111:443-455. Gardes M, Bruns TD. 1993. ITS primers with e nhanced specificity for basidiomycetes--application to the identification of mycorrhizae and rusts. Mol Ecol 2:113-118. Geiser DM, Frisvad JC, Taylor JW. 1998. Evolutionary relationships in section inferred from partial -tubulin and hydrophobin DNA sequences. Mycologia 90:831-845. Jimnez-Gasco M de M, Kang S, Makalowska I, Veeraraghavan N, Ward TJ, Zhang N, Kuldau GA, ODonnell K. 2004. FUSARIUM-ID v. 1.0: a DNA sequence database for identifying Europ J Plant Pathol 110:473-479. Giblin R. 2005. Avocado fruit responses to PhD dissertation. The University of Queensland, Australia, 216 pp. Gilbert GS and Webb CO. 2007. Phylogenetic signa l in plant pathogen-host range. Proc Natl Acad Sci USA 104:4979-4983. Gonzalez E, Sutton TB. 2004. Population Diversity within isolates of spp. Causing Glomerella leaf spot and bitter rot of apples in three orchards in North Carolina. Plant Dis 88:1335-1340. Gray A. 2001. The World Cocoa Market Outl ook. London: LMC International, Ltd. 29p. Gray MW. 1989. Origin and evolution of mitochondrial DNA. Annu Rev Cell Biol 5:25-50. Griffith GW, Bravo-Velasquez E, Wilson FJ, Lewis DM, Hedger JN. 1994. Autoecology and evolution of the witches broom pathogen ( ) of cocoa. In: Blakeman JP, Williamson B, eds. Ecology of Plant Pat hogens. Cambridge, MA: CAB International. p. 245-267. Hedger JN. 1993. A novel method for producing basidiocarps of the cocoa pathogen using a bran-vermiculite medium. Eur J Plant Pathol 99:227-230. 1994a. The breeding biology of biotypes of the witches broom pathogen of cocoa, Heredity 72: 278-289. 1994b. Spatial distribution of mycelia of the liana (L-) biotype of the agaric (Stahel) Singer in tropical forest. New Phytol 127: 243-259. 218

PAGE 219

Guerber JC, Correll JC. 2001. Characterization of the teleomorph of Mycologia 93:216-229. Liu B, Correll JC, Johnston PR. 2003. Characterization of diversity in by sequence analysis of two gene introns, mtDNA and intron RFLPs, and mating compatibility. Mycologia 95:872-895. Guetsky R, Kobiler I, Wang X, Perlman N, Gollop N, Avila-Quezada G, Hadar I, Prusky D. 2005. Metabolism of the flavonoid epicatechin by laccase of and its effect on pathogenicity on avocado fruits. Phytopathology 95:1341-1348. Gimil S, Chang HS, Zhu T, Sesma A, Osbourn A, Roux C, Ioannidis V, Oakeley EJ, Docquier M, Descombes P, Briggs SP, Paszkowski U. 2005. Comparative transcriptomics of rice reveals an ancient pattern of response to mi crobial colonization. P Natl Acad Sci USA 102: 8066. Hammerschmidt R, Nicholson RL. 1977. Resistance of maize to anthracnose: changes in host phenols and pigments. Phytopathology 67:251-258. Harris SA. 1999. RAPDs in systematics-a useful methodology? In: Hoolingsworth PM, Bateman RM, Gornall RJ, eds. Molecular systematic s and plant evolution. London: CRC Press. p. 211-228. Hassan MK, Dann EK, Irving DE, Coates LM 2007. Concentrations of constitutive alk(en)ylresorcinols in peel of commercial mango varietie s and resistance to postharvest anthracnose. Physiol Mo l Plant Pathol 71:158-165. Hawksworth DL. 1991. The fungal dimensions of biodiversity: magnitude, significance, and conservation. Mycol Res 95:641-655. 2001. The magnitude of fungal diversity: the 1.5 million species estimate revisted. Mycol Res 105:1422-1432. Hayden HL, Pegg KG, Aitken EAB, Irwin JAG. 1994. Genetic relationships as assessed by molecular markers and crossinfection among strains of Aust J Bot 42:9-18. Hedger JN, Pickering V, Arangundi J. 1987. Variab ility of population of the witches broom disease of cocoa ( ). Trans Brit Mycol Soc 88:553-545. Hibbett DS. 2006. A phylogenetic overview of th e Agaricomycotina. Mycologia 98:917-925. Hillis DM, Dixon MT. 1991. Molecular evolution and phylogenetic inference. Q Rev Biol 66:411-453. 219

