DETECTION METHODS AND TAXONOMIC CLARIFICATIONS FOR THREE EMERGING TREE PATHOGENS IN FLORIDA DIPLODIA CORTICOLA, DIPLODIA QUERCIVORA AND RAFFAELEA LAURICOLA By TYLER JAMES DREADEN A DISSERTATION PRESENTED TO THE GRADUATE S CHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014
2014 Tyler Dreaden
To my w ife
4 ACKNOWLEDGMENTS I am tr uly appreciative for the continuous support, advice, patience, and time my Committee Chair, Dr. Jason Smith, and committee members, Dr. John Davis, Dr. Randy Ploetz, Dr. Pamela S oltis and Dr. Michael Wingfield have shown me. I feel privileged to have had such wonderful guidance and assistance from my advisors. I would also like to express my gratitude to the University of Florida Forest Genomics Lab and the University of Pretoria Forestry & Agricultural Biotechnology Institute, for allowing me to use the ir facilities and their assistance, during the course of my research; and the members of the UF Forest Pathology lab, for their assistance. I also must thank my family and friends that have done so much to get me to this point
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION TO LAUREL WILT DISEASE, RAFFAELEA, DIPLODIA CORTICOLA AND D. QUERCIVORA ................................ ................................ ..... 10 The Genus Raffaelea and Laurel Wilt ................................ ................................ ..... 10 Cankers of Oaks Caused by Diplodia corticola and D. quercivora .......................... 13 2 PHYLOGENETIC STUDY OF THE GENUS RAFFAELEA ................................ ..... 16 Introduction ................................ ................................ ................................ ............. 16 Materials and Methods ................................ ................................ ............................ 18 Taxon Sampling ................................ ................................ ............................... 18 DNA Extraction, PCR Amplification, and Sequencing ................................ ...... 19 Phylogenetic Analyses ................................ ................................ ..................... 19 Results ................................ ................................ ................................ ............. 21 Discussion ................................ ................................ ................................ .............. 23 3 DEVELOPMENT OF MULTILOCUS PCR ASSAYS FOR RAFFAELEA LAURICOLA CAUSAL AGENT OF LAUREL WILT DISEASE ............................... 31 Introduction ................................ ................................ ................................ ............. 31 Materials and Methods ................................ ................................ ............................ 33 DNA Extraction ................................ ................................ ................................ 33 SSU Based Detection Method ................................ ................................ .......... 34 Microsatellite Based Detection Method Sequencing and Primer Design ........ 35 Microsatellite Based Detection Method Primer Screening ............................. 35 Microsatellite Based Detection Method PCR Conditions ................................ 36 Results ................................ ................................ ................................ .................... 38 SSU Based Detection ................................ ................................ ...................... 38 Microsatellite Based Detection ................................ ................................ ......... 39 Discussion ................................ ................................ ................................ .............. 39
6 4 DEVELOPMENT OF A PCR RFLP BASED DETECTION METHOD FOR THE OAK PATHOGENS DIPLODIA CORTICOLA AND D. QUERCIVORA ................... 48 Introduction ................................ ................................ ................................ ............. 48 Primer Design and Evaluation ................................ ................................ ................ 49 Restriction Enzyme Digestion ................................ ................................ ................. 51 Surve y of D. corticola / D. quercivora Isolates ................................ .......................... 52 Summary and Conclusions ................................ ................................ ..................... 53 5 CONCLUSION ................................ ................................ ................................ ........ 60 APPENDIX A DIPLODIA PLANT DISEASE NOTES ................................ ................................ ..... 62 B ADDICTIONAL RAFFAELEA PHYLOGENETICS FIGURES ................................ .. 66 LIST OF REFER ENCES ................................ ................................ ............................... 72 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 80
7 LIST OF TABLES Table page 2 1 Taxon names, isolate a nd GenBank accession numbers used in the study. ...... 25 3 1 Isolates used to screen Raffaelea lauricola specific microsatellite primers. ....... 43 3 2 Primer pairs used in the study. ................................ ................................ ........... 44 4 1 Taxa and sequences used to test the in silico specificity of the forward primers DcDq1 and DcDq1a. ................................ ................................ .............. 55 4 2 Size in bp of D. corticola and D. quercivora taxon specific PCR amplicons after digestion with Mse I, differences are italicized. ................................ ............ 56 4 3 Isolates used in the survey of symptomatic live oaks. ................................ ........ 56
8 LIST OF FIGURES Figure page 2 1 Raffaelea ML phylogeny from the combined, LSU, SSU, and BT dataset ......... 27 2 2 Raffaelea ML phylogram from the combined, LSU, SSU, and BT dataset. ........ 28 2 3 Raffaelea ML phylogeny with unidentified isolates. ................................ ............ 30 3 1 Standard curve for the qPCR R. lauricola detection. ................................ .......... 44 3 2 Photo of the 329 bp qPCR amplicons from primer pair chk. ............................... 45 3 3 Photo of the 322 bp qPCR amplicons from primer pair ifw. ................................ 46 3 4 Photo of qPCR a mplicons using primer pairs chk and ifw from healthy wood. ... 47 3 5 Images showing the sequences of the microsatellite loci, chk and ifw, from 9 R. lauricola isolates ................................ ................................ ........................... 47 4 1 L ocations of t he D. corticola and D. quercivora taxa specific forward PCR primers, DcDq1 and DcDq1a. ................................ ................................ ............. 57 4 2 Gel image from D. corticola / D. quercivora taxa specific PCR assay. ................. 58 4 3 Gel images of Mse I restriction enzyme digestion of DcDq1, DcDq1a and DcDq3 PCR amplicons. ................................ ................................ ...................... 59 B 1 LSU based phylogeny of Raffaelea ................................ ................................ ... 66 B 2 SSU based pylogeny of Raffaelea ................................ ................................ ..... 67 B 3 BT based phylogeny of Raffaelea ................................ ................................ ...... 68 B 4 Alignment of the ITS sequences from 10 Raffaelea spp. ................................ .... 69 B 5 Strict consensus of MP trees using the LSU dataset. ................................ ......... 70 B 6 Line gr aph comparing colony diameter for a range of temperatures for isolates PL1356 and PL1357 to the R. canadensis ................................ ........... 71
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in P artial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DETECTION METHODS AND TAXONOMIC CLARIFICATIONS FOR THREE EMERGING TREE PATHOGENS IN FLORIDA, DIPLODIA CORTICOLA, DIPLODIA QUERCIVORA AND RAFFAELEA LAURICOLA By Tyler James Dread en May 2014 Chair: Jason A Smith Major: Forest Resources and Conservation Laurel wilt of redbay ( Persea borbonia ) and other Lauracea e caused by Raffaelea lauricola and cankers of live oak ( Quercus virginiana ) caused by Diplodia corticola and D. quer civora affect the health of trees i n Florida. A phylogenetic study of Raffaelea and related taxa revea led that as currently defined the genus is not monophyletic and that Raffaelea and Grosmannia need to be reevaluated. Two methods to detect R. lauric ola were developed O ne utilizes the SSU locus and qPCR, is not taxon specific but is useful experimentally. Another uses multiple taxon specific loci and PCR reduces diagnostic time from 1 week to a day and can be used to diagnose laurel wilt A PCR RFLP based method detects and differentiates D. corticola and D. quercivora Using this method, a survey of isolates recovered from symptomatic live oaks confirm D. corticola in Duval, Highlands, Levy, Polk, and Washington Counties and D. quercivora in Alachua, Duval, and Marion Counties in Florida.
10 CHAPTER 1 INTRODUCTION TO LAUREL WILT DISEASE, RAFFAELEA, DIPLODIA CORTICOLA AND D. QUERCIVORA T he Genus Raffaelea and Laurel Wilt Ambrosia beetles have obligate mutualisms with asexual fungi and are one o f the few examples of agriculture in animals The ambrosial habit has evolved at least 7 times in these beetles ( Batra 196 7 Farrel et al. 2001 ). The beetles excavate galleries in host tree xylem in which the ambrosial fungi are cultivated as thick myce lial mats lining the galleries, and serve as the food source for both the adult and larva l stages Most ambrosia beetle symbionts fall into the Ambrosiella and Raffaelea genera (Batra 1967 Harrington et al. 2010). The genus Raffaelea (Fungi, Ascomycota, Sordariomycetes, Ophiostomatales, Ophiostomataceae) was created by Arx and Hennebert (1965) to accommodate R. ambrosiae a symbiont of Platypus ambrosia beetles. Raffaelea spp. usually colonize sapwood form close associations with ambrosia beetles and a re characterized by their production of sporodochia and by sympodial conidia development (Bartra 1967 Gebhardt and Oberwinkler 2005 not exhibit distinct morphological characteristics because conidiop hores lack pigmentation, are mostly simple, and the mode of conidiogenesis is barely visible with been recently described ( Harrington et al. 20 10). This lack of distinct characters and discovery of new taxa has led to some confusion in the taxonomy of the genus. Adding to this confusion has been debate regarding how to classify Raffaelea and some related genera. Gebhardt et al. (2005) created Dryadomyces to accommodate D. amasae clade in their rDNA small
11 ribosomal subunit (SSU) analysis but conidiogen esis differed from that of other Raffaelea spp Har r ington et al. (2008) synonymized Dryadomyces to Raffaelea so that all ambrosia beetle symbionts with similarities to Ophiostoma were included in Raffaelea Massoumi Alamouti et al. (2009) conducted a study of ambrosia fungi that included an analysis of morphology and a multigene phylogen e tic study. T heir results were consistent with the morphology reported by Gebhardt et al. (2005) and found that D. amasae formed a highly supported clade with R. montetyi and A. sulphurea They conclude d that Raffaelea should be revised. Harrington et al. (2010) maintain ed the synonym y of Dryadomyces wit h Raffaelea ( Harr ington et al. 2008) despite their SSU analysis placing D. amasae, R. montetyi and A. sulphurea in a clade with Grosmannia T hey used their rDNA large ribosomal subunit (LSU) analysis which did not include D. amasae for a new descripti on of Raffaelea and did not address the discrepancies in their SSU and LSU results. In summary the taxonomy of Raffaelea is presently unclear due to the lack of distinct morphological characters, discovery of new taxa and discord in the taxonomic litera ture. Laurel wilt is a vascular disease caused by R. lauricola T.C. Harr., Fredrich & Aghayeva It is transmitted by the exotic redbay ambrosia beetle, Xyleborus glabratus Eichhoff, and affect s redbay, Persea borbonia L. Spreng, and other members of the L auraceae in the southeastern United States including the commercial crop avocado, Persea americana in Florida (Rabaglia et al. 2006, Fraedrich et al. 2008, Harrington et al. 2008, Ploetz et al. 2011). The beetle was first detected in 2002 in Port Wentwo rth, GA and redbay mortality subsequently began in Georgia and South Carolina in 2003 and Florida in 2005 (Rabaglia et al. 2006, Fraedrich et al. 2008). As of Feb r uary 2014
12 t he disease wa s found in North and South Carolina, Georgia, across most of Florid a (5 2 of the 67 counties) Mississippi and Alabama (USDA Forest Service 201 4 ). The laurel wilt disease cycle in redbay is not completely understood I t is hypothesized that female redbay ambrosia beetles are guided to host trees by volatile compounds an d bore into the trees resulting in the ir inoculation of R. lauricola stored in the beetle s mandibular mycangia (Batra 1967, Hanula et al. 2008, Fraedrich et al. 2008, Harrington et al. 2008 Hughes et al. 2009, Hulcr and Dunn 2011 Niogret et al. 2011 ). This initial hole into a healthy tree is thought to be aborted by the beetle because ambrosia beetles usually construct galleries in stressed or dead tree s and the healthy tree is thought to be un suitable for gallery and brood development (Farrell 2001 Fr aedrich et al. 2008 ). Following this initial inoculation the fungus colonizes the host; tyloses are produced in vessels limiting wat er and nutrient transport leading to wilt and death of the tree in 5 to 12 weeks ( Fraedrich et al. 2008 ) thereby becoming suitable for gallery development and brood production Subsequent mass attack by the beetle is evidence d by frass tubes at gallery entrances and powdered wood at the base of affected tree s (Batra 1967, Fraedrich et al. 2008). The beetles lay eggs in gall eries and the resultant larvae feed upon R. lauricola that line s the galleries (Hanula et al. 2008, Pea et al. 2011). After 30 to 60 days mature females exit the host carrying R. lauricola in their mycangia, and identify new hosts thus completing the l aurel wilt disease cycle. Laurel wilt is unusual U nlike other ambrosial fungi which are typically saprobes, R. lauricola is able to kill susceptible trees with a single inoculation with as few as 100 conidia ( Hughes 2013, Ploetz et al. 2013). A few ot her pathogenic Raffaelea spp., such
13 as R. quercivora and R. quercus mongolicae do not cause systemic wilt s and require multiple inoculations to cause significant disease (Kubono and Ito 2002, Kim et al. 2009 Ploetz et al. 2013). X. glabratus is also unu sual because in its native range, S outh east Asia where it is associated with Lauraceous and other native trees, it is assumed although lacking experimental evidence, to act as a typical ambrosia beetle and bore into stressed and recently dead trees, but X. glabratus here in the United States, where it is exotic, is attracted to healthy redbay, along w ith other Lauraceae (Wood 1982 Harrington et al. 2011 Ploetz et al. 2013). It has been suggested this beetle has become a pest here in the United States b ecause of an olfactory mismatch where the volatiles from healthy trees here resemble that of hosts, stressed or recently dead trees, in its native range (Hulcr and Dunn 2011). Laurel wilt disease resulting from R. lauricola and its vector X. glabratus has been called a black swan event because the disease has had an extreme impact, was outside the realm of expectation and was unpredictable (Ploetz et al. 2013). Cankers of Oaks Caused by Diplodia corticola and D. quercivora S ince the 1980s a severe decli ne of oak trees has been observed in the Mediterranean region that has been associated with branch and trunk cankers caused by members of the Botryosphaeriaceae (Luque and Girbal 1989, Ragazzi et al. 1989, Brasier et al. 1993, Sanchez et al. 2003). M ember s of the Botryosphaeriaceae are pathogens, en d ophytes and saprophytes on a wide range of mainly woody hosts (Crous et al. 2006, Slippers and Wingfield 2007, Phillips et al. 2012, Phillips et al. 2013 Slipper s et al. 2013 ). Among t he most common and impor tant of these canker and dieback pathogens is Diplodia corticola A.J.L. Phillips, A. Alves & J. Luque (Fungi, Ascomycota, Dothideomycetes, Botryosphaeriales, Botryosphaeriaceae) which was
14 known previously as Botryosphaeria stevensii and B. corticola (Alve s et al. 2004, Crous et al. 2006). Recently D. corticola has been reported on Quercus agrifolia and Q. chrysolepis in California, Q. virginiana Mill. in Florida and Vitis vinifera in California and Texas (Urbez Torres et al. 2009 Lynch et al. 2010, Urb ez Torres et al. 2010, Dreaden et al. 2011 Appendix A ) Symptoms include branch and trunk cankers, quick decline, dieback and mortality on Quercus spp. and cankers on V. vinifera A recent study of Botryosphaeriaceae associated with declining Quercus sp p. in Tunisia found the main cause of cankers and branch dieback to be D. corticola although a new species D. quercivora Linaldeddu & A.J.L. Phillips caus ed similar symptoms on Q. canariensis (Linaldeddu et al. 2009, Linaldeddu et al. 2013). Similarly, during a survey of symptomatic Q. virginiana in Florida both D. corticola and D. quercivora were recovered from and shown to be pathogens on this host (Dreaden et al. 2011, Dreaden et al. 2014 a Appendix A ). It is interesting that D. quercivora was isola ted from Quercus spp. in both Tunisia and Florida at approximately the same time and that these are the only locations this fungus has been reported (Linaldeddu et al. 2013, Dreaden et al. 2014 a ). Where this species originated is not known. Botryosphaeria ceae are known to infect via wounds but they can also enter directly through lenticels, stomata and other openings on healthy plants ; in the absence of stress, they may establish only as endophyt es (Slippers and Wingfield 2007). O n Q. agrifolia i n Califor nia there is evidence for and against the involvement of the gold spotted oak borer, Agrilus auroguttatus in D. corticola infection ( Lynch et al. 2013 ) H ow D. corticola and D. quercivora infect Q. virginiana in Florida, whether these fungi are able to l ive as en d ophytes in th is host and the extent to which stress facilitates
15 disease development are not known. Since these fungi have been reported from cultivated settings but not natural areas i n Florida (Dreaden et al. 2011, Dreaden et al. 2014 a ) the y may have originated in the nursery, and there may be some unique stress in cultivated settings that makes the trees more susceptible, or perhaps a large enough survey has not been conducted B ut, at this time it is not known why the pathogens are not fo und in natural areas.
16 CHAPTER 2 PHYLOGENETIC STUDY OF THE GENUS RAFFAELEA Introduction The genus Raffaelea was created by Arx and Hennebert (1965) to accommodate R. ambrosiae a symbiont of Platypus ambrosia beetles and currently contains 20 species ( Harrington et al. 2010) Raffaelea w as traditionally distinguished from Ambrosiella which is also associated with ambrosia beetles, by the sympodial proliferation of the conidiogenous cell leaving lateral scars in Raffaelea and percurrent proliferation t hat leaves rings at the tips of the conidiogenous cell in Ambrosiella (Batra 1967, Harrington et al. 2008). This distinction is difficult to discern microscopically and its utility to distinguish the two genera has been questioned (Gebhardt and Oberwinkl er 2005, Harrington et al 2008). Both genera usually colonize tree host sapwood and form close associations with ambrosia beetles (Batra 1967). In general, the taxonomy of Raffaelea is difficult because they lack distinct morphological characters and t he mode of conidiogenesis is barely visible with light microscopy (Gebhardt and Oberwinkler 2005). Additionally, many new taxa have been recently described ( Harrington et al. 2010). Although most Raffaelea spp. live as saprophytes, colonizing dead and dy ing wood, some can be phyto pathogen s ( e.g. R. quercivora R. quercus mongolicae and R. canadensis ) that can cause significant damage to forests and fruit crops (Kubono and Ito 2002, Kim et al. 2009, Eskalen and McDonald 2011, Ploetz et al. 2013). The caus al agent of laurel wilt, R. lauricola is highly virulent and able to cause systemic wilt from a single inoculation ; it threatens native Lauraceae in the southeastern United States and avocado production in Florida (Ploetz et al. 2011, Ploetz et al. 2013)
17 The relationships between Raffaelea and related genera and their placement with in the Ophi ostomatales are unresolved Dryadomyces was introduced by Gebhardt et al. (2005) to accommodate D. amasae (= R. amasae ) because it fell in the Raffaelea clade in the ir phylogenetic analysis of the rDNA small ribosomal subunit (SSU) but its conidiogenesis is different Harrington et al. (2008) synonymized Dryadomyces with Raffaelea indicating that all ambrosia beetle symbionts with similarities to Ophiostoma should be included in Raffaelea until Ophiostoma was better resolved. Massoumi Alamouti et al. (2009) conducted a multigene (rDNA large ribosomal subunit (LSU), SSU, and beta tubulin (BT)) phylogenetic study of ambrosia fungi. They found the morphology of D. am asae was consistent with Gebhar d t et al. (2005) recognized D. amasae as a distinct monotypic lineage that formed a highly supported clade with R. montetyi and A mbrosiella sulphurea (= R. sulphurea ) conclud ing that Raffaelea should be revised. Harrington et al. (2010) revised Raffaelea maintaining the Harrington et al. (2008) synonym y of Dryadomyces with Raffaelea even though their SSU analysis plac ed D. amasae, R. montetyi and A. sulphurea in a clade with Grosmannia T o justif y this treatment they dis regard ed the SSU analysis and instead used their LSU analysis, which did not include D. amasae in their new description of Raffaelea. Molecular phylogenetic approaches have been used to clarify the relationships on which the taxonomy of most groups of fungi, including the O phiostomatales are being built on (Farrell et al. 2001, James et al. 2006, Duong et al. 2012, Slippers et al. 2013). Massoumi Alamouti et al. (2009) noted that phylogenies using the SSU indicate d that both Ambrosiella and Raffaelea are polyphyletic, suggesting that their similar morphology and an intimate association with beetles originated more than once
18 in these genera (Cassar and Blackwell 1996, Farrell et al. 2001). The ambrosial habit in beetles is also polyphyletic, having ar isen at least 7 times (Farrell et al. 2001). The polyphyletic nature of both ambrosial fungi and beetles indicates that the use of these associations to define Raffaelea as s uggested by Harrington et al. (2008) may not be warranted Although t he compre hensive ophiostomatoid fungi phylogenies by Zipfel et al. (2006) and Duong et al. (2012) may have helped clarify the taxonomy of Raffaelea it was not included in their studies. Plant pathogens in Raffaelea and other genera in the Ophiostomatales signific ant ly impact cultivated and natural area s (e.g. laurel wilt disease caused by R. lauricola in the United States and Dutch elm disease caused by Ophiostima novo ulmi among others, in North America and Europe ) (Ploetz et al. 2013 ) The discovery of new tax a the dissemination of known taxa to new areas, and current taxonomic uncertaint ies for both complicates the diagnos is and management of the diseases these pathogens cause Clarifying the taxonomy of Raffaelea will aid both researchers and diagnosticians The goals of this study were to conduct a multi gene phylogenetic analysis of Raffaelea test the monophyly of the genus as it is currently defined, and classify isolates with unclear affinities Materials and Methods Taxon S ampling The study by Masso umi Alamouti et al. (2009) was used as a base for loci (LSU, SSU, and BT because most of the sequences are available in GenBank and the loci have been phylogenetically useful for these fungi) and out groups used in this study. The description of Raffaelea followed Harrington et al. (2010) A total of 77 strains,
19 including 10 Microascales and 55 Ophiostomatale s ( 18 Ophiostoma 3 Ceratocystiopsis 11 Grosmannia 1 Esteya 2 Fragosphaeria and all 20 Raffaelea as defined by Harrington et al. (2010) ) taxa w ere used in this study (Table 2 1). DNA Extraction, PCR A mplification and Sequencing Sequences were either acquired from GenBank ( Table 2 1 ) or obtained by sequencing from DNA that was extracted from cultures (Table 2 1) with methods adapted from Justesen et al. (2002) and Duong et al. (2012). PCR s w ere performed using PCR primers NL1/LR3, NS1/NS4, and Bt2a/Bt2b for the LSU, SSU, and BT loci, respectively (Vilgalys and Hester, 1990, White et al. 1990, O'Donnell 1993, Glass and Donaldson 1995). Sanger seq uencing was performed using the above primers at the U niversity of F lorida I nterdisciplinary C enter for B iotechnology R esearch and consensus sequences were then constructed using both the forward and reverse sequences using Geneious Pro 5.6.6 (Biomatters Ltd., Auckland, New Zealand) See Appendix B for a discussion on amplifying and sequencing of the rDNA internal transcribed spacer region, ITS1 5.8s ITS2 (ITS). Phylogenetic Analyses DNA sequences generated for this study were aligned with sequences retri eved from GenBank (Table 2 1) using Geneious Pro 5.6.6, with the Geneious alignment default settings, then adjusted manually and trimmed The introns in the BT loci could not be unambiguously aligned and were removed from the dataset. The presence or ab sence of the BT introns were also coded, but gave similar MP results as the non intron coded dataset and was not used in subsequent analyses (see Appendix B). Concordance of the three gene datasets was evaluated by conducting a maximum likelihood (ML) ana lysis on each gene and then comparing the results visually The ML
20 analyses were conducted at the U niversity of F lorida High P erformanc e Computing C enter (HPC) using RAxML version 7.3.5 using the GTRGAMMAI model, as determined by JModelTest, with 100 dist inct starting trees and 1000 bootstrap analyses (BS) (Stamatakis 2006, Posada 2008). The concatenated data set, with 1849 characters total was analyzed using ML, as described above, with each gene in a separate partition. The MP analysis was conducted u sing PAUP* 4.0a129 with gaps treated as missing data, a heuristic search with 10 random stepwise additional replicates, tree bisection reconnection (TBR) branch swapping algorithum and branches with zero branch lengths were collapsed (Swofford 2003). Sup port was assessed by BS analysis using 1000 MP heuristic searches using TBR. The Bayesian Inference ( BI ) analysis was conducted at the HPC using MrBayes 3.2.1 using the GTR+I+G model with all parameters unlinked (adapted from JModelTest, see above), each gene in a separate character set, and 5 million generations that were sampled every 1000 generations (Ronquist et al. 2012). The burn in of 0 .25 was determined using Tracer 1.4 and the remaining trees were used to calculate the posterior probabilities (P P) and construct the majority rule consensus tree using MrBayes (Rambaut and Drummond 2007). To test the monophyly of Raffaelea B ayes factors (BF) were calculated by first conducting a BI analysis, as described above, with the addition of a constraint tha t the Raffaelea taxa form a clade. Next, BFs were calculated using the harmonic mean from MrBayes and the BF from T racer (Kass and Raftery 1995, Rambaut and Drummond 2007, Ronquist et al. 2012). Expected likelihood weight (ELW) and Shimodaira Hasegawa (S H) tests were conducted in RAxML, as described above, with the addition of a monophyletic Raffaelea constraint tree (Stamatakis 2006).
21 An additional ML analysis was performed, as above, to place potentially new undescribed species To do this, sequences from 7 additional isolates were added to the concatenated dataset, 5 from nutmeg, Myristica fragrans in Gr e nada that were isolated by Dr. Randy Ploetz and sequenced by Dr. Robert Blanchette (S_ID_21, S_ID_22, S_ID_ 28, S_ID_31, S_ID_32), 1 from avocado in California isolated by Dr. Akif Eskalen (PL1001 GenBank accession JF327799 from Eskalen and McDonald (2011)), and 1 from Thailand isolated and sequenced by Craig Bateman (PL1635). The 5 Gr e nada isolates had only SSU sequences available; the other isolate s had SSU, LSU, and BT sequences. Results The ML analyses of the individual genes showed weak support for both deeper nodes and terminal branches but the general topologies were similar (Append i x B). The most notable differences were the placement of Cera tocystiopsis and Fragosphaeria ; however I concluded as did Massoumi Alamouti et al. (2009), the genes were similar enough to combine. The taxa from the order O phiostomatales formed a highly supported clade with 100, 100, and 99 ML BS, BI PP, and MP BS, respectively. All three analyses supported placement of Ceratocystiopsis and Fragosphaeria in the O phiostomatales with high support ; however they could not be placed relative to the other genera because the individual gene phylogenies had different topol ogies (Figure 2 1, Figure 2 2, Append i x B). The Ophiostoma clade was well supported with 88, 100, and 77 ML BS, BI PP, and MP BS, respectively. Raffaelea spp. fell into two clades, one of which included R. amasae R. sulphurea R. quercus mongolicae R. quercivora R. montetyi and E. vermicola (97, 100, and 89 ML BS, BI PP, and MP BS, respectively) within the
22 Grosmannia clade (87, 100, 67, ML BS, BI PP, and MP BS, respectively) The second Raffaelea clade contained R. brunnea R. lauricola R. scolytodi s R. arxii R. gnathotrichi R. fusca R. subfusca R. ellipticospora R. ambrosiae (type for the genus) R. canadensis R. albimanens R. subalba R. tritirachium R. santoroi and R. sulcati (93, 100, and 87 ML BS, BI PP, and MP BS, respectively) was si ster to Grosmannia The placement of Fragosphaeria was disregarded because of the uncertainty in its placement resulting from the incongruence of the individual loci The log likelihood values from the ML unconstrained and the monophyletic Raffaelea con straint analyses were ( 15790.81 and 15822.69 ) and for the BI analyses were ( 15943.84 and 15973.97 from T racer) and ( 15960.43 and 15997.19 from MrBayes), respectively. The ELW test found the monophyletic constrain ed hypothesis to be significantly wor se than the polyphyletic Raffaelea unconstrained hypothesis (0.954 PP) but the SH test did not find a significant difference between the hypotheses at alpha < 0.05. The BFs were greater than 30 for both methods used, indicating very strong support for t he polyphyletic Raffaelea hypothesis (Kass and Raftery 1995 ). The ML analysis with the unidentified isolates provided evidence of 6 new taxa and supports the finding from Massoumi Alamouti et al. (2009) that isolate TR25 represents a distinct taxon (Figure 2 3). In the Grosmannia clade, S_ID_S28 was near R. sulphurea and (S_ID_S31 + S_ID_S32) was near R. amasae In the Raffaelea cl ade, (S_ID_S21 + S_ID_S22) and PL1004 (see chapter 3 for more information on this isolate) was near R. brunnea PL1001 was ne ar R. canadensis and PL1635 was near R. scolytodis The BS support was lower for many clades when compared to the analysis without the unidentified isolates (Figure 2 1 and Figure 2 3 ). This is probably
23 because of uncertainty in the placement of the Gre nada isolates because only SSU sequences were available for these isolates. Discussion This analysis found Raffaelea not to be monophyletic which is contrary to the MP LSU but consistent with the MP SSU analyses of Harrington et al. (2010). Three prob able explanations for this discrepancy include differences in taxa sampling, loci used, and methodology (MP vs. ML). The ML phylogenies of the individual gene datasets all found Raffaelea to be polyphyletic. A MP analysis was conducted on the LSU dataset to investigate if the difference in results was an artifact of the LSU dataset and MP (Appendix B 5). The MP study on the LSU dataset resulted in Raffaelea forming a clade, providing evidence that the discrepancy between this study and the Harrington et al. (2010) MP LSU study is an artifact of MP and the LSU dataset, though the effect, if any, of taxa sampled is not known. The ML study that included the unknown isolates revealed evidence of 6 possible new taxa (Figure 2 3). This suggests there are likel y many more Raffaelea spp. to be found and care must be taken in identifying isolates and designing diagnostic/ detection methods because new undescribed taxa are likely to be encountered. For example, PL1004 was identified as R. lauricola based on SSU da ta, but was shown later to be nonpathogenic and is now thought to be a new species ( see chapter 3 ) Likewise, PL1001 was previously reported as R. canadensis ( Eskalen and McDonald 2011) but shown in the present study to differ from that species ( s ee Appen dix Figure B 6 for more information on these isolates) A more detailed study that includes more isolates of these taxa is needed to determine which represent new species.