PAGE 220

Hodson A, Mills PR, Brown AE. 1993. Ribosom al and mitochondrial DNA poly morphisms in isolated from tropical fr uits. Mycol Res 97:329-335. Hong SK, Kim WG, Yun HK, Choi KJ. 2008. Morphological Variati ons, genetic diversity and pathogenicity of species causing grape ripe rot in Korea. Plant Pathology J 24:269-278. Hu G, St. Leger RJ. 2004. A phylogenomic approach to reconstructing th e diversification of serine proteases in fungi. J Evol Biol 17:1204-1214. Iwen PC, Hinrichs SH, Rupp ME. 2002. Utilization of the internal transcribed spacer regions as molecular targets to detect and identify human fungal pathogens. Med Mycol 40:87. James C. 1971. A manual of a ssessment keys for plant diseases. Canada Department of Agriculture. St. Paul, MN: APS Press. 90p. James TY, Kauff F, Schoch C, Matheny PB, Hofstette r V, Cox C, Celio G, Gueidan C, Fraker E, Miadlikowska J, Lumbsch HT, Rauhut A, R eeb V, Arnold AE, Amtoft A, Stajich JE, Hosaka K, Sung G-H, Johnson D, ORourke B, Crockett M, Binder M, Curtis JM, Slot JC, Wang Z, Wilson AW, Schler A, Longcore JE, ODonnell K, Mozley-Standridge S, Porter D, Letcher PM, Powell MJ, Taylor JW, White MM, Griffith GW, Davies DR, Humber RA, Morton JB, Sugiyama J, Rossma n AY, Rogers JD, Pfister DH, Hewitt D, Hansen K, Hambleton S, Shoemaker RA, Kohlmeyer J, Volkmann-Kohlmeyer B, Spotts RA, Serdani M, Crous PW, Hughes KW, Mats uura K, Langer E, Langer G, Untereiner WA, Lcking R, Bdel B, Geiser DM, Aptroot A, Diederich P, Schmitt I, Schultz M, Yahr R, Hibbett D, Lutzoni F, McLaughlin D, Spat afora J, Vilgalys R. 2006a. Reconstructing the early evolution of the fungi using a six ge ne phylogeny. Nature 443:818-822. Letcher PM, Longcore JE, Mozley-Standridge SE, Porter D, Powell MJ, Griffith GW, Vilgalys R. 2006b. A molecular phylogeny of th e flagellated fungi (Chytridiomycota) and description of a new phylum (Blast ocladiomycota). Mycologia 98:860-871. Johnson LA, Soltis DE. 1998. Assessing congruence: empirical examples from molecular data. In: Soltis DE, Soltis PS, Doyle JJ, eds. Molecular Systematics of Plants II: DNA Sequencing. Boston, MA: Kluwer Academic Publishers. 574 p. Johnson LJ, Koufpanou V, Goddard MR, Hetheri ngton R, Schafer RM, Burt A. 2004. Population genetics of the wild yeast Genetics 166:43-52. Junior P. 2006. Terrorism o Biolgico. Veja 1961:60-63. Kerekes JF, Desjardin DE. 2009. A monograph of the genera and from Southeast Asia including a molecu lar phylogeny. Fungal Diversity 37:101-152. 220

PAGE 221

Kilaru A, Bailey BA, Hasenstein KH. 2007. produces hormones and alters endogenous auxin and salicylic acid in fected cocoa leaves. FEMS Microbiol Lett 274:238-244. Knight, Jr. RJ. 1993. Evaluating important fruit characters in m ango germplasm. Fruit Varieties J 47:25-30. Schnell RJ. 1994. Mango introduction in Florida and the Haden cul tivars significance to the modern industry. Econ Bot 48:139-145. Kndler M, Berardini N, Kammerer DR, Carle R, Schieber A. 2007. Characterization of major and minor alk(en)ylresorcinols from mango ( L.) peels by highperformances liquid chromatography/atmospheric pressure chemical ionization mass spectrometry. Rapd Commun Mass Spectrom 21:945-951. Kobayashi T, Okamoto T. 2003. Notes on some pl ant inhibiting fungi collected at Hahajima, Bonin Islands. J Agric Sci 48:89-104. Kohn LM. 2005. Mechanisms of fungal sp eciation. Annu Rev Phytopath 43:279-308. Koomen I, Jeffries P. 1993. Effects of antagonistic microorganisms on the post-harvest development of on mango. Plant Pathol 42:230-237. Kraft JM. 1977. The role of delphinidin and sugars in the resistance of pea seedlings to Fusarium root rot. Phytopathology 67:1057-1061. Kramer-Haimovich H, Servi E, Katan T, Rollin s J, Okon Y, Prusky D. 2006. Effect of ammonia production by on activation, pectate lyase secretion, and fruit pathogenicity. A ppl Env Microbiol 72:1034-1039. Krauss U, Soberanis W. 2002. Effect of fertiliz ation and biocontrol a pplication frequency on cocoa pod diseases. Biol Control 24:82-89. Kronstadt JW, Staben C. 1997. Mating type in filamentous fungi. Annu Rev Genet 31:245-76 Kummuang N, Smith BJ, Diehl SV, Graves, Jr. CH 1984. Muscadine grape be rry rot diseases in Mississippi: disease iden tification and inciden ce. Plant Dis 80:238-243. Kuo KC. 2001. Sensitivity of mango anthracnose pathogen, to the fungicide prochloraz in Taiwan. Proc Natl Sci. Counc ROC (B) 25:174-179. Laker HA, Ram A. 1992. Investigations on integrated control of witches broom disease of cocoa in Rondonia State, Brazil. Trop Pest Manag 38:354-358. Leslie JF. 1991. Mating populations in ( section ). Phytopathology 81:1058-1060. 221

PAGE 222

. 1995. : available populations and vari able traits. Can J Bot 73:S282S291. Linzer RE, Otrosina WJ, Gonthi er P, Bruhn J, Laflamme G, Bu ssires G, Garbelotto M. 2008. Inferences on the phylogeography of the fungal pathogen, including evidence of interspecific horizontal gene transfer and of human-mediated, longrange, dispersal. Mol Ph ylogenet Evol 46:844-862. Lonsdale JH, Kotz JM. 1993. Chemical control of mango blossom diseases and the effect on fruit set and yield. Plant Dis 77:558-562. Lubbe CM, Denman S, Cannon PF, Groenewa ld JZ, Lamprecht SC, Crous PW. 2004. Characterization of species associated with diseases of Proteaceae. Mycologia 96:1268-1279. Lutzoni F, Kauff F, Cox CJ, Mc Laughlin D, Celio G, Dentinger B, Padamsee M, Hibbett D, James TY, Baloch E, Gruber M, Reeb V, Hofstetter V, Schoch C, Arnold C, Arnold AE, Miadlikowska J, Spatafora J, Johnson D, Hamb leton SM, Crockett M, Shoemaker R, Sung G, Lcking R, Lumbsch T, ODonnell K, Binde r M, Diederich P, Ertz D, Gueidan C, Hansen K, Harris RC, Hosaka K, LimY, Matheny B, Nishida H, Pfister D, Rogers J, Rossman A, Schmitt I, Sipman H, Stone J, Sugiyama J, Yahr R, Vilgalys R. 2004. Assembling the Fungal Tree of Life: progress, classification and evolution of subcellular traits. Am J Bot 91:1446-1480. Macagnan D, Romeiro RS, Pomella AWV, de S ouza JT. 2008. Prodcution of lytic enzymes and siderophores, and inhibition of ge rmination of basidiospores of ( ) by phylloplane actinomycetes. Biol Control 47:309-314. MacKenzie SJ, Mertely JC, Seijo TE, Peres NA. 2008. is a pathogen on hosts other than strawberry. Plant Dis 92:1432-1438. Peres NA, Barquero MP, Arauz LF, Timmer LW. 2009. Host range and genetic relatedness of isolates from fruit crops and leatherleaf fern in South Florida. Phytopathology 99:620-631. Seijo TE, Legard DE, Timmer LW, Peres NA. 2007. Selection for pathogenicity to strawberry in population of from native plants. Phytopathology 97:1130-1140. Romeiro RS, de Souza, JT, Pomella AWV. 2006. Isolation of actinomycetes and endospore-forming bacteria from the cacao pod surface and their antagonistic activity against witches broom and black po d pathogens. Phytoparasitica 34:122-132. Mahuku GS, Riascos JJ. 2004. Virulence and molecular diversity within isolates from Andean and Mesoameri can bean varieties and regions. Eur J Plant Pathol 110:253-263. 222