24 The ML study of the individual gene datasets along with the ML, BI, and MP studies of the combined dataset all indicate that Raffaelea, as currently defined, is polyphyletic. Esteya vermicola along with R. amasae R. sulphurea R. quercus mongolicae R. quercivora and R. montetyi formed a highly supported clade in Grosmannia with the r emaining Raffaelea spp. forming a highly supported clade sister to Grosmannia Of the 3 tests of Raffaelea monophyly that were used, only the SH test found the constrained tree not to be different than the unconstrained tree. This finding is not surprisi ng as the SH test has been shown to be conservative (Shimodaira and Hasegawa 1999, Strimmer and Rambaut 2002, Czarna et al. 2006). These results in combination with the ambrosial habit in beetles being polyphylet ic, having arisen at least 7 times, indica te the association of ambrosial fungi and ambrosial beetles to define Raffaelea as suggested by Harrington et al. (2008), is of limited use (Farrell et al. 2001). Taken as a whole the evidence suggests that Raffaelea needs to be reevaluated and that Gros mannia needs to be included in this reevaluation. This study recognizes R. brunnea R. lauricola R. scolytodis R. arxii R. gnathotrichi R. fusca R. subfusca R. ellipticospora R. ambrosiae R. canadensis R. albimanens R. subalba R. tritirachium R. santoroi and R. sulcati as Raffaelea sensu stricto and that R. amasae R. sulphurea R. quercus mongolicae R. quercivora and R. montetyi should be removed from Raffaelea b ut it is not clear where they should be placed. One possibility is to place t hem in Grosmannia Alternatively, these species might be accommodated in a reinstated Dryadomyces with D. amasae as the type. Hopefully, ongoing phylogenomic work by Vanderpool and Mccutcheon (personal communication) will help clarify how to proceed.
25 Ta ble 2 1 Tax on names isolate and GenBank accession numbers used in the study. Taxon_Isolate Accession LSU SSU BT Ambrosiella_ferruginea _CBS408.68 EU984285 EU984254 EU977461 Ambrosiella_ferruginea _JB13 EU984286 EU984255 EU977462 Ambrosiella_hartig ii _CBS404.82 EU984288 EU984256 EU977463 Ambrosiella_ips _CBS435.34 EU984289 AY858657 EU977464 Ambrosiella_macrospora _CBS367.53 EU984290 EU984257 EU977465 Ambrosiella_tingens _CBS366.53 EU984293 EU984258 EU977468 Ambrosiella_xylebori _CBS110.61 EU984294 AY 858659 EU977469 Ceratocystiopsis_manitobensis _UM237 DQ268607 EU984266 DQ268638 Ceratocystiopsis_minuta _CBS463.77 DQ268615 EU984267 EU977481 Ceratocystiopsis_minuta bicolor _CBS635.66 DQ268616 EU984268 EU977482 Ceratocystis_adiposa _CBS600.74 EU984304 EU9 84263 EU977479 Ceratocystis_coerulescens _CL13 12 AY214000 EU984264 AY140945 Ceratocystis_moniliformis _CBS155.62 EU984305 EU984265 EU977480 Claviceps_fusiformis _ATCC26019 U17402 DQ522539 AF263569 Daldinia_concentrica U47828 U32402 FJ185285 Epichloe_typ hina U17396 AB105953 X52616 Esteya_vermicola _CBS115803 EU668903 FJ490552 Fragosphaeria_purpurea _CBS133.34 AF096191 AF096176 Fragosphaeria_reniformis _CBS134.34 AB189155 AB278193 Grosmannia_abiocarpa _MUCL18351 AJ538339 EU984269 DQ097857 Grosmannia_cl avigera _ATCC18086 AY544613 EU984270 AY263194 Grosmannia_cucullata AJ538335 AY497513 EU977483 Grosmannia_penicillata DQ097851 AY858662 DQ097861 Grosmannia_piceiperda AY707209 AY497514 AY707195 Grosmannia_serpens DQ294394 AY497516 AY707188 Leptographium _abietinum _DAOM60343 DQ097852 EU984271 AY263182 Leptographium_fruticetum _DAOM234390 DQ097848 EU984272 DQ097855 Leptographium_longiclavatum AY816686 EU984273 AY288934 Leptographium_lundbergii _UAMH9584 AY544603 EU984274 AY263184 Leptographium_terebrantis _UAMH9722 AY544606 EU984275 AY263192 Microascus_cirrosus _CBS217.31 AF275539 EU984279 EU977490 Ophiostoma_abietinum AF155685 EU984276 EU977484 Ophiostoma_bicolor DQ268604 AY497512 DQ268635 Ophiostoma_canum AJ538342 EU984277 EU977485 Ophiostoma_floccosu m AJ538343 AF139810 AY789142 Ophiostoma_ips AY172022 AY172021 GU170412 Ophiostoma_montium _CBS15178 AY194947 EU984278 AY194963 Ophiostoma_novo ulmi _CMW10573 DQ294375 FJ430508 Ophiostoma_piceae AJ538341 AB007663 AY305698
26 Table 2 1. Continued Taxon_Iso late Accession LSU SSU BT Ophiostoma_pulvinisporum _CMW9022 DQ294380 EU977487 Ophiostoma_quercus DQ294376 AF234835 AY789157 Ophiostoma_setosum AF128929 AY305703 Ophiostoma_stenoceras _CMW3202 DQ294350 FJ176850 DQ296074 Ophiostoma_ulmi DQ368627 M832 61 EU977489 Ophiostomataceae _sp._TR25 EU984281 EU984251 EU977457 Penicillium_expansum U15483 DQ912698 AF003248 Petriella_setifera _CBS385.87 AF027666 EU984280 EU977491 PL1001 1 * PL1004 2 * Raffaelea_albimanens _CBS271.70 EU984296 EU984259 EU97 7471 Raffaelea_amasae _CBS116694 EU984295 AY858660 EU977470 Raffaelea_ambrosiae _CBS185.64 EU984297 AY497518 EU977472 Raffaelea_arxii _CBS273.70 EU984298 AY497519 Raffaelea_brunnea _CBS378.68 EU984284 AY858654 EU977460 Raffaelea_canadensis _CBS168.66 EU98 4299 AY858665 EU977473 Raffaelea_canadensis _CBS805.70 EU984291 AY858658 EU977466 Raffaelea_ellipticospora HQ688664 C2345 3 C2345 3 Raffaelea_fusca _C2394 3 EU177449 * Raffaelea_gnathotrichi _CBS379.68 EU177460 AY858655 Raffaelea_lauricola EU123077 EU123076 Raffaelea_lauricola _PL159 2 EU257806 Raffaelea_montetyi EU984301 AY497520 EU977475 Raffaelea_montetyi _PC06.001 JF909540 JF909512 Raffaelea_quercivora _MAFF410918 AB496454 AB496428 GQ225691 Raffaelea_quercus mongolicae _KACC44405 GQ225700 GQ225688 Raffaelea_santoroi _CBS399.67 EU984302 EU984261 EU977476 Raffaelea_scolytodis _CCF3572 AM267270 AM267261 Raffaelea_subalba _C2401 3 EU177443 * Raffaelea_subfusca _C2335 3 EU177450 * Raffaelea_sulcati _CBS806.70 EU177462 AY858666 EU977477 Ra ffaelea_sulphurea _CBS380.68 EU984292 EU170272 EU977467 Raffaelea_tritirachium _CBS726.69 EU984303 EU984262 EU977478 Sporothrix_humicola _CMW7618 EF139114 EF139100 Sporothrix_schenckii DQ294353 M85053 DQ296076 Sporothrix_schenckii _CMW7614 DQ294352 AY280 477 Taphrina_populina _CBS337.55 AF492050 D14165 AF170968 Xylaria _sp AY327481 U32417 AY951763 Sequenced in this study, 1 isolate UCR1073 from Eskalen and McDonald (2011) 2 from authors collection s 3 from Dr. T. C. Harrington Iowa State University
27 Figure 2 1. Raffaelea ML phylogeny from the combined, LSU, SSU, and BT dataset. Clades with multiple support values (ML bootstrap, BI posterior probabilities, and MP bootstrap) are labeled whereas clades with single support values have ML bootstrap val ues. T ype species for select genera are indicated in blue and isolates missing gene sequences have the genes that were used listed in red
28 Figure 2 2. Raffaelea ML phylogram from the combined, LSU, SSU, and BT dataset with BS support values. The scal e bar represents substitutions per site Penicillium expansum and Taphrina populina branches are not to scale.
30 Figure 2 3. Raffaelea ML phylogeny with unidentified isolates from the combined, LSU, SSU, and BT dataset. Clade support values are ML bootstrap percentages. Notice there is support for 6 new taxa 1. S_ID_S28, 2. S_ID_S31 and S_ID_S32, 3. S_ID_S21 and S_ID_S22, 4. PL1004, 5. PL1001, and 6. PL1635. Type species for select genera are labeled in blue and isolates missing gene sequences hav e the genes that are available listed in red
31 CHAPTER 3 DEVELOPMENT OF MULTILOCUS PCR ASSAYS FOR RAFFAELEA LAURICOLA CAUSAL AGENT OF LAUREL WILT DISEASE Introduction The ambrosia beetle symbiont, Raffaelea lauricola, is a unique pathogen of trees that has created an unprecedented phenomenon in American forests, laurel wilt ( Fraedrich et al. 2008 Harrington et al. 2008 Hulcr and Dunn 2011 ). Typically, ambrosia fungi are non pathogenic or weakly pathogenic to their tree hosts ( Hulcr and Dunn 2011 Ploe tz et al. 2013 ), but R. lauricola has challenged this paradigm in dramatic fashion. R. lauricola is transmitted by the Asian redbay ambrosia beetle ( Xyleborus glabratus ), which is not known to cause damage in its native range ( Rabaglia et al. 2006, Hulcr a nd Dunn 2011 ). Less than 2 years after its appearance in an area, more than 92% of the redbays > 7.6 cm in dia and 100% of those greater tha n 10.2 cm in dia were k illed ( Fraedrich et al. 2008 Goldberg and Heine 2009 ). As of August 2013 the disease was found in North and South Carolina, Geo rgia, across most of Florida ( 52 of the 67 counties) and Mississip pi and A labama ( Riggins et al. 2010, USDA Forest Service 201 4 ). Laurel wilt affects members of the Lauraceae plant family, including redbay ( Persea bor bonia ) ( Fraedrich et al. 2004 Fraedrich et al. 2008 ), swamp bay ( P. borbonia var. pubescens ) ( Mayfield 2007 ), silkbay (P. humilis ) ( Hughes et al. 2012 ), sassafras ( Sassafras albidum ) ( Smith et al. 2009 a ), and camphor ( Cinnamomum camphora ) ( Smith et al. 20 09 b ). Reprinted with Permission from Dreaden TJ, Davis JM, Smith JA, Harmon CL, Ploetz RC, Palmateer AJ, Soltis PS. 2014 b Development of multilocus PCR assays for Raffaelea lauricola causal age nt of laurel wilt disease. Plant Dis 98:379 383
32 Three shrubs (northern spicebush ( Lindera benzoin ), pondberry ( Lindera melissifolia ), endangered, and pondspice ( Litsea aestivalis ), critically endangered) are also susceptible ( Surdick and Jenkins et al. 2009 Fraedrich et al. 2011 Hughes et al. 2011 ). Substantial losses of a fruit crop, avocado ( Persea americana ) ( Mayfield et al. 2008 b Ploetz et al. 2011 Inch and Ploetz 2012 Ploetz et al. 2012 ), are developing in commercial groves in Florida ( Ploetz et al. 2012 ). In Florida alone, avocado is worth in excess of $60 million annually, and more valuable production is threatened in California and Latin America ( Evans et al. 2010 ). Management of laurel wilt is limited to injection of trees with systemic fungicides prior to infection ( Mayfield et al 2008 a ) and the judicious use of sanitation, particularly in avocado orchards ( Spence et al. 2013 ). It has been demonstrated that the vector and fungus do not survive in wood that has been chipped ( Spence et al. 2013 ). Thus, early detection of symptomat ic trees and subsequent sanitation may slow the development of laurel wilt epidemics, particularly in areas devoid of significant densities of natural hosts. Monitoring based on trapping the vector is problematic because trapping efficiencies for X. glabr atus are poor and the disease often develops before it is detected. In addition, several additional potential vector species have been identified ( Harrington and Fraedrich 2010 Carrillo et al. 2014 ). Currently, diagnosis of laurel wilt requires recovery of R lauricola on a semi selective medium ( Ploetz et al. 2012 ), followed by sequencing of the small subunit (18s) or large subunit (28s) of the rDNA and/or the demonstration of pathogenicity for examined isolates (unpublished, Dreaden et al.). No taxon specific PCR assay for the
33 pathogen has yet been reported. The aim of this study was to develop reliable and accurate real time and traditional PCR based methods to identify R lauricola and diagnose laurel wilt. Materials and Methods DNA Extraction The DNA extraction method for samples used in this study was adapted from Justesen et al. (2 00 2). Wood samples were ground using a mortar and pestle with liquid nitrogen. Fungal conidia/mycelia samples (10 50 mg) were lysed by adding 0.8 ml of extraction buf fer and three cycles of freezing at 80C for 5 min followed by incubating at 65C until thawed. The ground wood samples were suspended in 0.8 ml extraction buffer (25 g liter 1 D sorbitol, 10 g liter 1 N lauroylsarcosine, 8 g liter 1 hexadecyltrimethylam monium bromide, 0.8 M NaCl, 20 mM EDTA, 10 g liter 1 polyvinylpolypyrrolidone, 0.1 M Tris, pH 8), 2 l RNase A (100 mg ml 1 ) was added, inverted 5 times, and incubated at 65C for 25 min. The mixture was cooled to room temperature, 0.8 ml chloroform was a dded, inverted 30 times and centrifuged for 5 min at 16,000 RCF (relative centrifugal force). The supernatant was transferred to new 2 ml tubes, 20 l of proteinase K (20mg l 1 ) was added, the suspension was inverted 4 6 times and incubated for 20 min at 55C. The tubes were cooled to room temperature, 0.8 ml chloroform was added, inverted 30 times and centrifuged for 5 min at 16,000 RCF. The supernatant was transferred to new 2 ml tubes, one volume of ice cold 100% isopropanol was added, and the tubes were inverted and centrifuged for 15 min at 16,000 RCF. The DNA pellets were washed with 350 l 75% ethanol, centrifuged for 5 min at 16,000 RCF, ethanol was removed, and then the pellets were air dried and resuspended in 100 l TE buffer. PCR was condu cted on all DNA extractions using the
34 rDNA SSU primers NS1 and NS4 ( White et al. 1990 ) to confirm that all DNA extractions contained amplifiable DNA before using. DNA from R. lauricola PL1390, culture was used to construct a serial dilution, which was us ed to determine the detection limit and as a positive control. DNA was obtained for R. brunnea, R. albimanens, R. ambrosiae and R. tritirachium from the Centraalbureau voor Schimmelcultures (Utrecht, The Netherlands) and for R. canadensis C592 R. montet yi C2221 R. albimanens C2223 R. brunnea C2229 R. subfusca C2253, and R. fusca C2336 and C2349 from Dr. T. C. Harrington (Iowa State University). SSU Based Detection Method An alignment of multiple 18s SSU rDNA sequences of R. lauricola and closely rela ted species, retrieved from GenBank, was evaluated for potential R. lauricola taxon specific PCR primer sites. Three primers PCR primers LWD1 and LWD3 that matched only R. lauricola but later produced a false positive and were discarded and a second se t of qPCR primers LWD3 and LWD4 that matched R. lauricola and R. brunnea were designed. PCR conditions for LWD3 and LWD4 were optimized by testing a range (50 60C) of annealing temperatures. The primers were also tested against DNA from a R. lauricola serial dilution, both redbay and avocado laurel wilt diseased (naturally infected and artificially inoculated with isolate PL159, respectively) and healthy wood, and DNA from related fungi: R. canadensis, R. montetyi, R. albimanens, R. brunnea, R. subfusc a and R. fusca The thermocycling protocol was 95C for 2 min, followed by 40 cycles of 95C for 15 s, 56C for 15 s, and 68C for 30 s with a melting curve analysis performed at the end. SYBR green was used as the detection method. The PCR mixture inc luded 11.25 l Eppendorf 2.5x Realmastermix/SYBR solution, 10.75 l water, 1 l LWD3 10 M primer, 1 l LWD4 10
35 M primer, and 1 l DNA template. All PCR s were replicated in triplicate with the R. lauricola serial dilution and no DNA negative control pres ent on every run using an Eppendorf Mastercycler ep Realplex, and Realplex 2.0 software was used for all qPCR data analysis (Eppendorf Inc., Hauppauge, NY). Microsatellite Based D etecti on Method Sequencing and Primer D esign One eighth of a Roche 454 GS FLX Titanium sequencer plate was used to generate sequences from R. lauricola isolate (PL716 gq996063) by the Interdisciplinary Center for Biotechnology Research (ICBR) at the University of Florida (Gainesville, FL ). Primer_designer.pl ( Castoe et al. 2010 ; modified by Matt Gitzendanner) was used to find simple sequence repeats, also known as microsatellites, and design primers to amplify them. Design parameters specified that each repeat contain a minimum of 8 repeat units and primers not be located withi n 25 bp of the repeat. The primers were then sorted by the number of repeats (large to small), repeat unit size (small to large) and product size (small to large). The 454 sequences were assembled using Geneious 5.4.6 de novo assembler using the high sen sitivity option with all other opt ions left to default settings (Biomatters Ltd. Auckland, New Zealand ). Microsatellite Based D etecti on Method Primer Screening Microsatellite primer pairs were first prescreened to limit the total number of PCR s The p rescreening consisted of testing the primer pairs with DNA from R. lauricola PL1390 to confirm the correct amplicon was produced and its two closest relatives: R. brunnea ( Massoumi Alamouti et al. 2009 ) and Raffaelea sp. PL1004 (undescribed fungus that pro duced the false positive with the LWD1 LWD3 SSU based method) to confirm no amplicons were produced Primer pairs that passed the prescreening were then tested against all of the isolates (Table 1). The primers were
36 tested against DNA from Raffaelea spp. from the two clades closest to R. lauricola ( R. montetyi, R. sulphurea, R. brunnea, R. ellipticospora, R. ambrosiae, R. subfusca and R. fusca ) ( Harrington et al. 2010 ) and included representatives ( R. subalba, R. tritirachium, R. albimanens and R. arxii ) from more distantly related clades ( Harrington et al. 2010 ). DNA from healthy redbay and avocado was included in the screening process to insure that the primers would not produce a false positive from host material. The selected R. lauricola taxon sp ecific primer pairs, chk and ifw (GenBank KF381410 and KF381411 respectively), were further tested against 9 isolates of R. lauricola This was done to confirm that the primers produced the correct amplicon for all isolates. Six isolates were from the US A; 3 contained polymorphisms (PL388, PL692, and PL735) and 3 were identical to all other isolates (PL159, PL570, PL571) according to an AFLP study (Marc Hughes, personal communication ). Two isolates, CBS 129001 and CBS 129006, from Taiwan, and 1 isolate, CBS 129007, from Japan ( Harrington et al. 2011 ), were also tested. The amplicons from the 2 primer pairs for the Microsatellite Based D etecti on Method PCR Conditions The optimal annealing temperature for the primer prescreening was determined by testing a range of temperatures, 60 68 C. 65 C was chosen as optimal because it was the highest temperature that did not reduce amplification and the higher temperature wou ld increase PCR stringency when compared to the lower temperatures ( Wu et al. 1991 ). The thermocycle profile consisted of 95C for 180 s, followed by 40 cycles of 95C for 35 s, 65C for 30 s, and 72C for 60 s with a melting curve analysis performed upon completion. All PCR s used in this experiment were replicated in triplicate with R. lauricola positive controls and non template negative controls present
37 for every PCR run using an Eppendorf Mastercycler ep Realplex (Eppendorf Inc., Hauppauge, NY). All PCR products were then visualized on agarose gels. Four DNA polymerases were tested: Biolase, (Bioline USA Inc. Randolph, MA), AmpliTaq, AmpliTaq Gold and AmpliTaq Gold 360 ( Life Technologies Corp., Carlsbad, CA ) for each. AmpliTaq Gold was used for the primer screening because it produced the lowest overall cycle threshold ( Ct ) values. PCR s for the primer screening consisted of 2.5 l of GeneAmp 10x PCR Gold buffer ( Life Technologies Corp ., Carlsbad, CA ), 2.5 l of 2 mM deoxynucleotide triphosphates (Bioline USA Inc. Taunton, MA ), 3 l of 25mM MgCl 2 1 l of 10 mM forward primer, 1 l of 10 mM reverse primer, 0.4 l of 10x SYBR Green I solut ion (Lonza Inc., Rockland, ME ), 0.125 l AmpliTaq Gold DNA polymeras e (Life Technolo gies Corp., Carlsbad, CA ), 13.475 l H 2 O, and 1 l of DNA template. All of the primers used in this study were synthesized by Integrated DNA Tech nologies, Inc. (Skokie, IL ). After the taxon specific primers were chosen, a range of primer annealing temperatures 60 65C was tested. A fifth DNA polymerase, Immolase (Bioline USA Inc. Randolph, MA) was also tested, which produced the same detection limit as Amplitaq Gold, produced more amplicons, costs less and was used in subsequent testing The PCR s were the same as above only 1.5 l of 50mM MgCl 2 and 14.975 l of H 2 O were used. The thermocycle profile for Immolase DNA polymerase consisted of 95C for 4 m, followed by 35 cycles of 95C for 35 s, 65C for 30 s, and 72C for 60 s with a me lting curve analysis performed upon completion. The primer pairs, chk and ifw (Table 2) were chosen for the diagnostic test because they did not produce amplicons in
38 the presence of related taxa or host DNA and amplified well from R. lauricola DNA. The t axon specific primers and the above methodology were tested in four independent laboratories (Carrie L. Harmon, Aaron J. Palmateer, Randy C. Ploetz and Jason A. Smith) to confirm their reliability. Results SSU Based D etecti on The primers, LWD1 and LWD3 ( Table 2), produced an 832 bp amplicon, and when a 68C primer annealing temperature was used the primers did not produce amplicons in the presence of R. amasae, R. arxii, R. ellipticospora, R. fusca, R. montetyi, R. subalba, R. subfusca, R. sulphurea or he althy redbay and avocado DNA However, when used to examine a suspect, dead avocado tree in August, 2009, they resulted in a spurious diagnosis for R. lauricola /laurel wilt A fungus from this tree, Raffaelea sp. PL1004, resembled R. lauricola phenotypic ally, and produced a SSU sequence (Genbank KF026302) that was 99% (971/980) similar to R. lauricola (JF797172, e value of 0.00). However, it was not pathogenic to avocado in two pathogenicity tests (Ploetz, unpublished ) and is now considered a closely rel ated but undescribed Raffaelea species. The qPCR primers LWD3 and LWD4 (Table 3 2) have a detection limit of 1 x 10 4 ng of R. lauricola DNA per qPCR. The standard curve constructed using the serial dilution (6.3 to 1 x 10 4 ng per qPCR) produced a slop e of 4.193, PCR efficiency of 0.73, R 2 of 0.997 and a Ct Std. Dev. of 0.64 at the lowest limit. There was no amplification or primer dimer formation with the no DNA template, healthy avocado or redbay wood negative controls. However, at least one replic ate of each related Raffaelea spp. tested produced an amplicon with approximately the correct melting
39 temperature. However, the amplification occurred late in the amplification process (mean Ct values > 30). Microsatellite Based D etecti on During 454 sequ encing, 80,237 reads were generated with an average length of 324 bp. Among the sequences, Primer_designer.pl found 3,103 microsatellites that contained fewer than 8 repeat units, 3,117 microsatellites without suitable flanking primers and 212 microsatell ites with suitable primers. The sequences containing the chk and ifw loci were located in the 454 sequence assemblies to confirm that they were not in the same contig, adding evidence that they were located in different parts of the genome. Both of the R lauricola taxon specific primer pairs had a detection limit of 0.1 ng of R. lauricola DNA per PCR, were able to quickly screen DNA from cultures, but were not able to reliably detect R. lauricola from diseased wood samples (see Figures 3 1 to 3 5 for all gel images and additional information). The primer pairs produced the correct amplicon for all isolates of R. lauricola that were tested. The 6 R. lauricola isolates from the southern USA and the 3 from Asia had identical DNA sequences for the ifw locus and all except the CBS 129007 isolate from Japan, which had 14 additional TCT repeats, had the same chk sequences. All fou r labs and Carrillo et al. (2014 ) were able to use the microsatellite based detection method to screen suspect R. lauricola isolate s confirming the consistency of the methodology. Discussion Given the emergent nature of this pathogen and its potential threat to natural forests and avocado culture globally, surveillance programs have been initiated in several countries (Australia, Isra el, Mexico and South Africa). The need for a practical diagnostic method that is consistently reliable in different labs is imperative. To date,
40 diagnostic methodology has relied on sequencing of the LSU and/or SSU, however, the time and extra resources needed for this approach make it unrealistic for many diagnostic labs. The false positive that was obtained with the SSU (LWD1 and LWD3 primers) and Raffaelea sp. (PL1004) illustrates the difficulty of designing taxon specific PCR assays for new pathogen s, as well as the dangers that are associated with the use of a single marker or locus for pathogen identification. These problems are especially acute when unknown, closely related taxa are present. To address these hurdles, microsatellite loci were use d as the basis for a new method to detect R. lauricola. Microsatellites are ideal for such an objective as they have limited inter taxon transferability ( Barbar et al. 2007 Cristancho and Escobar 2008 ), vary in length ( Schlotterer 2000 ), and have been u sed previously to detect pathogens ( Abd Elsalam et al. 2011 ). In addition, numerous methods have been developed to locate microsatellites and design primers for their amplification ( Thurston and Field 2005 You et al. 2008 Meglecz et al. 2010 ). The pri mers that were designed for the chk and ifw loci in the present study distinguished different isolates of R. lauricola from Asia and the USA from a wide array of closely related taxa. Their accuracy and successful use in 4 different laboratories demonstra tes their value in identifying this pathogen and usefulness in diagnosing laurel wil t. Recent ly, Carrillo et al. ( 2014 ) used the chk and ifw markers to verify the presence of R. lauricola in 6 xyleborine ambrosia beetles other than X. glabratus The new microsatellite based method reduced the time needed (from 1 week to 1 day) and cost of screening candidate R. lauricola isolates, because it eliminates the need for Sanger sequencing.