PAGE 223

Manicom B, Ruggiero C, Ploetz RC, and de Goes A. 2003. Diseases of passion fruit. In: Ploetz RC, ed. Diseases of Tropical Fr uit Crops. Cambridge, MA: CABI Publishing. 527 p. Marita JM, Nienhuis J, Pires JL, Aitken WM. 2001. Analysis of Genetic Diversity in with emphasis on witches broom disease resistance. Crop Sci 41:1305-1316. Martinez-Culebras PV, Barrio E, Suarez-Fern andez MB, Garcia-Lopez MD, Querol A. 2002. RAPD analysis of species isolated from stra wberry and the design of specific primers for the identification of J Phytopathology 150:680-686. Querol A, Suarez-Fernandez MB, Garc ia-Lopez MD, Barrio E. 2003. Phylogenetic relationships among pathogens of strawberry and design of PCR primers for their identification. J Phytopath 151:135-143. Matheny PB, Curtis JM, Hofstetter V, Aime MC, Moncalvo JM, Ge ZW, Yang ZL, Slot JC, Ammirati JF, Baroni TJ, Bougher NL, Hughe s KW, Lodge DJ, Kerrigan RW, Seidl MT, Aanen DK, DeNitis M, Daniel GM, Desjar din DE, Kropp BR, Norvell LL, Parker A, Vellinga EC, Vilgalys R. 2006. Major clades of Agaricales: a multilocus phylogenetic overview. Mycologia 98:982-995. Matheny PB, Wang Z, Binder M, Curtis JM, Lim YW, Nilsson RH, Hughes KW, Hofstetter V, Ammirati JF, Schoch CL, Langer E, Langer G, McLaughlin DJ, Wilson AW, Frslev T, Ge ZW, Kerrigan RW, Slot JC, Yang ZL, Baroni TJ, Fischer M, Hosaka K, Matsuura K, Seidl MT, Vauras J, Hibbett DS. 2007. Contribu tions of rpb2 and tef1 to the phylogeny of mushrooms and allies (Basidiomycota, Fungi). Mol Phylogenet Evol 43:430-451. Wang Z, Binder M, Curtis JM, Lim YW Nilsson RH, Hughes KW, Hofstetter V, Ammirati JF, Schoch CL, Langer E, Langer G, McLaughlin DJ, Wilson AW, Frslev T, Ge ZW, Kerrigan RW, Slot JC, Yang ZL, Baroni TJ, Fischer M, Hosaka K, Matsuura K, Seidl MT, Vauras J, Hibbett DS. 2007. Contribu tions of rpb2 and tef1 to the phylogeny of mushrooms and allies (Basidiomycota, F ungi). Molec Phylogenet Evol 43:430-451. May G, Taylor JW. 1988. Patterns of mating and mitocondrial dNA inheritance in the agaric basidiomycete Genetics 118:213-220. Mayr E. 2000. Biological species concept. In: Wheeler QD, Meier R, eds. Species Concepts and phylogenetic theory: a debate. New York: Columbia University Press. p. 17-29. Mbofung GY, Hong SG, Pryor BM. 2007. Phylogeny of f. sp. lactucae inferred from mitocondrial small subunit, elong ation factor 1-alpha, and nuclear ribosomal intergenic spacer sequence data. Phytopathology 97:87-98. McDermott JM, Brandle U, Dutly F, Haemerli UA, Keller S, Muller KE, Wolf MS. 1994. Genetic variation in powdery mildew of barley: development of RAPD, SCAR, and VNTR markers. Phytopathology 84:1316-1321. 223