41 Despite the detection of the non specific amplicons late in the amplific ation cycles (mean Ct values > 30), the qPCR assay (primers LWD3 and LWD4) is still useful due to the highly sensitive and nearly specific nature of this assay. For example, the assay could be used to assess the rate of spread of R. lauricola in different hosts or under different environmental conditions after artificial inoculation, and has been used to detect and quantify R. lauricola in a P. americana cultivar screening study ( Ploetz e al. 2012 ). The detection limits for the SSU and microsatellite meth ods were quite different (0.0001 vs 0.1 ng per PCR, respectively). Some of this discrepancy may be due to differences in copy number between the PCR targets. Ribosomal DNA loci have ( Ga nley and Kobayashi 2007 Rooney and Ward 2007 ) while the microsatellite loci are assumed to be low in copy number. The increase in copy number means that less genomic DNA template is required for a PCR to reach the detection limit. This is one drawbac k of using low copy loci for detection methods. Still, the difference in copy number does not explain all of the difference in detection limits. Results with the rDNA SSU primers LWD1 and LWD3 demonstrate problems that can arise when detection of a new p athogen/disease is based on a single locus and undescribed, closely related taxa are present in host tissue. The qPCR primers LWD3 and LWD4 are not taxon specific and, therefore, cannot be used for diagnosing R. lauricola and laurel wilt. However, their high sensitivity and ability to quantitatively detect R. lauricola in wood samples makes them useful in experimental work with this pathogen ( Ploetz et al. 2012 ). In contrast, the microsatellite based assay is taxon
42 specific. Although it is not sensitive enough to reliably detect the fungus in wood samples (titers of this pathogen are low in diseased tissue) ( Ploetz et al. 2013 ), it can be used to quickly screen DNA from cultures for the presence of R. lauricola The microsatellite assay is now used rout inely in diagnostic clinics of the University of in new host trees and ambrosia beetles.
43 Table 3 1. Isolates used to screen Raffaelea lauricola specific microsatel lite primers. PL# Other n ames Sp. or notes 1008 Paraphaeosphaeria sp. 1009 CBS 271.70 Raffaelea albimanens 893 C2750 R. amasae 894 CBS 185.64 R. ambrosiae 896 C2372 R. arxii 1194 CBS 378.68 R. brunnea 890 C2345 R. ellipticospora 891 C2394 R. fusca 159 Avocado B R. lauricola Duval Co. FL 388 Pierce Co. R. lauricola Pierce Co. GA 570 MK1 R. lauricola Clay Co. FL 571 MK2 R. lauricola Clay Co. FL 692 R. lauricola Appling Co. GA 716 MS2 R. lauricola Jackson Co. MS 717 MS3 R. lauricola Jackson Co. MS 735 Columbia Co. R. lauricola Columbia Co. FL 1467 CBS 129007 R. lauricola Japan 1468 CBS 129001 R. lauricola Taiwan 1469 CBS 129006 R. lauricola Taiwan 1007 R. lauricola from Ambrosiodmus lecontei 894 C2221 R. montetyi 1001 1073 Raf faelea. sp 1002 1080 R. sp 1003 1081 R. sp 1005 R. sp from X. affinis 1000 3649 R. sp from X. ferrugineus 996 3645 R. sp from X. glabratus 999 3648 R. sp from X. glabratus 1006 R. sp from X. sp. 1004 RL5 R. sp. from dead Persea americana 892 C2 401 R. subalba 889 C2335 R. subfusca 895 C593 R. sulphurea 896 CBS 726.69 R. tritirachium
44 Table 3 2. Primer pairs used in the study. Forward p rimer Reverse p rimer Name Sequence Name Sequence Amplicon s ize (bp) chk f GTTCCACAGCCTGGAAAACC chk r GGTAGAGGACGATGGTTGGC 329 371 ifw f TCGAGGTCGTGGACTACAGC ifw r GTGGGACTCGCTGATGAGG 322 LWD1 CCCTGGTGATTCATGATAACT TCT LWD3 AACGCGTCAAAAGACAACAG 832 LWD4 TTTCTAGGACCGCCGTAATG LWD3 AACGCGTCAAAAGACAACAG 232 Figure 3 1. Standard curve for the qPCR R. lauricola detection method using primers LWD3 and LWD4. Standards are from three replicates with 6.3 ng, 1ng, 0.1ng, 0.01ng, 0.001 ng and 0.0001ng of template DNA per qPCR ( slope 4.193, PCR efficiency 0.73, R 2 of 0.997 and a Ct Std. Dev. of 0.64 at th e lowest limit ). Unknowns are DNA extractions from an unidentified R. lauricola culture and infected wood samples.
45 Figure 3 2. Photo of the 3 29 bp qPCR amplicons from primer pair chk using DNA from related and unknown fungi recovered from ambrosia beetles. All reactions are repeated in triplicate: 1 R. subfusca, 2 PL1002, 3 R. ellipticospora, 4 PL1003, 5 R. fusca 6 PL1004, 7 R. subalba, 8 PL1008, 9 R. amasae 10 PL1006, 11 R. montetyi 12 R. lauricola, 13 R. sulphurea 14 PL1008, 15 R. arxii 16 R. brunnea 17 PL996, 18 R. albimanens 19 PL999, 20 R. ambrosiae 21 PL1000, 22 R. tritirachium 23 PL1001, 24 no DNA negative control, 25 1ng R. lauricola 26 0.1ng R. lauricola and 27 0.01ng R. lauricola Photos by Tyler Dreaden.
46 Fig ure 3 3. Pho to of the 322 bp qPCR amplicons from primer pair ifw using DNA from related and unknown fungi recovered from ambrosia beetles. All reactions are repeated in triplicate: 1 R. subfusca 2 PL1002, 3 R. ellipticospora, 4 PL1003, 5 R. fusca 6 PL1004, 7 R. sub alba, 8 PL1008, 9 R. amasae 10 PL1006, 11 R. montetyi 12 R. lauricola, 13 R. sulphurea 14 PL1008, 15 R. arxii 16 R. brunnea 17 PL996, 18 R. albimanens 19 PL999, 20 R. ambrosiae 21 PL1000, 22 R. tritirachium 23 PL1001, 24 no DNA negative control, 25 1ng R. lauricola 26 0.1ng R. lauricola and 27 0.01ng R. lauricola Photos by Tyler Dreaden.
47 Fig ure 3 4. Photo of qPCR amplicons using primer pairs chk, top, and ifw, bottom. All react ions are repeated in triplicate using DNA from: 1 1ng R. lauri cola 2 healthy P. borbonia wood, 3 healthy P. americana wood, and 4 no DNA negative control Photos by Tyler Dreaden. Fig ure 3 5. Images showing the sequences of the microsatellite loci, chk top and ifw bottom, from 9 R. lauricola isolates. Not e the 14 additional TCT repeats in the Japanese isolate, CBS 129007 in the chk locus and that all of the sequences from the ifw locus are identical.
48 CHAPTER 4 DEVELOPMENT OF A PCR RFLP BASED DETECTION METHOD FOR THE OAK PATHOGENS DIPLODIA CORTICOLA AND D. QUERCIVORA Introduction Diplodia corticola A. J. L. Phillips, Alves et Luque is a well known canker pathogen of oak ( Quercus ) that is contributing to the decline of oaks in the Mediterranean region ( Alves et al. 2004 ). Recently it has been affecting Q uercus spp. in California, Vitis vinifera in California and Texas ( Urbez Torres et al. 2009, Lynch et al. 2010 Urbez Torres et al. 2010 ) and live oak ( Q. virginiana Mill.) in Florida ( Dreaden et al. 2011 Appendix A ). D. quercivora Linaldeddu & A. J. L. Phillips is closely related to D. corticola and causes similar cankers on oaks and has only been found in Tunisia and Florida ( Linaldeddu et al. 2013, Dreaden et al. 2014 a Appendix A ). In 2010, live oak trees in a landscape in Florida were found with nume rous branch cankers that were caused by D. corticola ( Dreaden et al. 2011, Appendix A ). In 2011 after recovering additional morphologically similar Diplodia isolates, DNA sequencing revealed that they fell into two groups, D. corticola ( Dreaden et al. 201 1, Appendix A ) and a fungus that was later described as D. quercivora ( Linaldeddu et al. 2013, Dreaden et al. 2014 a Appendix A ). The two species can be distinguished based on morphology and by comparing DNA sequences ( Linaldeddu et al. 2013 ), however, th ese methods are time consuming and increase the cost of differentiating isolates. The goal of this study was to develop a method to rapidly and cost effectively screen and identify suspect isolates from oak as either D. corticola or D. quercivora based on DNA, without the need for sequencing. Here we describe the development of an assay that involves PCR with taxa specific primers and restriction enzyme digestion to amplify and differentiate DNA from D. corticola and D. quercivora
49 Primer Design and Evalua tion One isolate each of D. corticola and D. quercivora recovered from Q. virginiana in Florida (isolate PL1010 from Highlands County and isolate PL1345 from Alachua County, respectively) was selected and used for primer development. DNA was extracted fr om pure cultures using a method adapted from Justesen et al. ( 2002 ). Five loci: rDNA small subunit (SSU), rDNA internal transc ribed spacer, ITS1 5.8s ITS2 (ITS), tubulin (Bt) and translation factor 1 alpha (EF1) were amplified and Sanger sequenced for both isolates using primers NS1/NS4, ITS1F/ITS4, LR0R/LR5 ( White et al. 1990 ), Bt2a/Bt2b ( Glass and Do naldson 1995 ), and EF1 728F/EF1 986R ( Carbone and Kohn 1999 ), respectively. The sequences from each locus were then aligned with sequences of related taxa, retrieved from GenBank (http://www.ncbi.nlm.nih.gov/genbank/), and regions that had sequence diverg ence were identified. One suitable site was found in the ITS region (Figure 4 1). The ITS LSU sequences used for primer design were deposited in GenBank (isolate PL1010 accession is KF500478; isolate PL1345 accession is KF500479). Geneious Pro 5.6.6 (Bio matters Ltd., Auckland, New Zealand) was used to design taxa specific primers and to check their specificity in silico allowing a maximum mismatch of 3 bp between primer and target sequences compared to ITS sequences from related taxa (Table 4 1 and Figur e 4 1). The phylogenetic study by Phillips et al. ( 2012 ) was used as the basis for included taxa and sequences used with the addition of D. quercivora ( Linaldeddu et al. 2013 ), D. agrifolia ( Lynch et al. 2013 ) and others retrieved from GenBank. Matching primer binding sites were found only in D. corticola and D. quercivora The taxa specific forward primers DcDq1 (CGAGTGCTACGAGCGA G A) and DcDq1a (CGAGTGCTACGAGCGA A A) have 1
50 nucleotide difference, bold font, and were 100% matches to D. corticola and D. que rcivora respectively A mixture of both forward primers was required for PCR because both only weakly amplify from the other, non matching, taxon. Primer specificity to D. corticola and D. quercivora was also tested empirically by performing PCR using DN A extracted ( Aljanabi and Martinez 1997 ) from cultures (obtained from The Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, Pretoria, South Africa) using the closest related species ( D. cupressi CBS168.87 and D. tsugae CBS41 8.64) and several more distantly related species ( D. agrifolia ATCC MYA4895 obtained from Akif Eskalen, UC Riverside, D. africana CBS120835, D. mutila CBS431.82, D. pinea CBS117911, D. scrobiculata CBS118110 D. seriata CBS112555, Neodeightonia phoenicum C BS122528 N.subglobosa CBS448.91 and Haplosporella acacia CBS469.69) ( Philips et al. 2012, Linaldeddu et al. 2013 Lynch et al. 2013 ). D. corticola (CBS112549 from FABI and isolates UCROK1246 and UCROK488 from Akif Eskalen) and D. quercivora (PL1265 reco vered from a live oak in Jacksonville, Florida) isolates were used as positive controls. The quality of all DNA samples used in this study was checked by amplifying a portion of the SSU, using primers NS1/NS4 to confirm they contained amplifiable DNA ( Whi te et al. 1990 ). PCR consisted of 2.5 l of 10x ImmoBuffer, 2.5 l of 2 mM deoxynucleotide triphosphates (Bioline USA Inc. Taunton, Ma), 1.5 l of 50 mM MgCl 2 1 l of 10 mM DcDq1 primer, 1 l of 10 mM DcDq1a primer, 1 l of 10 mM DcDq3 primer, 0.125 l immolase DNA polymerase (Bioline USA Inc. Taunton, Ma), 14.375 l H 2 O, and 1 l of DNA template. PCR conditions for DcDq1 (CGAGTGCTACGAGCGAGA) / DcDq1a (CGAGTGCTACGAGCGAAA) / DcDq3 (GAGCCACATTCCCGAGGT) were optimized
51 by testing a range (52 68 C) of anne aling temperatures. The final thermocycler conditions consisted of 94 C for 5 min followed by 37 cycles of 94 C for 30 s, 65 C for 30 s, and 72 C for 45 s. D. corticola and D. quercivora positive controls were included in each series of PCRs and subs equent Mse I digestions. PCR products were then viewed on 1% (w/v) agarose gels using SYBR Green added to the loading buffer The sensitivity of the PCR was tested using dilutions of D. corti cola and D. quercivora DNA (1 and .1 ng of DNA per reaction) with 1 ng per reaction needed for consistent amplification (Figure 4 2). The taxa specific primers only produced an amplicon (1,039 bp) in the presence of D. corticola and D. quercivora DNA with no amplicons being detected from reactions with related taxa DNA (Figure 4 2). Restriction Enzyme Digestion Geneious Pro 5.6.6 was used to identify restriction enzyme digestion sites that could distinguish PCR amplicons, using primers DcDq1/DcDq1a/DcDq3, of D. corticola from D. quercivora A Mse I site was found that produces different sized fragments that could readily separate D. corticola from D. quercivora (Table 4 2). The restriction enzyme digestion consisted of .25 l Mse I 10,000 units/ml, 10 l P CR amplicons in PCR buffer, incubation at 37 C for 8 h and then viewed on 2% (w/v) agarose gels (Figure 4 3). D. corticola and D. quercivora could be distinguished by the absence of the 173 bp band and presence of the 83/88 bp band in D. quercivora for a ll DNA samples with at least 1ng of DNA. When digested amplicons were stained by mixing SYBR Green with loading buffer as described above, the resulting gel revealed bands that were less clear (Figure 4 3) when compared to post staining with SYBR Green ( data not shown) or staining with
52 ethidium bromide (Figure 4 3). But, D. corticola and D. quercivora were differentiated with all three staining methods tested. Under these non optimal restriction digestion conditions (digestion in PCR buffer), signs of i ncomplete digestion were observed (Figure 4 3). This could likely be eliminated by using greater quantities of enzyme, less DNA, or by purifying the PCR product and digesting it using the recommended buffer. However, the small levels of incomplete digest ion did not affect the ability to differentiate the two species, so no extra steps are required for their identification. The use of a restriction enzyme digestion of the D. corticola / D. quercivora amplicons eliminates the need for DNA sequencing; rather polymorphism at an enzyme TTAA subunit (LSU) rDNA from D. quercivora holotype (CBS 133852) is not available to us, but we infer the holotype is near identical to D. quercivora PL1345 (from this study), since the two isolates have only 2bp differences in the EF1 and ITS loci ( Dreaden et al. 2014 a Appendix A ). Additionally, this region is highly conserved with only 1 bp difference between D corticola PL1010 and D. quercivora PL1 345 Survey of D. corticola / D. quercivora Isolates A survey of fungal isolates from symptomatic live oaks was conducted which included 9 isolates from 6 counties in Florida (Table 4 3). The samples were first cultured on artificial media ( Dreaden et al. 2011, Dreaden et al. 2014 a Appendix A ) as it was difficult to extract DNA from some of the very hard and woody tissue DNA was extracted from the cultures and the quality confirmed, as described above. PCR and the Mse I digestion was then preformed, as described, using D. corticola / D. quercivora positive controls and a no template negative control present for all PCR runs. The unknown isolates were then identified by the distinctive bands (Table 4 2) when the
53 Mse I digested PCR amplicons were fractionize d on an agrose gel. The species identities are given in Table 4 3. Summary and Conclusions We have developed the first assay for the rapid and accurate detection of the globally invasive oak pathogens D. corticola and D. quercivora Our results provide a solution to the problem of distinguishing among these species, which incite similar disease symptoms on live oaks. Previously, identifying and differentiating these pathogens required isolation on artificial medium followed by DNA extraction, PCR, sequ encing and sequence analysis or time consuming morphological comparisons. Using this method a PCR followed by restriction digestion with Mse I is sufficient. This method could potentially be applied directly to symptomatic plant tissues, eliminating the ne ed for cultures. However, this has not been assessed experimentally yet. Since D. quercivora is a potentially destructive pathogen ( Linaldeddu et al. 2013 ), which has only been re cently discovered in the U.S. (Dreaden et al. 2014 a Appendix A ), agencies a nd plant disease diagnostic clinics should monitor its distribution and evaluate its contribution to oak damage and mortality. As a result of this detection method and the subsequent survey, D. corticola and D. quercivora has now been confirmed from live oak in (Duval, Highlands, Levy, Polk, and Washington) and (Alachua, Duval, and Marion) Counties, respectively, in Florida. D. quercivora and D. corticola can be distinguished from each other based on morphology ( Linaldeddu et al. 2013 ) but this is not id eal because of their many similarities. Although it is not known whether these pathogens have been misidentified previously, Diplodia isolates in fungal collections should be evaluated to determine if these species are present. Such surveys would provide improved information regarding
54 the origin and distribution of these new pathogens. The protocols developed in this study will make it possible to shorten the length of time required for such work without the need for DNA sequencing.
55 Table 4 1. Taxa an d sequences used to test the in silico specificity of the forward primers DcDq1 and DcDq1a. Accession Species Strain EU673315 Diplodia acerina CBS 910.73 EF445343 Diplodia africana STE U 5908 JX174054 Diplodia agrifolia ATCC MYA 4895 FJ888461 Diplodia alatafructa CMW22635 FJ888478 Diplodia alatafructa CMW22721 GQ923853 Diplodia bulgarica CAP332 AY259100 Diplodia corticola CBS 112549 FJ790842 Diplodia corticola UCD2397TX KF500478 Diplodia corticola PL1010 EU673317 Diplodia coryli CBS 242.51 DQ458 893 Diplodia cupressi CBS168.87 GQ923855 Diplodia fraxini CAP302 GQ923858 Diplodia intermedia CAP273 EU673316 Diplodia juglandis CBS 188.87 GQ923865 Diplodia malorum CAP271 EU673318 Diplodia medicaginis CBS 500.72 AY259093 Diplodia mutila CBS 112553 DQ458886 Diplodia mutila CBS230.30 EU392302 Diplodia olivarum CBS121887 GQ923873 Diplodia olivarum CAP301 DQ458895 Diplodia pinea CBS393.84 DQ458897 Diplodia pinea CBS109727 DQ458898 Diplodia pinea CBS109943 EU080922 Diplodia pseudoseriata UY671 EU 080933 Diplodia pseudoseriata UY1263 JX894205 Diplodia quercivora BL8 JX894206 Diplodia quercivora BL9 KF500479 Diplodia quercivora PL1345 JQ694105 Diplodia ramulicola PDD:95987 EU430265 Diplodia rosulata CBS116470 AY253292 Diplodia scrobiculata CMW1 89 DQ458899 Diplodia scrobiculata CBS109944 DQ458900 Diplodia scrobiculata CBS113423 AY259094 Diplodia seriata CBS 112555 EU673319 Diplodia spegazziniana CBS 302.75 DQ458888 Diplodia tsugae CBS418.64 DQ458890 Lasiodiplodia theobromae CBS124.13
56 Ta ble 4 2. Size in bp of D. corticola and D. quercivora taxon specific PCR amplicons after digestion with Mse I, differences are italicized. D. corticola isolate PL1010 D. quercivora isolate PL1345 173 88 83 361 361 354 354 145 145 Table 4 3. Isol ates used in the survey of symptomatic live oaks. Isolate number Date collected FL county Species PL1346 8/29/11 Alachua D. quercivora PL1347 9/13/11 Marion D. quercivora PL1476 4/23/12 Levy D. corticola PL1477 4/23/12 Levy D. corticola PL1478 8/29/1 2 Washington D. corticola PL1480 7/3/12 Duval D. corticola PL1518 9/19/12 Polk D. corticola PL1519 9/19/12 Polk D. corticola PL1520 9/19/12 Polk D. corticola
57 Figure 4 1. Top, shows the locations of the D. corticola and D. quercivora taxa specif ic forward PCR primers, DcDq1 and DcDq1a, in the rDNA internal transcribed spacer 1 and the reverse primer, DcDq3, in the rDNA large subunit of D. quercivora isolate PL1345 GenBank accession KF500479. Bottom, shows the consensus sequence for D. corticola (entry #38) and D. quercivora (#39) primer sequences along with the primer binding sites in D. corticola 34) and D. quercivora 37) with the corresponding locations in the rDNA ITS sequences of related taxa (#1 #31) used in the in silico spe cificity test.
58 Figure 4 2. Gel (1% w/v agarose) image from D. corticola / D. quercivora taxa specific PCR assay showing the 1,039 bp amplicon in duplicate (using primers DcDq1, DcDq1a and DcDq3) with DNA from 1. D. scrobiculata 2. D. pinea 3. D. afric ana 4. D. seriata 5. D. cupressi 6. Neodeightonia phoenicum 7. Haplosporella acacia 8. D. tsugae 9. N. subglobosa 10. D. mutila 11. D. agrifolia 12. live oak negative control, 13. D. corticola CBS112549, 14. D. corticola UCROK1246, 15. D. corticol a UCROK488, 16. D. quercivora PL1265, 17. D. corticola PL1010 1ng, 18. D. corticola PL1010 0.1ng, 19. D. quercivora PL1345 1ng, 20. D. quercivora PL1345 0.1ng, and 21. no DNA template. Photo by Tyler Dreaden.
59 Figure 4 3. Gel (2% w/v agarose) images o f Mse I restriction enzyme digestion of DcDq1, DcDq1a and DcDq3 PCR amplicons from Figure 4 2, using DNA from D. corticola (13. CBS112549, 14. UCROK1246, 15. UCROK488 and 17. PL1010 1ng) and D. quercivora (16. PL1265 and 19. PL1345 1ng). D. quercivora is differentiated from D. corticola by the absence of the 173 bp band and presence of the 83/88 bp band in the D. quercivora digestions. The top gel was stained using ethidium bromide (the marker starts at 800 bp and decrease in 100 bp increments), the botto m gel was stained by mixing SYBR Green with the loading buffer (the marker starts at 650 bp and decreases in 50 bp increments). Photos by Tyler Dreaden.
60 CHAPTER 5 CONCLUS ION In this work, detection methods were developed and the taxonomy was clarified fo r pathogens that cause 2 tree diseases in Florida, Diplodia canker of live oak ( Diplodia spp. ) and laurel wilt ( Raffaelea lauricola ). A phylogenetic study of Raffaelea did not support the monophyly of the genus as it is currently described and Raffaelea s ensu stricto was defined. The phylogenomic work by Vanderpool and Mccutcheon in combination with this study should clarify the genera in which E. vermicola R. amasae R. sulphurea R. quercus mongolicae R. quercivora and R. montetyi belong. This study also found evidence for 6 new taxa and suggests that more undescribed taxa are likely to be encountered in future work The high likelihood of encountering undescribed taxa in conjunction with the lack of distinct morphological characters in these fungi indicates that diagnosticians and researchers must be caut ious when identifying isolates and developing methodology. The need for a reliable and inexpensive detection method along with the finding of multiple undescribed Raffaelea spp. lead to the develo pment of a multilocus R. lauricola detection method. This method is based on the amplification of 2 taxon specific microsatellite loci Due to the unique nature of these markers, the assay w ould not be expected to produce false positives if undescribed a nd closely related taxa are assessed in the future. This method can be used to aid laurel wilt disease management by accurately and quickly identifying the pathogen. Th e approach laid out in this study can also be applied to other pathogens in which trad itionally used loci do not have enough sequence divergence for the development of taxon specific PCR assays.
61 D etection and identification p roblems that new undescribed taxa can cause were also observed with t he Diplodia spp. on live oak. Since D. cortico la and D. quercivora are similar and cause the same types of symptoms on live oak a method was needed that could quickly and cost effectively differentiate these taxa. This was achieved with PCR primers that produced an amplicon for both species which, i n turn, produced diagnostic fragments after digestion with the restriction enzyme Mse I. This assay makes a large scale survey feasible that could be used to determine the distribution s of these pathogens in individual trees and across landscapes, and coul d help assess the disease predisposing factors and endophytic infection
62 APPENDIX A DIPLODIA PLANT DISEASE NOTE S First Report of Diplodia corticola Causing Branch Cankers on Live Oak ( Quercus virginiana ) in Florida Numerous cank ers on small branches showing dieback were observed on live oak ( Quercus virginiana ) trees in September 2010 in Marion County, FL. Approximately 24 12 year old landscape trees planted on a farm displayed symptoms. Samples were collected from six of the sym ptomatic trees and returned to the laboratory for processing. Isolations were made from canker margins after surface sterilization of samples in 2.5% sodium hypochlorite and by plating on potato dextrose agar (PDA). A suspect Botryosphaeriaceae sp. (based on colony morphology) was consistently isolated from the symptomatic branches fr om all 6 trees sampled. Fungal colonies consisted of plentiful, white, aerial mycelium that turned dark olive after 5 to 7 days at 23C with the underside of the cultures turni ng black ( Alves et al. 2004 ). Total genomic DNA from three representative Botryosphaeriaceae sp. isolates was extracted and the internal transcribed spacer (ITS1 5.8s ITS2) region of the rDNA (GenBank Accessions Nos. JF798638, JF798639, and JF798640) us ing the primers ITS1 and ITS4 (White et al. 1990 tubulin gene (Bt), (GenBank Accession Nos. JF798641, JF798642, and JF798641) using the primers Bt2a and Bt2b ( Glass and Donaldson 1995 ) were amplified, sequenced, and deposited in GenBan k. BLASTn searches of the ITS rDNA sequences resulted in 100% homology (467 of 467, 467 of 467, and 540 of 540, respectively) with Diplodia corticola isolate CBS 112074 (GenBank Accession No. Reprinted with permission from Dreaden TJ, Shin K, Smith J A. 2011. First report of Diplodia corticola causing branch cankers on live oak ( Quercus virginiana ) in Florida. Plant Dis. 95:1027.
63 AY268421). BLASTn searches of the Bt sequences resulted in 99, 9 8, and 99% (391 of 393, 396 of 400, and 392 of 394, respectively) matches with D. corticola strain UCD2397TX, GenBank Accession No. GU294724. To complete Koch's postulates, 9 seedlings of Q. virginiana 0.6 to 0.9 cm in diameter at ground line maintained i n a greenhouse, were inoculated with isolate PL949 (GenBank Accession Nos. JF798638 and JF798641) by making a 1.5 cm incision with a single edge razor blade into the xylem 10 cm above ground line. Inoculations were done by placing mycelial plugs (1 0.25 cm) from cultures on PDA in the incision with the mycelium facing the center of the stem. Wounds were sealed by wrapping them with Parafilm. Three negative controls were mock inoculated as previously described except sterile PDA plugs were used. Eight week s postinoculation, the lengths of the necrotic lesions were measured. Mean lesion length of the inoculated seedlings was 4.12 cm SE .45 and ranged between 2.7 and 6.3 cm. The negative control inoculations showed no necrotic lesions. Three of the inoculat ed seedlings were plated on PDA in an effort to reisolate the inoculated fungus. D. corticola was reisolated from each and all had the same ITS sequence as D. corticola strain CBS 112074. To our knowledge, this is the first report of D. corticola causing c ankers on Q. virginiana and the first report of the disease occurring in Florida. D. corticola has been reported to cause cankers and dieback in several Quercus spp. in Greece, Hungary, Italy, Morocco, Portugal, and Spain and has recently been reported to cause cankers on Q. chrysolepis and Q. agrifolia in California.