PAGE 224

McGuire RG, Ca mpbell CA. 1993. Imazalil for posth arvest control of anthracnose on mango fruits. Acta Hortic 341:371-376. McLaughlin DJ, Berres ME, Szabo LJ. 1995. Molecules and morphology in basidiomycete phylogeny. Can J Bot 73:S684-S692. McMillan RT. 1984. Control of mango anthracnose w ith foliar sprays. Proc Fla State Hort Soc 97:344-345.Meinhardt LW, Bailey BA, Guiltinan M, Carazzolle MF, Lacerda G, Mondego J, Formigheri E, Periera GA. 2008. Comparative genome analyses of and : two closely related phytopathogenic basidiomycetes that cause distinctly different diseases of theobroma cacao. 2008 Mid Atlantic Plant Molecular Biology Society Conf erence. Savage Mill, MD. Rincones J, Bailey BA, Aime MC, Griffith GW, Zhang D, Pereira GAG. 2008. the causal agent of witches broom disease of cacao: whats new from this old foe? Mol Plant Pathol 9:577-588. Meja LC, Rojas EI, Maynard Z, Van Bael S, Arnold AE, Hebbar P, Samuels GJ, Robbins N, Herre EA. 2008. Biol Control 46:4-14. Merino ST, Nelson MA, Jacobson DJ, Natvig DO. 1996. Pseudohomothallism and evolution of the mating-type chromosome in Genetics 143:789-799. Mickevich MF, Farris JS. 1981. The im plications of congruence in Syst Zool 30:351370. Miechelse CB, Rep M. 2009. Pahogen profile update: Mol Plant Pathol 10:311-324. Milgroom MG. 1996. Recombination and the mu ltilocus structure of fungal populations. Ann Rev Phytopath 34:457-477. Mills PR, Sreenivasaprasad S, Brown AE. 1992. Detection and differentiation of isolates using PCR. FEMS Microbiol Lett 98:137-143. Miyara I, Shafran H, Kramer Haimovich H, Rollins J, Sherman A, Prusky D. 2008. Multi-factor regulation of pectate lyase secretion by pathogenic on avocado fruits. Mol Plant Pathol 9:281-291. Moore D. 1998. Fungal Morphogenesis. Cambridge, UK: Cambridge University Press. 488 p. Moriwaki J, Sato T, Tsukiboshi T. 2003. Morp hological and molecula r characterization of sp. nov. from Japan. Mycoscience 44:47-53. Moriwaki J, Tsukiboshi T, Sato T. 2002. Grouping of species in Japan based on rDNA sequences. J Gen Pl ant Pathol 68:307-320. 224

PAGE 225

Mossler MA, Nesheim ON. 2002. Florida Crop/P est Manageme nt Profile: Mango. Publication CIR 1401, Florida Cooperative Extension Serv ice, IFAS, University of Florida. Muirhead IF. 1976. Post-harvest c ontrol of mango anthracnose with benomyl and hot water. Aust J Exp Agr Animal Husbandry 16:600-603. Munaut F, Hamaide N, Maraite H. 2002. Genomic and pathogenic diversity in from wild native Mexican spp., and taxonomic implications. Mycol Res 106:579-593. Niella GR, Resende ML, Castro HA, Silva LHCP, Carvalho JA. 1999. Aperfeioamento da metodologia de produo artifi cial de basidiocarpos de Fitopatol Bras 24:523-527. Nilsson RH, Kristiansson E, Ryberg M, Halle nberg N, Larsson KH. 2008. Intraspecific ITS variablility in the Kingdom Fungi as expresse d in the international sequence databases and its implications for molecular species identification. Evol Bioinform 4:193-201. Nirenberg HI, Feiler U, Hagedorn G. 2002. Descriptiono f comb. nov. in modern terms. Mycologia 94:307-320. ODonnell K, Cigelnik E, Nirenberg HI. 1998a. Molecular systematic and phylogeography of the species complex. Mycologia 90:465-493. Kistler HC, Cigelnik E, Ploetz R. 1998b. Multiple evolutionary origins of the fungus causing Panama disease of banana: concordant evidence from nuclear and mitochondrial gene genealogies. Proc Natl Acad Sci USA 95:2044-2049. Lutzoni FM, Ward TJ, Benny GL. 2001. Evol utionary Relationships among Mucoralean Fungi (Zygomycota): Evidence for Family Po lyphyly on a Large Scale. Mycologia 93:286297. Nirenberg HI, Aoki T, Cigelnik E. 2000. A multigene phylogeny of the species complex: Detection of additional phylogenetically distinct species. Mycoscience 41:61-78. Ward TJ, Geiser DM, Kistler HC, Aoki T. 2004. Geneological concordance between the mating type locus and seven other nuclear genes supports formal recognition of nine phylogenetically distinct species with the clade. Fungal Genet Biol 41:600-623. Oerke EC, Dehne HW. 2004. Safeguarding producti onlosses in major crops and the role of crop protection. Crop Prot 23:275. 225

PAGE 226

Orchard J, Collin HA, Hardwick K, Isaac S. 1994 Changes in morphology and measurement of cytokinin levels during the development of witches brooms on cacao. Plant Pathol 43:6572. Paim VRL de M, Luz EDMN, Pires JL, Silva SDVM, de Souza JT, Albuquerque PSB, Filho LP dos S. 2006. Sources of resistance to in progenies of cacao accession collected in the Brazilian Amazon. Sc i Agric (Piracicaba, Braz.) 63:572-578. Paquin B, Laforest MJ, Forget L, Roewer I, Wang Z, Longcore J, Lang BF. 1997. The fungal mitochondrial genome project: evolution of f ungal mitochondrial genomes and their gene expression. Curr Genet 31:380-395. Patio-Vera M, Jimnez B, Balderas K, Ortiz M, Allende R, Carrillo A, Galindo E. 2005. Pilotscale production and liquid formulation of a potential biocontrol agent of mango anthracnose. J Appl Microbiol 99:540-550. Peever TL, Carpenter-Boggs L, Timmer LW, Carri s LM, Bhatia A. 2005. Citrus black rot is caused by phylogenetically distinct lineages of Phytopathology 95:512-517. Pegler DN. 1978. (Agaricales). Kew Bullet 32: 731-736. Pegus JE. 1972. Aspects of the hostparasite relationship in the L./ Stahel disease complex. M.Sc. Thesis, Univ ersity of the West Indies, Trinidad. Pereira JL, de Almeida LCC, Santos SM. 2006. Witches broom disease of cocoa in Bahia: attempts at eradication and c ontainment. Crop Prot 15:743-752. Barreto RW, Bezerra JL. 1997. Distribuio de em solanceas selvagens no sudeste do Brasil. Fitopatol Bras 22S: 295. Peres NAR, Kuramae EE, Dias MSC, and de Souza NL. 2002. Identification of spp. affecting fruit after harvest in Brazil. J Phytopathol 150:128-134. MacKenzie SJ, Peever TL, Timmer LW. 2008. Postbloom fruit drop of citrus and key lime anthracnose are caused by di stinct phylogenetic lineages of Phytopathology 98:345-352. Timmer LW, Adaskaveg JE, Correll JC. 2005. Lifestyles of Plant Dis 89:784-796. Persley DM, Ploetz RC. 2003. Diseases of Papaya. In: Ploetz RC, ed. Diseases of Tropical Fruit Crops. Cambridge, MA: CABI Publishing. 527 p. 226