64 First Report of Diplodia quercivora Causing Shoot Dieback and Branch Cankers on Live Oak ( Quercus virginiana ) in the United States In September 2010, live oak ( Quercus virginiana Mill.) tr ees in an Alachua County, FL, shopping center parking lot were observed with shoot dieback and cankers on small branches. Isolations were made from canker margins by surface sterilizing tissue in 2.5% sodium hypochlorite and plating on potato dextrose agar (PDA) and incubating at 23C. Fungi morphologically similar to Diplodia quercivora Linaldeddu & A.J.L. Phillips (mycelium initially velvety and white and later turning pale to dark olivaceous and grayish in reverse) were consistently iso lated from symptom atic tissue (Linaldeddu et al. 2013 ). The two l oci used by Linaldeddu et al. (2013 ) in the description of D. quercivora were sequenced to identify a representative isolate (PL1345) as D. quercivora The internal transcribed spacer (ITS) (GenBank Accession No. KF386635) and translation factor 1 alpha (EF sequenced using primers ITS1F/I TS4 (White et al. 1990) and EF1 728F/EF1 986R (Carbone et al. 1999 ). BLASTn searches of the two sequences resulted in 99% (467 of 469 and 257 of 259, respectively) homology with D. quercivora CBS 133852, confirming the fungal isolates' identity as D. quercivora In October 2011, Koch's postulates were verified by inoculating, repeated twice, three Q. virginiana saplings (stem diameters, 12 to 14 mm; at inoculation sites approximately 50 mm above soil line) with isolate PL1345. Agar plugs (3 3 mm) taken from the margin of a 12 day old culture on PDA were inserted into flaps in the stems made by a sterile blade with the mycelia facing the Reprinted w ith permission from Dreaden TJ, Black AW, Mullerin S, Smith J A. 2014a First Report of Diplodia quercivora Causing Shoot Dieback and Branch Cankers on Live Oak ( Quercus virginiana ) in the USA. Plant Dis. 98:282.
65 cambial tissue. One negative control tree was mock inoculated with a sterile PDA plug. All inoculation sites were sealed with Parafilm and maintained in a greenhouse (19 to 29C). Trees were assessed for symptoms 90 days after inoculation. External bleedi ng was noted on all but one tree, and all flaps became necrotic. Pycnidia were observed on the outer surface of the flap on one inoculated tree. Negative controls showed no bleeding and their tissue flaps remained alive. Vertical length of phloem necrosis and percent of stem girdling were measured after removing the bark. Mean necrotic length and percent girdling for inoculated saplings were 48 mm (standard error [SE] = 10.6) and 26.6% (SE = 5.7) for the first inoculation and 46 mm (SE = 17) and 25% (SE = 5 ) for the second, respectively. Controls showed no internal necrosis and all produced healthy callus tissue at inoculation sites. Two of the pathogen inoculated trees per inoculation were sampled and the pathogen was re isolated from each. Recovered fungal isolates were confirmed as D. quercivora based on morphology and 100% ITS sequence homology to PL1345. D. quercivora was first described as causing shoot dieback and cankers on Q. canariensis in Tunisia and was found to be pathogenic to three additional M editerranean oak species, Q. ilex, Q. pubescens and Q. suber ( Linaldeddu et al. 2013 ). To our knowledge, this is the first report of D. quercivora causing cankers on Q. virginiana and the first report of the fungus outside of Tunisia. Given the damage thi s pathogen has caused there, efforts to monitor the spread of this disease would seem warranted. More research is needed to assess the risk this pathogen poses to North American oaks, however
66 APPENDIX B ADDICTIONAL RAFFAELEA PHYLOGENETICS FIGURES Figu re B 1 LSU based p hylogeny of Raffaelea and related taxa. The bipartition tree from the RAxML maximum likelihood analysis with 1000 bootstrap replicates is shown. Five clades are shaded to aid compari sons of the individual gene phylogenies (B 1 through B 3).
67 Figure B 2 SSU based p ylogeny of Raffaelea and related taxa. The bipartition tree from the RAxML maximum likelihood analysis with 1000 bootstrap replicates is shown. Five clades are shaped to aid comparisons of the individual gene phylogenies B 1 through B 3.
68 Figure B 3 BT based p hylogeny of Raffaelea and related taxa. The bipartition tree from the RAxML maximum likelihood analysis with 1000 bootstrap replicates is shown. Five clades are shaped to aid comparisons of the individual gene phylogenies, B 1 through B 3
69 Figure B 4. Alignment of the rDNA internal transcribed spacer region, ITS, from 10 Raffaelea spp N ote that only the 5.8s and portions of the 28s LSU could be aligned unambiguously. The bar graph at the top of the figure indicates the degree of similarity among sequences. The ITS in Raffaelea is unfortunately hard to PCR amplify and sequence (Harrington et al. 2011). With the use of FastStart Taq and the GC RICH solution (Roche Applied Science, Basel, Switzerland) the IT S region could be amplified from many Raffaelea isolates H owever few could be sequenced and those that were sequenced could not be aligned. Because of the sequencing and alignment problems with the ITS region in Raffaelea spp. it was not used in the ph ylogenetic study.
70 Figure B 5. Strict consensus of 20,700 most parsimonious trees using the LSU dataset and PAUP* (465 total characters, 230 were constant, and 195 were parsomony informative ) G aps were treated as missing data, and the TBR branch swapp ing algorithm was used (Swofford 2003). This LSU MP analysis found Raffaelea as defined by Harrington et al. (2010) to be monophyletic, as did their LSU MP analysis Where as, the ML (LSU, SSU and BT) analyses (Figure B 1 through B 3) as well as the ML, BI and MP analyses using the combined data sets, in this study, found Raffaelea spp. in both the Raffaelea and the Grosmannia clades (Figure 2 1 and Figure 2 2). These findings provide evidence the cause of the incongruence between our findings, that Raff ealea is not monophyletic, and the Harrington et al. (2010) LSU MP study is because of an interaction between the LSU dataset and MP.
71 Figure B 6. Line graph comparing colony diameter, after 7 days, for a range of temperatures for isolates PL1356 and PL1 357 to the R. canadensis type CBS168.66. Error bars are 95% confidence intervals The growth rates of isolates identified as R. canadensis by Eskalen and McDonald (2011) were compared to that of the R. canadensis t ype ( CBS168.66 ). Eight mm d iameter plu gs from each isolate actively growing on MEA were placed on 2 MEA plates per isolate and incubat ed at constant temperature s (10, 15, 20, 25, 30, and 35 C) in the dark. Colony diameter s w ere measured at right angles after 7 days. The growth of isolates w as similar from 10 to 30C but at 35 C PL1356 and PL1357 did not grow and CBS168.66 had grown to 27 mm in diameter. These findings, of different growth rates, are consistent with the results from the phylogenetic study, chapter 2, that found the isolat e from Eskalen and McDonald (2011) was closely related to but separate from R. canadensis 0 5 10 15 20 25 30 35 10 15 20 25 30 35 Mean colony diameter mm Temperature C PL1356 PL1357 CBS168.66
72 LIST OF REFERENCES Abd Elsalam K, Bahkali AH, Moslem M, De Wit PJGM, Verree t J. 2011. Detection of Mycosphaerella graminicola in wheat leaves by a microsatellite di nucleotide specific p rimer. Int J Mol Sci 12 :682 693 Aljanabi SM, Martinez I. 1997. Universal and rapid salt extraction of high quality DNA for PCR base d techniques. Nucleic Acids Res 25:4692 4693. Alves A Co rreia A, Luque J, Phillips A. 2004. Botryospha eria corticola sp. nov. on Quercus species, with notes and description of Botryosphaeria stevensii and its anamorph, Diplodia mutila Mycologia 96:598 613. Arx JA von, Hennebert GL. 1965. Deux champignons ambrosia. Mycopathol ogia et Mycologia Applicata 25 : 309 315. Barbar T, Palma Silva C, Paggi GM, Bered F, Fay MF, Lexer C. 2007. Cross species transfer of nuclear microsatellite markers: potential a nd limitations. Mol. Ecol. 16 :3759 3767. Batra LR. 1967. Ambrosia fungi: a taxonomic revision and nutritional status of some species. Mycologia 59:979 1017 Brasier CM, Robredo F, Ferraz JFP. 1993. Evidence for Phytophthora cinnamomi involvement in Iberian oak decline. Plant Pathol 42:140 145, doi:10.1111/j.1365 3059.1993.tb01482.x Carbone I Kohn L M. 1999. A me thod for designing primer sets for speciation studies in filamentous ascomycetes. Mycologia 91:553 556. Carrillo D, Duncan RE, Ploetz JN, Campbell AF, Ploetz RC, Pea JE. 2014 Lateral transfer of a phytopathogenic symbiont among native and exotic ambrosia beetles. Plant Pathol 63:54 62, DOI: 10.1111/ppa.12073. Cassar S, Blackwell M. 1996. Convergent origins of ambrosia fungi. Mycologia 88: 596 601. Castoe TA, Poole AW, Gu W, Jason De Koning AP, Daza JM, Smith EN, Pollock D D. 2010. Rapid identification of th ousands of copperhead snake ( Agkistrodon contortrix ) microsatellite loci from modest amounts of 4 54 shotgun genome sequence. Mol Ecol Resour 10:341 347. Cristancho M, Escobar C. 2008. Transferability of SSR markers from related Uredinales species to the co ffee rust Hemileia vastatrix Genet Mol Res 7:1186 1192. Crous PW, Slippers B, Wingfield MJ, Rheeder J, Marasas WFO, Philips AJL, Alves A, Burgess TI, Barber P, Groenewald JZ 2 006. Phylogenetic lineages in the Botryosphaeriaceae. Stud Mycol 55: 235 253.
73 Czarna A, Sanjun R, Gonzlez Candelas F, Wrbel B. 2006. Topology testing of phylogenies using least squares methods. Evol Biol 6:105. Dreaden TJ, Black AW, Mullerin S, Smith J A. 2014 a First Report of Diplodia quercivora Causing Shoot Dieback and Branch Cankers on Live Oak ( Quercus virginiana ) in the United States. Plant Dis 98:282 Dreaden TJ, Davis JM, Smith JA Harmon CL, Ploetz RC, Palmateer AJ, Soltis P S. 2014 b Development of multilocus PCR assays for Raffaelea lauricola causal agent of laurel wilt disease. Plant Dis. 98:379 383. Dreaden TJ, Shin K, Smith J A. 2011. First report of Diplodia corticola causing branch cankers on live oak ( Quercus virginiana ) in Florida. Plant Dis 95:1027. Duong TA, de Beer ZW, Wingfield BD, Wingfield MJ. 2012. Phylogeny and taxonomy of species in the Grosmannia serpens complex. Mycologia 104:715 732, doi:10.3852/11 109. Eskalen A, McDonald V. 2011. First Report of Raffaelea canadensis Causing Laurel Wilt Disease Symptoms on Avocad o in California. Plant Dis 95:9 1189. Eva ns EA, Crane JH, Hodges A, Osborne J L. 2010. Potential economic impact of laurel wilt disease on the Florida avocado industry. HortTechnology 20:234 238. Farrell BD, Sequeira ASO, Meara BC, Normark BB, Chung JH, Jordal BH. 2001. The evolution of agricultur e in beetles (Curculionidae: Scolytin ae and Platypodinae). Evolution 55: 2011 2027 Fraedrich SW, Harrington TC, Bates CA, Johnson J, Reid LS, Best GS, Leininger TD, Hawkins T S. 2011. Susceptibility to Laurel Wilt and Disease Incidence in Two Rare Plant Spe cies, Pon dberry and Pondspice. Plant Dis 95:1056 1062. Fraedrich SW, Harrington TC, Rabaglia RJ, Ulyshen MD, Mayfield IIIAE, Hanula JL, Eickwort JM, Miller D R. 2008. A fungal symbiont of the redbay ambrosia beetle causes a lethal wilt in redbay and other L auraceae in the so utheastern United States. Plant Dis 92:215 224. Fraedrich SW, Harrington TC, Rabaglia RJ. 2004. Laurel wilt: A new and devastating disease of redbay caused by a fungal symbiont of the exotic redbay ambrosia beetle. Newsletter of the Mich igan Entomological Society 52:14 15. Ganley ARD, Kobayashi T. 2007. Highly efficient concerted evolution in the ribosomal DNA repeats: Total rDNA repeat variation revealed by whole genome sh otgun sequence data. Genome Res 17:184 191. Gebhardt H, Oberwinkle r F. 2005. Conidial development in selected ambrosial species of the genus Raffaelea Antonie von Leeuwenhoek 88: 61 66.