PAGE 227

Phillips-Mora W, Aim e MC, W ilkinson MJ. 2007. Biodiversity and biogeography of the cacao ( ) pathogen in tropical America. Plant Pathol 56:911-922. Cawich J, Aime MC. 2006a. First report of frosty pod rot (moniliasis disease) caused by on cacao in Belize. Plant Pathol 55:584. Coutio A, Ortiz CF, Lpez AP, Hernndez J, Aime MC. 2006b. First report of frosty pod rot (moniliasis disease) caused by on cacao in Mexico. Plant Pathol 55:584. Photita W, Taylor PWJ, Ford R, Hyde KD, Lumyong S. 2005. Morphological and molecular characterization of species from herbaceous plants in Thailand. Fungal Diversity 18:117-113. Ploetz RC, Prez-Martnez JM, Palmateer AJ, Tarnowski TL. 2009. Influence of temperature, light intensity, and isolate on the development of -induced dieback of eugenia, Plant Dis 93:804-808. Ploetz RC, Prakash O. 1997. Foliar, floral and so ilborne diseases. In: Litz RE, ed. The mango: botany, production and uses. New York: CAB International. p. 281-326. Poggeler S. 1999. Phylogenetic relationships betw een mating-type sequences from homothallic and heterothallic ascomycet es. Curr Genet 36:222-231. Pound FJ. 1943. Cacao and witches broom disease ( ). A report on a recent visit to the Amazon region of Peru, September 1942-February 1943. Department of Agriculture of Trinidad and Tobago. Yuille s Priaterie, St. Augustine, Trinidad. Prillinger H, Lopandic K, Schwei gkofler W, Deak R, Aarts HJM, Bauer R, Sterflinger K, Kraus GF, Maraz A. 2002. Phylogeny and sy stematics of the fungi with special reference to the Ascomycota and Basidiomycota. In: Breitenb ach M, Crameri R, Lehrer SB, eds. Fungal allergy and pathogenicity. Chem Immunol vol 81. 207. 310 p. Promputtha I, Saisamorn L, Dhanasekaran V, McKenzie E, Hyde K, Jeewon R. 2007. A phylogenetic evaluation of whether endophytes become saprophytes at host senescence. Microbial Ecol 53:579-590. Prusky D, Keen NT. 1993. Involvement of preforme d antifungal compounds in the resistance of subtropical fruits to fungal decay. Plant Dis 72:114-119. Kobiler I, Miyara I, Alkan N. 2009. Fruit Diseases. In: Litz RE, ed. The mango: botany, production and uses. Cambridge, MA: CAB International. 680 p. Lichter A. 2007. Activation of quiescent infections by postharvest pathogens during transition from the biotrophic to the necrot rophic stage. FEMS Mi crobiol Lett 268:1-8. 227

PAGE 228

, 2008. Mechanisms modulating fungal attack in post-harvest pathogen interactions and their control. Eur J Plant Pathol 121:281-289. McEvoy JL, Leverentz B, Conway WS 2001. Local modulation of host pH by species as a mechanism to increase virulence. Mol Plant Microbe In 14:1105-1113. Purdy LH, Dickstein ER. 1990. Basidiocar p development on mycelia mats of Plant Dis 74:493-496. Queiroz VT, Guimares CT, Anhert D, Schuste r I, Daher RT, Pereira MG, Miranda VRM, Loguercio LL, Barros EG, Moreira MA. 2003. Id entification of a major QTL in cocoa ( L.) associated with resistance to witches broom disease. Plant Breeding 122:268-272. Rahman MA, Mahmud TMM, Kadir J, Abdul Ra hman R, Begum MM. 2008. Major postharvest fungal diseases of papaya cv. Sekaki in Selangor, Malaysia. Per tinaka J Trop Agric Sci 31:27-34. Redecker F, Raab P. 2006. Phylogeny of the Gl omeromycota (arbuscular mycorrhizal fungi): recent developments and new gene markers. Mycologia 98:885-895. Reeb V, Lutzoni F, Roux C. 2004. Contribution of RPB2 to multilocus phylogenetic studies of the euascomycetes (Pezizomycotina, Fungi) w ith special emphasis on the lichen-forming Acarosporaceae and evolution of polyspory. Molec Phylogenet Evol 32:1036-1060. Resende MLV, Gutemberg BAN, Silva LHCP, Niella GR, Carvalho GA, Santiago DVR, Bezerra JL. 2000. proveniente de um novo hospedeiro, patognico ao cacauei ro. Fitopat Bras 25:88-91. Rincones J, Mazotti GD, Griffith GW, Pomela A, Figueira A, Leal GA, Queiroz MV, Pereira JF, Azevedo RA, Pereira GAG, Meinhardt LW. 2006. Genetic variability and chromosomelength polymorphisms of th e witches broom pathogen from various plant hosts in South America. Mycol Res 110:821-832. Riveras-Vargas LI, Lugo-Noel Y, McGovern RJ, Seijo T, David MJ. 2006. Occurrence and distribution of spp. on mango ( L.) in Puerto Rico and Florida, USA. Plant Pathol J 5:191-198. Roets F, Wingfield MJ, Crous PW, Dreyer LL. 2009. Fungal radiation in the Cape Floristic Region: an analysis based on and Mol Phylogenet Evol 51:111-119. Saha T, Kumar A, Ravindran M, Jacob C k, Roy B, Nazeer MA. 2002. Identification of from rubber using random amplified polymorphic DNAs and ribosomal DNA polymorphisms. Mycol Res 106:215-224. 228

PAGE 229

Salazar O, Julian MC, Hyakumuchi M, Rubio V. 2000. Phylogenetic Grouping of Cultural Types of Rhizoctonia solani AG 2-2 Based on Ri bosomal ITS Sequences. Mycologia 92:505-509. Sanders GM, Korsten L. 2003. Comparison of cro ss inoculation potential of South African avocado and mango isolates of Microbiol Res 158:143150. , Wehner FC. 2000. Survey of fungicide sensitivity in from different avocado and mango pr oduction areas in South Africa. Eur J Plant Pathol 106:745-752. Sanogo S, Pomella A, Hebbar PK, Bailey B, Costa JCB, Samuels GJ, Lumsden RD. 2002. Production and germination of conidia of a mycoparasite of on Cacao. Phytopathology 92:1032-1037. Santos RMF, Lopes UV, Bahia R de C, Machado RCR, Ahnert D, Co rra RX. 2007. Marcadores microssatlites relacionados com a resistn cia vassoura-de-bruxa do cacaueiro. Pesq Agropec Bras, Braslia 42:1137-1142. Santos RX, Melo SCO, Cascardo JCM, Br endel M, Pungartnik C. 2008. Carbon sourcedependent variation of acqui red mutagen resistance of : similarities in natural and artificial systems. Fungal Genet Biol 45:851-860. Schnell, RJ, Brown JS, Olano CT, Meerow AW, Campbell RJ, Kuhn DN. 2006. Mango genetic diversity analysis and pedigr ee inferences for Florida cultivars using microsatellite markers. J Am Soc Hort Sci 131:214-224. Schoeman MH, Manicom BQ, Wingfield MJ. 1995. Epidemiology of powdery mildew on mango blossoms. Plant Dis 79:524-528. Schulz B, Rmmert A-K, Dammann U, Aust H-J, Strack D. 1999. The endophyte-host interaction: a balanced an tagonism? Mycol Res 103:1275-1283. Schssler A, Schwarzott D, Walker C. 2001. A new fungal phylum, the Glomeromycota: phylogeny and evolution. Mycol Res 105:1413-1421. Seelanen T, Schnabel A, Wendel JF. 2007. Congr uence and consensus in the cotton tribe (Malvaceae). Syst Bot 22:259-290. Seifert KA, Samson RA, deWard JR, Houbraken J, Lvesque J-M, Louis-Se ize G, Hebert PDN. 2007. Prospects for fungus identification using DNA barcodes, with as a test case. Proc Natl Ac ad Sci USA 104:3901-3906. Sesma A, Osbourn AE. 2004. The rice leaf blast pathogen undergoes developmental processes typical of root-infecti ng fungi. Nature 431:582-586. 229

PAGE 230

Shi Y, Correll JC, Guerber JC, Rom CR. 1996. Frequency of species causing bitter rot of apple in the southeastern United St ates. Plant Dis 80:692-696. Simmonds JH. 1965. A study of the species of causing fruit rot in Queensland. Qld J Agric Anim Sci 22:437-459. Simmons MP, Ochoterena H. 2000. Gaps as characters in sequence-based phylogenetic analyses. Syst Biol 49:369-381. Singer R. 1942. A monographic study of the genera Crinipellis and Chaetocalathus. Lilloa 8:441-531. 1976. Monograph No. 17. Marasmiaceae (Bas idiomycetes-Tricholomataceae). Flora Neotropica 17: 1. Smilanick JL, Schmidt L, Ghosph J, Margosan DA. 2003. Mutation at be ta-tubulin gene codon 200 of conferred thiabendazole resi stance. Plant Dis 90: 765-770. Smith BJ, Black LL. 1990. Morphological, cultural and pathogenic variation among species isolated from st rawberry. Plant Dis 74:69-76. Soberanis W, Rios R, Arvalo E, Ziga L, Cabezas O, Krauss U. 1999. Increased frequency of phytosanitary pod removal in cacao ( ) increases yield economically in eastern Peru. Crop Prot 18:677-685. Spatafora JW. 1995. Ascomal evolution of filamentous ascomycetes: evidence from molecular data. Can J Bot 73:S811-S815. Sung G, Johnson D, Hesse C, ORourke B, Seda ni M, Spotts R, Lutzoni F, Hofstetter V, Miadlikowska J, Reeb V, Gueidan C, Fraker E, Lumbsch T, Lcking R, Schmitt I, Hosaka K, Aptroot A, Roux C, Miller AN, Geiser DM, Hafellner J, Hest mark G, Arnold AE, Bdel B, Rauhut A, Hewitt D, Untereiner WA, Cole MS, Scheidegger C, Schultz M, Sipman H, Schoch CL. 2006. A five-gene phylogeny of Pezizomycotina. Mycologia 96:1018-1028. Spiers AG, Brewster DT, Slade A, Gardiner SE 2000. Characterization of New Zealand isolates of with regard to morphology, growth, pathogenicity and RAPD banding patterns Mycol Res 104:395-402. Sreenivasaprasad S, Brown AE, Mills PR. 1993. Coffee berry disease pathogen in Africa: genetic structure and relati onship to the group species Mycol Res 97:995-1000. Mills PR, Meehan BM, Brown, AE. 1996. Phylogeny and systematic of 18 species based on ribosomal DNA spacer sequences. Genome 39:499-512. 230

PAGE 231

, Sharada K, Brown AE, Mills PR. 1996. PCR-based detection of on strawberry. Pl ant Pathol 45:650-655. Talhinas P. 2005. Genotypic and phenotypic diversity in a cosmopolitan pathogen causing anthracnose on a wide range of hosts. Mol Plant Pathol 6:361-378, Stahel G. 1915. nov. spec. Dept Landbouw in Surname Bull 33:1-27. Steenkamp ET, Wingfield BD, Coutinho TA, Zelle r KA, Wingfield MJ, Marasas WFO, Leslie JF. 2000. PCR-based id entification of and in the species complex. Appl E nviron Microbiol 66:4378-4382. , Desjardins AE, Marasas WFO, Wi ngfield MJ. 2002. Cryptic speciation in Mycologia 94:1032-1043. Stuckenbrock EH, McDonald BA, Grnlund-Jrg ensen F, Heide MS. 2009. Deciphering the genetic basis of speciation of a fungal pl ant pathogen through comparative genomics. 2009 Mycological Society Annual Meeting, Snowbird, Utah. Sukno SA, Garca VM, Shaw BD, Thon MR. 2008. R oot infection and systematic colonization of maize by Appl Environ Microb 74:823-832. Sugiyama J, Hosaka K, Suh SO. 2006. Early-diverging Ascomy cota: phylogenetic divergence and related evolutionary enigmas. Mycologia 98:996-1005. Suh SO, Blackwell M, Kurtzman CP, Lachance MA. 2006. Phylogenetics of Saccharomycetales, the ascomycete yeasts. Mycologia 98:1006-1017. Sundravadana S, Alice D, Kuttalam S, Samiyappan R. 2006. Control of mango anthracnose by azoxystrobin. Tunis J Plant Prot 1:109-114. Sutton BC. 1980. The Coelomycetes. Kew, London: Commonwealth Mycological Institute. 696 p. 1992. The genus and its anamorph In: Bailey JA, Jeger MJ, eds. : biology, pathology and control. Wallingford, UK: CAB International. 402 p. Swofford DL. 2000. PAUP*. Phylogenetic Analys is Using Parsimony (*and Other Methods). Version 4. Sunderland, MA: Sinauer Associates. Talhinhas P, Sreenivasaprasad S, Neves-Martin s J, Oliveira H. 2005. Molecular and phenotypic analyses reveal association of diverse groups and a low level of with olive anthracnose. A ppl Environ Microbiol 71:2987-2998. 231

PAGE 232

Tamura K, Dudley J, Nei M, Kuma r S. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software vers ion 4.0. Mol Biol Evol 24:1596-1599. Tanabe Y, Saikawa M, Watanabe MM, Sugiya ma J. 2004. Molecular phylogeny of Zygomycota based on EF-1 and RPB1 sequences: limitations and utility of alternative markers to rDNA. Mol Phylogenet Evol 30:438-449. Tapia-Tussell R, Quijano-Ramayo A, Cortes-Vel asquez A, Lappe P, Larque-Saavedra A, PerezBrito D. 2008. PCR-based detection and ch aracterization of the fungal pathogens and causing anthracnose in papaya ( L.) in the Yucatan peninsula. Mol Biotechnol 40:293-298. Tarnowski TL, Ploetz RC. 2008. Mango anthracnose in south Florida: Assessing the respective roles of and Phytopathology 98: S155. Taylor JW. 1995. Making the Deuteromycota redunda nt: a practical integr ation of mitosporic and meiosporic fungi. Can J Bot 73:S754-S759. Jacobson DJ, Kroken S, Kasuga T, Geiser DM, Hibbett DS, Fisher MC. 2000. Phylogenetic species recognition and species co ncepts in fungi. Fungal Genet Biol 31:2132. Taylor M. 1994. The World Cocoa Situation. International Forum in Cocoa, Lima, Peru, October 28-29, 1998. LMC International, Ltd., London, UK. Templeton AR. 1983. Phylogenetic in ferences from restriction endonuclease cleavage site maps with particular reference to the evolution of humans and the apes. Evolution 37:221-224. Than PP, Jeewon R, Hyde KD, Pongsupasa mit S, Mongkolporn O, Taylor PWJ. 2008. Characterization and pathogenicity of species associated with anthracnose disease on chilli ( spp.) in Thailand. Plant Pathol 57:562. Thompson AK. 1987. The development and adaptation of methods for control of anthracnose. In: Prinsley RT, Tucker G, eds. Mangoes: a review. London: Commonwealth Science Council. p. 29-38. Thompson JD, Higgins DG, Gibson TJ. 1994. CL USTAL W: improving the sensitivity of progressive multiple sequence alignments through sequence weighting, position specific gap penalties and weight matrix choice. Nucl Acids Res 22:4673-4680. Timmer LW, Agostini JP, Zitko SE, Zulfiger M. 1994. Postbloom Fruit Drop, an increasingly prevalent disease of citrus in the Americas. Plant Dis 78:329-334. Turgeon BG. 1998. Application of mating type ge ne technology to probl ems in fungal biology. Annu Rev Phytopath 36:115-137. 232

PAGE 233

Turner BC, Perkins DD, Fairfield A. 2001. from natural populations: a global study. Fungal Genet Biol 32:67-92. Urea-Padilla AR, MacKenzie SJ, Bowen BW Legard DE. 2002. Etiology and population genetics of spp. causing crown and fruit rot of strawberry. Phytopathology 92:1245-1252. Vakalounakis DJ, Fragkiadakis GA. 1999. Genetic diversity of isolates from cucumber: differentitation by pathogeni city, vegetative comp atibility, and RAPD fingerprinting. Phytopathology 89:161-168. Vialle A, Feau N, Allaire M, Didukh M, Mart in F, Moncalvo JM, Hamelin RC. 2009. Evaluation of mitochondrial genes as DNA barcode for Basidiomycota. Mol Ecol Res 9:99-113. Vinnere O, Fatehi J, Wright SAI, Gerharson B. 2002. The causal agen t of anthracnose of Rhododendron in Sweden and Latvia. Mycol Res 106:60-69. Vivekananthan R, Ravi M, Ramanathan A, Samiyappan R. 2004. Lytic enzymes induced by and other biocontrol organism s mediate defence against the anthracnose pathogen in mango. Worl d J Microbiol Biotech 20:235-244. von Arx JA. 1957. Die Arten der Gattung Cda. Phytopathol Z 29:413-468. Waller JM, Bridge PD, Black R, Hakiza G. 1993. Characterization of th e coffee berry disease pathogen, sp. nov. Mycol Res 97:989-994, Waugh J. 2007. DNA barcoding in animal species: pr ogress, potential and pitfalls. BioEssays 29:188-197. Weeds PL, Chakraborty S, Fernandes CD, Charchar MJ dA, Ramesh CR, Kexian Y, Kelemu S. 2003. Genetic diversity in from spp. at centers of origin and uti lization. Phytopathology 93:176-185. Wegulo SN, Yang XB, Martinson CA. 1998. Soybean cultivar response to in field and controlled environment studies. Plant Dis 82:1264-1270. Wendel JF, Doyle JJ. 1998. Phylogenetic incong ruence: window into genome history and molecular evolution. In: Soltis DE, Soltis PS, Doyle JJ, eds. Molecular Systematics of Plants II: DNA Sequencing. Boston, MA: Kluwer Academic Publishers. 574 p. Weng FY, Chuang TY. 1995. Grouping of mango anthracnose fungus in Taiwan. Plant Protec Bull (Taichung 37) 3:295-309. Went FAFC. 1904. Krulloten en versteende vrucht en van de cacao in Suriname. Verhandeligen der K akademie van wetenschappen. Amsterdam 2:1-40. 233

PAGE 234

Wheeler HE. 1954. Genetics and evol ution of heterothallism in Phytopathology 44:342-345. White MM, Jame s TY, ODonnell K, Cafaro MJ, Tanabe Y, Sugiyama J. 2006. Phylogeny of the Zygomycota based on nuclear ribosomal sequence da ta. Mycologia 98:872-884. White TJ, Bruns T, Lee S, Taylor JW. 1990. Amp lification and direct sequencing of fungal RNA genes for phylogenetics. In: Innis MA, Gelgard DH, Sninsky JJ, White TJ, eds. PCR protocols. San Diego, CA: Academic Press. pp 315-322. Whitehouse HLK. 1949. Heterothallism and sex in the fungi. Biol Rev Camb Philos Soc 24:411447. Wiley EO. 1978. The Evolutionary species c oncept reconsidered. Syst Zool 27:17-26. Wolfe KH. 2006. Comparative genomics and genome evolution in yeasts. Philos Trans R Soc Lond B Biol Sci 361, 403-412. Xiao CL, MacKenzie SJ, Legard DE. 2004. Genetic and pathogenic analyses of isolates from strawberry and noncu ltivated host. Phytopathology 94:446453. Xu J, Ali RA, Gregory DA, Amick D, Lamber t SE, Yoell HJ, Vilgalys RJ, Mitchel TG. 2000. Uniparental mitochondrial transm ission in sexual crosses in Curr Microbiol 40:269-273. 2005. The inheritance of organelle genes and genomes: patterns and mechanisms. Genome 45:951-958. Yagushi Y, Uchara K, Nojima H, Kobayash i T, Nakamura S. 1998. Occurrence of a new anthracnose in Japan, caused by Edinburgh, Scotland: ISPP/ICPP, 1998. Abstract 6.11. Yakoby N, Beno-Moualem D, Keen NT, Dinoor A, Pines O, Prusky D. 2001. is an important virulence factor in avocado fruit-fungus interaction. Mol Plant Microbe In 14:989-995. Freeman S, Dinoor A, Keen NT, Prusky D. 2000a. Expression of pectate lyase from in promotes pathogenicity. Mol Plant Microbe In 13:887-891. Kobiler I, Dinoor A, Prusky D. 2000b. pH regulation of pectate lyase secretion modulates the attack of on avocado fruits. Appl Environ Microbiol 66:1026-1030. 234

PAGE 235

235 Yoder AD, Irwin JA, Payseur BA. 2001. Failure of the ILD to determine data combinability for slow loris phylogeny. Syst Biol 50:408-424. Yun S-H, Berbee ML, Toder OC, Turgeon BG. 1999. Evolution of th e fungal self-fertile reproductive life style from self-sterile an cestors. Proc Natl Acad Sci 96:5592-5597. Zaparoli G, Cabrera OG, Medrano FJ, Tiburcio R, La Cerda G, Pereira GG. 2009. Identification of a second family of genes in the causal agent of witches broom disease in cacao, encoding necrosis-induc ing proteins similar to cerato-platanins. Mycolog Res 113:61-72. Zare R, Gams W, Starink-Willemse M and Summerbell RC. 2007. a suitable genus for and a new genus for Nova Hedwigia 85:463-489.

PAGE 236

BIOGRAPHICAL SKET CH Tara Tarnowski is originally from Des Moin es, IA. Although she is from an agricultural state, she grew up in the city. As a kid, she help ed her mother in their large vegetable garden every summer, and it is here that her interest in plants and agricultur e began. In high school, she had the opportunity to spend two summers in Cost a Rica, and these trips sp arked her interest in how people lived in different place s in the world. She attended Iowa State University for her Bachelor of Science degree, where she majored in Plant Health & Protection and International Agriculture, and minored in Spanish. At Iowa State she had the opportunity to work with Dr. Mark Gleason on a range of fruit crop diseas e management research, and completed two independent research projects on soot y blotch and flyspeck on apple. In 2003 she moved to Athens, GA to begin an MS in Plant Pathology. She studied with Dr. Harald Scherm, characterizing fungicide efficacy in blueberry flowers against which causes mummy berry of blueberry. After completing her degree she moved south again to Gainesville to begin her PhD, and later to the Tropical Research and Education Center in Homestead, FL, to complete her dissertation research. 236