74 Gebhardt H, Weiss M, Oberwinkler F. 2005. Dryadomyces amasae : a nutritional fungus associated with ambrosia beetles of the genus Amasa ( Coleoptera: Cu rculionidae, Scolytinae). Mycol Res 109: 6 87 696. Glass NL, Donaldson G C. 1995. Development of primer sets designed for use with the PCR to amplify conserved genes fro m filamentous ascomycetes. Appl Environ Microbiol 61:1323 1330. Goldberg N, Heine J. 2009. A comparison of arborescent vegetation pre (1983) and post (2008) outbreak of the invasive species the Asian ambrosia beetle Xyleborus glabratus in a Flori da maritime hammock. Plant Ecol Divers 2:77 83. Hanula JL, Mayfield AE, Fraedrich SW Rabaglia R J. 2008. Biology and host associations of redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae), exotic vector of laurel wilt killing redbay trees in the southea stern United States. J Harrington TC, Yun HY, Lu SS, Goto H, Aghayeva DN, Fraedrich SW. 2011. Isolations from the redbay ambrosia beetle, Xyleborus glabratus confirm that the laurel wilt pathogen, Raffaelea lauricola originated in Asia. Mycologia 103:1028 1036. Harrington TC, Aghayeva DN, Fraedrich S W. 2010. New combinations in Raffaelea Ambrosiella, and Hyalorhinocladiella, and four new species from the redbay ambrosia beetle, Xyleborus glabratus Mycotaxon 111:337 361. of Econ Entomol 101:1276 1286. Harrington TC, Fraedrich S W. 2010. Quantification of Propagules of the Lau rel Wilt Fungus and Other Mycangial Fungi from the Redbay Ambrosia Beetle, Xyleborus glabratus Phytopathology 100:1118 1123. Harrington TC, Fraedrich SW, Aghayeva DN. 2008. Raffaelea lauricola a new ambrosia beetle symbiont and pathogen on the Lauraceae. Mycotaxon 104:399 404. Hughes MA. 2013. The Evaluation of Natural Resistance to Laurel Wilt Disease in Redbay (Persea borbonia). (Doctoral dissertation). University of Florida, Gainesville, Florida. Hughes MA, Shin K, Eickwort J, Smith J A. 2012. First rep ort of laurel wilt disease caused by Raffaelea lauricola on silk bay in Florida. Plant Dis 96:910. Hughes M, Smith JA, Mayfield AE, Minno MC Shin K. 2011. First Report of Laurel Wilt Disease Caused by Raffaelea lauricola on Pondspice in Florida. Plant Dis 95:1588. Hughes M, Thomas J, Mayfield A E. 2009. Graphical Depiction of the Laurel Wilt Disease in Redbay. http://www.fs.fed.us/r8/foresthealth/laurelwilt/disease_cycle.shtml
75 Hulcr J, Dunn R R. 2011. The sudden emergence of pathogenicity in insect fungus s ymbioses i n nave forest ecosystems. Proc of Br R Soc 278:2866 2873. Inch SA, Ploetz R C. 2012. Impact of laurel wilt, caused by Raffaelea lauricola on xylem function in avocado, Persea americana For Pathol 42: 239 245. James TY, Kauff F, Schoch CL, Mat heny PB, Hofstetter V, Cox CJ, Celio G, Gueidan C, Fraker E, Miadlikowska J (70 co authors). 2006. Reconstructing the early evolution of Fungi using a six gene phylogeny. Nature 443: 818 822. Justesen AF Ridout CJ, Hovmoller M S. 2002. The recent history o f Puccinia striiformis f.sp. tritici in Denmark as revealed by disease incidence and AFLP markers. Plant Pathol 51:13 23. Kass RE, Raftery AE. 1995. Bayes factors. J Amer Statist Assoc 90:773 795. Kim KH Choi YJ, Seo ST, Shin H D. 2009. Raffaelea quercus m ongolicae sp. nov. associated with Platypus koryoensis on oak in Korea. Mycotaxon 110:189 197. Kubono T, Ito S. 2002. Raffaelea quercivora sp. Novo associated with mass mortality of Japanese oak, and the ambrosia beetle ( Platypus quercivorus ). Mycoscience 43: 255 260. Linaldeddu BT, Franceschini A, Alves A, Phillips A J. 2013. Diplodia quercivora sp. nov.: a new species of Diplodia on declining Quercus canariensis trees in Tunisia. Mycologia 105:1266 1274. Linaldeddu BT Hasnaoui F, Franceschini A. 2009. Fi rst report of Botryosphaeria corticola affecting Quercus afares and Q. canariensis in Tunisia. J Plant Pathol 91:234. Luque J, Girbal J. 1989. Dieback of cork oak ( Quercus suber ) in Catalonia (NE Spain) caused by Botryosphaeria stevensii Eur J For Path 19 :7 13, doi:10.1111/j.1439 0329.1989.tb00764.x Lynch SC Eskalen A Zambino P J Mayorquin J S Wang D H 2013. Identification and pathogenicity of Botryosphaeriaceae species associated with coast live oak ( Quercus agrifolia ) decline in southern California. M ycologia 105:125 140. Lynch SC, Eskalen A, Zambino PJ, Scott T. 2010. First report of Bot canker disease caused by Diplodia corticola on coast live oak ( Quercus agrifolia ) in California. Plant Dis 94:1510. Massoumi Alamouti S, Tsui CKM, Breuil C. 2009. Mu ltigene phylogeny of filamentous ambrosia fungi associated with a mbrosia and bark beetles. Mycol Res 113: 822 835.
76 Mayfield AEIII, Barnard EL, Smith JA, Bernick SC, Eickwort JM, Dreaden T J. 2008 a Effect of propiconazole on laurel wilt disease development i n redbay trees and on the pathogen in vitro. Urban Forestry 35:317 324. Mayfield AE Smith JA, Hughes M, Dreaden T J. 2008 b First Report of Laurel Wilt Disease Caused by a Raffaelea sp. o n Avocado in Florida. Plant Dis 92:976. Mayfield AE III. 2007. Laurel wilt: A serious threat to redbay and other related native plants. The Palmetto 24:8 11. Meglecz E, Costedoat C, Dubut V, Gilles A, Malausa T, Pech N, Martin J F. 2010. QDD: a user friendly program to select microsatellite markers and design primers from lar ge sequencing projects. Bioinformatics 26:403 404. N iogret J, Kendra PE, Epsky ND, Heath R R. 2011. Comparative analysis of terpenoid emissions from Florida host trees of the redbay ambrosia beetle, Xyleborus glabratus (Coleoptera: Curculioidae: Scolytinae) Fla Entomol 94: 1010 1017. O'Donnell K. 1993. Fusarium and its near relatives. In Reynolds R. Taylor JW. ed. The fungal holomorph: mitotic, meiotic and pleomorphic speciation in fungal systematics. Wallingford, United Kingdom: CBA International. p 225 23 3. Pe a JE, Crane JH, Capinera JL, Duncan RE, Kendra PE, Ploetz RC, McLean S, Brar G, Thomas MC, Cave R D. 2011. Chemical control of the redbay ambrosia beetle, Xyleborus glabratus and other Scolytinae ( Coleoptera: Curculionidae). Fla Entomol 94:892 896. Phillips AJL, Alves A, Abdollahzadeh J, Slippers B Wingfield MJ Crous PW 2013. The Botryosphaeriaceae: genera and spec ies known from culture. Stud in Mycol 76: 51 167. Phillips AJL Lop es J Abdollah zadeh J Bobev S, Alves A. 2012. Resolving the Diplodia complex on apple and other Rosaceae hosts. Persoonia 29:29 38. Ploetz RC, Hulcr J, Wingfield MJ, de Beer Z W. 2013. Destructive Tree Diseases Associated with Ambrosia and Bark Beetles: Black Swan Even ts in Tree Pathology? Plant Dis 97:856 872. Ploetz RC, P rez Martnez JM Smith JA, Hughes M, Dreaden TJ, Inch SA, Fu Y. 2012. Responses of avocado to laurel wilt, caused by Raffaelea lauricola Plant Pathol 61:801 808. Ploetz RC, Pena JE Smith JA, Dreaden TJ, Crane JH, Schubert T, Dixon W. 2011. Laurel Wilt, Caused by Raffaelea lauricola is Confirmed in Miami Dade County, Center of Florida's Commercia l Avocado Production. Plant Dis 95:1589. Posada D. 2008. jModelTest: phylogenetic model averaging. Mol Biol Evol 25:1253 1256.
77 Rabaglia RJ, Dole SA, Cognat A I. 2 006. Review of American Xyleborina (Coleoptera : Curculionidae : Scolytinae) occurring North of Mexic o, with an illustrated key. Ann Entomol Soc Am 99:1034 1056. Ragazzi A, Dellavalle Fedi I, Mesturino L. 1989. The oak decline: a new problem in Italy. Eur J For Pathol 19: 105 110, doi:10.1111/j.1439 0329.1989.tb00240.x Rambaut A, Drummond AJ. 2007. Tracer 1.4. Available at http://beast.bio.ed.ac.uk/Tracer Riggins JJ, Hughes M, Smith JA, Mayfield AEI Layton B, Balbalian C, Campbell R. 2010. First occurrence of laurel wilt disease caused by Raffaelea lauricola on redbay trees in Mississippi. Plant Dis 94:634. Ronquist F, Teslenko M, van der Mark P, Ayres D, Darling A, Hhna S, Larget B, Liu L, Suchard MA, Huelsenbeck JP. 2012. MrBayes 3.2: efficient Bayesian p hylogenetic inference and model choice across a large model space. Syst Biol 61:539 542. Rooney AP, Ward T J. 2007. Highly efficient concerted evolution in the ribosomal DNA repeats: Total rDNA repeat variation revealed by whole genome sh otgun sequence data Genome Res 17:184 191. Sanchez ME, Venegas J, Romero MA, Phillips AJL, Trapero A. 2003. Botryosphaeria and related taxa causing oak canker in southwestern Spain. Plant Dis 87:1515 1521, doi:10.1094/PDIS.2003.87.12.1515 Schlotterer C. 2000. Evolutionary dynamics of microsatellite DNA. Chromosoma 109:365 371. Shimodaira H, Hasegawa M. 1999. Multiple comparisons of log likelihoods with applications to phylogenetic infere nce. Mol Biol Evol 16:1114 1116 Slippers B Boissin E Phillips AJ Groenewald JZ Lomb ard L Wingfield MJ Postma A Burgess T Crous PW 2013. Phylogenetic lineages in the Botryosphaeriales: a systematic and evolution ary framework. Stud Mycol. 76:31 49, doi: 10.3114/sim0020 Slippers B, Wingfield MJ. 2007. Botryosphaeriaceae as endophytes a nd latent pathogens of woody plants: diversity, ec ology and impact. Fungal Biol Rev 21: 90 106. Smith JA, Dreaden TJ, Mayfield AE, Boone A, Fraedrich SW, Bates C. 2009 a First Report of Laurel Wilt Disease Caused by Raffaelea lauricola on Sassafras in Flor ida and So uth Carolina. Plant Dis 93:1079.
78 Smith JA, Mount L, Mayfield AE, Bates CA, Lamborn WA, Fraedrich S W. 2009 b First Report of Laurel Wilt Disease Caused by Raffaelea lauricola on Camphor in Florida and Georgia. Plant Dis 93:198 198. Spen ce DJ, Smit h JA, Mayfield AEIII, Hulcr J, Ploetz R, Stelinski L L. 2013. Effect of chipping on emergence of the redbay ambrosia beetle (Coleoptera: Curculionidae: Scolytinae) and recovery of the laurel wilt pathogen from infested wood c hips. J Econ Entomol 10 :2093 100 Stamatakis A. 2006. RAxML VI HPC: Maximum Likelihood based Phylogenetic Analyses with Thousands of Taxa and Mixed Models. Bioinformatics 22:2688 2690. Strimmer K, Rambaut A. 2002. Inferring confidence sets of possibly misspecified gene trees. Proc. R. So c. Lond. B 269:137 142. Surdick JA, Jenkins A M. 2009. Pondspice ( Litsea aestivalis ) population status and response to laurel wilt disease in northeast Florida. Florida Natural Areas Inventory, Tallahassee, Florida #025665:42. Swofford DL. 2003. PAUP* 4.0: phylogenetic analysis using parsimony (*and other methods). Sunderland, Massachusetts: Sinauer Associates. Thurston MI, Field D. 2005. Msatfinder: detection and characterization of microsatellites. Distributed by the authors at http://www.genomics.ceh.ac.u k/msatfinder/. date accessed December 2013. Urbez Torres JR, Peduto F, Rooney Lantham S, Gubler W D. 2010. First report of Diplodia corticola causing grapevine ( Vitis vinifera ) cankers and trunk cankers and dieback of canyon live oak ( Quercus chrysolepis ) i n California. Plant Dis 94: 785. rbez Torres JR, Adams P, Kamas J, Gubler W D. 2009. Identification, incidence, and pathogenicity of fungal species associated with grapevine dieback in Texas. Am J Enol Vitic 60: 497 507. USDA Forest Service. 201 4 Distribu tion of Counties with Laurel Wilt Disease* by year of Initial Detection. http://www.fs.fed.us/r8/foresthealth/laurelwilt/dist_map.shtml Date accessed (Febuary 2014). Vilgalys R, Hester M. 1990. Rapid genetic identification and mapping of enzymatically amp lified ribosomal DNA from several Cryptococcus species. J Bacteriol 172:4238 4246. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Inis MA, Gelfand DH, Sninsky JJ, White TJ eds. PCR protocols: a guide to methods and application s. San Diego: Academic Press. p 315 322.
79 Wood S L. 1982. The bark and ambrosia beetles of north and Central America (Coleoptera: Scolytidae), a taxonomic monograph. Great Basin Nat. Mem. No. 6. Wu DY, Ugozzoli L Pal BK, Qian J, Wallace R B. 1991. The Effect of Temperature and Oligonucleotide Primer Length on the Specificity and Efficiency of Amplification by the Polymeras e Chain Reaction. DNA Cell Biol 10:233 238. You FM, Huo N, Gu YQ, Luo M C Ma Y, Ha ne D, Lazo GR, Dvorak J, Anderson O D. 2008. BatchPrimer3: a high throughput web application for PCR and sequencing primer design. BMC Bioinformatics 9:253. Zipfel RD, de Beer ZW, Jacobs K, Wingfield BD, Wingfield MJ. 2006. Multi gene phylogenies define Cer atocystiopsis and Grosmannia distinct from Ophiostoma Stud Mycol 55:75 97.
80 BIOGRAPHICAL SKETCH Tyler Dreaden grew up in Crestview, a town in the western Panhandle of Florida. It is here that Tyler discovered his love of the outdoors and became interes ted in forestry. Tyler began his college career attending a then local community college, Northwest Florida State College, in Niceville, Florida where he r eceived his A ssociate of A rts From there, he transferred to the University of Flori da where he gr aduated with his B achelor of S cience in forest resources and conservation Forest Health class that sparked his interest in forest pathology and to make the decision to obtain a Master of S cience degree in forest resources and conse rvation, with an emphasis on forest pathology from the University of Florida. During the course of this degree with Dr. Jason Smith he found he enjoyed research and decided to continue his education and pursue a Doctor of P hilosophy in forest resources and conservation with a minor in plant pathology
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EAKKVVJYM_LT37M6 INGEST_TIME 2014-10-03T22:32:52Z PACKAGE UFE0046535_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC