This item is only available as the following downloads:
1 THE PATHOGENICITY OF DIPLODIA CORTICOLA AND DIPLODIA QUERCIVORA ON OAK SPECIES OF THE SOUTHEASTERN COASTAL PLAIN: A HOST RANGE STUDY By SONJA MULLERIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Sonja Mullerin
3 To Wendy and Dave Best Buds
4 A CKNOWLEDGMENTS This resear ch was designed by Assistant Professor Jason A. Smith, School of Forest Resources and Conse r vation, University of Florida and funded in part by the SFRC, as well as the United States Department of Agriculture, U.S. Forest Service, through G rant N o. 11DG11083150022. Molecular identification of the initial isolates was done by Tyler Dreaden, who also developed primers to streamline detect ion of the se isolate s Mr. Adam Black performed the bulk of data collection. Dr. James Colee provided his expert advice on statistical analysis and interpretation of results.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................6 LIST OF FIGURES .........................................................................................................................7 ABSTRACT .....................................................................................................................................8 CHAPTER 1 INTRODUCTION ....................................................................................................................9 2 LITERATURE REVIEW .......................................................................................................12 An Emergent Pathogen on Oaks and Grapevines ...................................................................12 Reported Symptoms of D. corticola Infection in Oak ............................................................14 Mechanism of Host Death. .....................................................................................................15 Fungal Morphology. ...............................................................................................................16 3 METHODS AND MATERIALS ...........................................................................................18 Isolation and Characterization of Fungal Strains P and J .......................................................18 Plant Materials ........................................................................................................................19 2011 Study .......................................................................................................................19 2012 Study .......................................................................................................................20 Pathogenicity Tests .................................................................................................................21 Experimental Design ..............................................................................................................23 Statistical Analysis ..................................................................................................................23 4 RESULTS ...............................................................................................................................25 Differences in Wound Response between Infected Trees and Controls. ................................25 Relative Susceptibility of Oak Species to the Two Isolates ...................................................26 Categorical Variables of Disease Assessment (Visual Observations) ............................26 Results of Statistical Analysis .........................................................................................27 Relationship between lesion length and stem diameter in 2012 ..............................27 Relationship between oak section and lesion length ................................................28 Statistical significance of the pathogenicity tests .....................................................28 5 DISCUSSION .........................................................................................................................38 APPENDIX: RESULTS FROM STATISTIC S PROGRAM ........................................................43 LIST OF REFERENCES ...............................................................................................................48 BIOGRAPHICAL SKETCH .........................................................................................................52
6 LIST OF TABLES Table page 41 Frequencies of signs and symptoms .................................................................................36 42 Species, showing section ...................................................................................................37
7 LIST OF FIGURES Figure page 11 Live oak ( Q. virginiana) showing dieback. .......................................................................11 12 Cankers on infected live oak branch ..................................................................................11 31 Isolates J and P in culture. ..................................................................................................24 41 Q MU J1 1: Pycnidia around flap .......................................................................................31 42 QMUJ1 1 Phloem necrosis ..............................................................................................31 43 QMUJ1 1. Xylem streaking ...........................................................................................32 44 Negative correlation between canker length and stem diameter in 2012 ..........................32 45 Mean canker length for isolates P and J, by species, 2012. ...............................................33 46 Mean lesion volume for isolates P and J, by species, 2012 ...............................................33 47 Mean girdling, in cm, for isolates P and J by species, 2012. ............................................34 48 Mean canker length for isolates P and J, by species, 2011 ................................................34 49 Mean girdling for isolates P and J, by species, 2011 ........................................................35 410 Mean lesion volume for isolates P and J, by species, 2011 ...............................................35 A 1 2012, species*isolate interaction using mean canker length as response variable, LS Means Differences Tukeys HSD ......................................................................................43 A 2 2012: Lesion volume LS M Differences, Tukeys HSD, for species ...............................44 A 3 2012: Mean girdling, LS M Differences Tukey s HSD, for species. ................................45 A 4 2011: Canker vertical measurement, LS M Differences, Tukeys HSD, for species. .......46 A 5 2011: Mean girdling, LSM Differences Tukeys HSD, for species. ...............................46 A 6 2011: L esion volume LS M Differences Tukeys HSD, for s pecies ...............................46 A 7 2011: L esion volume LS M Differences, Tukeys HSD, for isolate ................................47
8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE PATHOGENICITY OF DIPLODIA CORTICOLA AND DIPLODIA QUERCIVORA ON OAK SPECIES OF THE SOUTHEASTERN COASTAL P LAIN: A HOST RANGE STUDY By Sonja Mullerin August 2013 Chair: Jason Andrew Smith Major: Forest Resources and Conservation A host range study was conducted in 2011 and 2012 on 29 species of oaks, along with four cultivars and two seed sources of live oak ( Quercus virginiana Mill.) commonly grown on the Southeastern Coastal Plain of the United States to assess the pathogenicity of two p resumed strains of the ascomycete Diplodia corticola T wo isolates with 99% and 97% homology with the ITS region of D corticola dubbed J and P have been discovered in Florida which vary by 10 nucleotides in th e ITS region. Inoculations were performed on the stem and Kochs postulates completed All controls developed callus and healed. Only one of the 382 t rees inoculated with either pathogen form ed callus, in contrast, and none healed. All inoculated trees developed long vascular necroses which also girdled the stem to varying degrees most exhibit ed stem bleeding and pycnidia and some d ied Stem diameter w as negatively correlated with lesion length, and red oaks (section Lobatae) had both significantly longer lesions and more girdling than white oaks (section Quercus) During preparation of this thesis, a new species, Diplodia quercivora was identified, w ith which isolate P has 99% homology in the ITS and EF regions. R esults support a conclusion that both fungal species inhibit wound repair in oaks
9 CHAPTER 1 INTRODUCTION In September 2010, the University of Florida Forest Pathology Laboratory received reports of dieback and cankers1 on small branches of landscape live oaks ( Quercus virginiana Mill. ) on a horse farm in Marion County, Florida. Isolations from the cankers led to identification of the pathogen Diplodia corticola A.J.L. Phillips, Alves & Luque, sp. nov., by comparison of the internal transcribed spacer (ITS nucleotide sequences with sequences deposited in Genbank ( 16). A review of the literature revealed that D. corticola ha d been identified as the primary cause of a canker disease causing serious decline of the cork oak forests ( Quer cus suber) in Mediterranean countries ( 2). In December 2010, a first report appeared in Plant Disease documenting D. corticola in California, and establishing the pathogen as a factor in the widespread mortality seen since 2002 in coast live oaks, Quercus agrifolia Nee, in San Diego County ( 30). This report prompted a statewide survey in Florida for declining live oaks, leading to i dentification of D. corticola from several regions of the state ( Figures 11 and 12) Unlike the situation in California di sease incidence in Florida has been almost entirely limited to landscape planting s It has not been documented in forest s The symptoms reported in response to the survey have fallen into two general categories. The first is characterized by dieback of outer twigs throughout the crown, giving the canopy a sparse appearance. The d ead twigs frequently bear cankers as well as black pycnidia which are flask shaped fruiting bodies of the fungus The second is distinguished by localized flagging of 1 Cankers are localized wounds or dead areas in t he bark of the stem or twigs of woody or other plants that are often sunken beneath the surface of the bark. Dieback is the [p]rogressive death of shoots, branches, and roots, generally starting at the tip ( 1).
10 large b ranches randomly distributed in the crown, also frequently bearing canker s and pycnidia with or without twig dieback. Molecular studies of fungi cultured from these affected trees in Florida revealed the existence of two pr esumed strains of Diplodia cor ticola referred to here as P and J, which differ at 10 nucleotide sites in the ITS region. P was isolated from a declining live oak in a grocery store parking lot in Gainesville, and J from a condominium complex in Jacksonville.2 I solate P corr esponded with the first syndrome described above and i solate J with the second Since the first submission of this thesis, the ITS sequence of a species newly identified on three oak species in Tunisia has been deposited in GenBank, with which the ITS an d EFregions of isolate P have 99% homology ( Dreaden, T., personal communication ) Th e new species has been named Diplodia quercivora ( 24 ). It will continue to be referred to as Diplodia corticola i solate P in this thesis. The research objectives of this study w ere to assess the susceptibility of various members of the family Fagaceae to these two emergent and clo sely related pathogens, P and J, by completing Kochs postulates. The host tree species selected are native to or commonly cultivated on the coa stal plain of the Southeastern United States. A first study was conducted in 2011 on eleven oak species native to Florida including a Florida seed source of live oak ( Q. virginiana Mill. ) An expanded study was conducted in 201213 on an additional 17 s pecies of oaks commonly grown in the Southeast ern U.S., including an Asian species ( Quercus acutissima ); f our cultivars of live oak; a live oak grown from a Louisiana seed source; and Castanea pumila, a nonoak member of family Fagaceae. 2 P is culture no. PL 1345, and J is culture no. PL 1325, in the UF Forest Pathology Lab.
11 Figure 11. Live oak ( Q. virginiana) showing dieback. Photo: Jason Smith, Univ. of Florida Figure 12. Cankers on infected live oak branch. Photo: Tyler Dreaden, University of Florida
12 CHAPTER 2 LITERATURE REVIEW A n Emergent Pathogen on Oaks and Grapevines Since the 1960s, the ancient cork oak forests in Spain and Portugal have been declin ing, a condition which accelerated dramatically in the 1980 s ( 27 2) I n 1989, t he fungal pathogen believed to be the primary cause was identified as Botryosphaeria stevensii Shoemaker ( 27, 36). I n 2004, th e causative organism was re evaluated and recognized as a new species, Botryosphaeria corticola A.J.L. Phillips, Alves & Luque, sp. nov. ( 2). The name has since been changed to Diplodia corticola referring to its anamorph (asexual) stage ( 17) Renaming is due to reorganization of the genus Botryosphaeria to exclude species having pigmented conidia at maturity The pigmented species were thus p laced in the genera Di plodia and Neofusicoccum ( 12, 13). Since its identification as the cause of cork oak mortality (2) D. corticola has also been implicated in the die off of other species of European oaks, a malady which had been referred to as oak decline ( 36, 8, 25). In 2010, Diplodia corticola was then identified as the cause of rapidly spreading dieback and occasional death of live oaks in California and Florida ( Q. agrifolia Nee; Q. chrysolepis Liebm.; Q. virginiana Mill.), as well as grapevine mortality in Texas, California, and Mexico ( 16, 4 2, 43, 30, 29, 9). A more in depth examination of the literature has revealed that the pathogen might, in fact, have originated in North America and been identified as the cause of decline of chestnut oak ( Q. prinus ) and othe r oak species in Pennsylvania in 1912 ( 4 5, 22, 40, 6). Although referred to as Bot canker, that term w as already commonly used for diseases caused by other members of the family Botryosphaeriaceae which are generalist pathogens on
13 hundreds of plant gene ra ( 7, 14, 21). Diplodia corticola in contrast, has been found only on oak and grapevine.1 The Botryosphaeria ceae a re known to enter plants through wounds including leaf scars, but can also enter through stomata and lenticels open for gas exchange ( 4 6, 7) The se fungi are spread by air, water splash, or contaminated pruning tools ( 7 ). The traditional --and environmentally sustainable -enterprise of removing cork from oak trees2 causes mild injury, thus providing entry for Diplodia corticola into Q ue rcus suber trees ( 18 ). In addition, t he Botryosphaeriaceae often live harmlessly as endophytes within plant s but are believed to becom e pathogenic when the plant is stressed by environmental factors such as drought, heat, freezing, herbicide use, or soil compaction ( 7, 25).3 In California, Diplodia corticola has been implicated in the death of tens of thousands of acres of coast and canyon live oaks since 2002, commonly occurring with other, less aggressive pathogens ( 29). T he re is equivocal evidence that the fungus may be vectored in California by insects. D. corticola was isolated from coast live oak trees, both living and dead, that had been colonized by bark and ambrosia beetles in Marin County following artificial inoculations with Phytophthora ramorum ( 32). The exotic gold spotted oak borer, Agrilus auroguttatus Schaeffer 1From GenBank sequences, Alves et al (2 ) identified D. corticola on two other hosts: Eastern Redbud, Cercis canadensis L., in the District of Columbia, and Tsuga sp. in North Carolina. Both sequences were deposited by Jacobs in 1998. Cercis canadensis is listed as a host of D. corticola also in Table 2 of Slippers et al (41 ) (GenBank accession number AF027752), but Tsuga (GenBank accession number AF02755) is not. The USDA Fungus Host database inclu des C. c anadensis as a host (1 7 ). Botryosphaeria stevensii (anamorph: Diplodia mutila) has now been determined not to infect Quercus .. Thus, prior reports of B. stevensii infecting oaks must be interpreted as referring to Diplodia corticola instead (L uque, J., personal communication). Affected European oak species from which B. stevensii was identified, which are not listed as hosts in GenBank because the identifications predated molecular characterization, are: Q. cerris L., Q. ilex Q. petraea (Matt.) Liebl., Q. pubescens Willd. Q. robur L., and Q. trojana Webb. See citations in (2 8 ). 2 Cork harvesting, http://www.youtube.com/watch?v=ztr RP0XYd8 (accessed 6 2 7 13). 3 Luque et al. ( 2 6 ) n ote that it may be the other way round: that infection by a pathogen usually negatively affects water transport in the tree, in the absence of drought, as well as that the symptoms of pathogenic infection resemble those of drought stress in uninfected tre es.
14 (Coleoptera: Buprestidae) (GSOB) has also been associated with the colonization of coast live oak by D. corticola ; however, trees not infested by the GSOB are also infected by D. corticola and declining. A ttack by the GSOB increases drought stress in the trees, possibly awakening the latent fungal pathogen already in residence ( 2 9, 30, 10). Lastly, a report of lesions and branch dieback in Virginia in 1959 in Q. prinus and Q. alba found infections by the fungus Dothiorella quercina (Cke. & Ell) Sacc often occurr ing in association with scale insects ( 6). Vajna in 1986, concluded D. quercina was Diplodia corticola ( 45 ). Surveys in Florida have most often found symptoma tic live oaks ( Q. virginiana) in landscape plantings and most reports have concerned the cultivar Cathedral O aks in landscape settings, unlike those in natural forests, are exposed to various stresses such as transplanting, pruning, and herbicides M ost of Florida also experienced severe drought in 2010, the same year the decline was first reported. Reported Symptoms of D. corticola Infection in Oak In the cork oak in Europe ( Q. suber ), discolored bark is the first disease symptom, appearing 23 mo nths after cork removal ( 8) In the early period following infection t here may be foliar yellowing and discoloration, as well as epicormic shoot development (shoots which develop from buds under the bark) ( 8). After approximately six months, the tree develops necroses in the cambium of varying lengths (cankers) on branches and the trunk, and bark comes off easily. Wilting follows due to loss of vascular function and pycnidia are observed in and around necrotic areas. Death of the tree, if it occurs, usually comes between one and three years after symptoms are first noted ( 27). Infection by D. corticola in mature live oak trees in California and Florida is not visible in the early stages, since it is not preceded by cork removal. In Florida, c l umps of dead branches appear randomly distributed in the crown ( 16). Cankers are evident on branches and cutting into
15 the branch at these points reveals sapwood streaking and phloem necroses (dark brown or black discoloration) ( Figure 12) A lthough trunks are frequently cankered in California ( 29 ) only once have cankers been found on a trunk in Florida Bleeding sap and bark cracking are observed with the branch cankers, along with pycnidia which erupt through the bark or cankers. In artificial inoc ulations of coast live oak seedling s with D. corticola in California, Lynch et al ( 29) also reported epicormic shoots, and leaf desiccation. They report that D. corticola killed the xylem in advance of the living phloem and moved into the taproot on 70% of inoculated seedlings. Luque et al (28) conducted a study of the effects of 38 different fungal species isolated from cork oaks in Catalonia (Northeastern Spain) on Q. suber by inoculating stems, leaves, and roots under two watering regimes, one ade quate and the other water stressed Among the inocula was Botryosphaeria stevensii Shoemaker ( Diplodia corticola on oak) In contrast to the later findings of Lynch et al. ( 29 ), Luque et al ( 2 6) found that D. corticola was not pathogenic on roots or lea ves On stems, however, D. corticola was the most virulent of the pathogens tested not only inducing l ong cankers and vascular lesions, but causing death of almost all plants (one year old cork oak seedlings) within two weeks of inoculation. N otwithstan ding reports such as those cited above --that plants become more susceptible to infections by Botryosphaeriaceae during drought --th e Luque group determined that the length of lesions caused by the D. corticola w as independent of water stress in the plant. M echanism of Host Death Cork oaks and holm oaks ( Q. ilex ) artificially infected with D. corticola by Linaldeddu et al ( 25) exhibited significant reduction in net photosynthetic rate and stomatal conductance. Because the impact of the infection on gas exchange rates was independent of stem lesion length in both species, the authors suggested the cause was diffusible toxins.
16 D iplodia corticola produce s several phytotoxin s as secondary metabolites. The most important appears to be the compound diplopyrone which is toxic to cork oak at concentrations from 0.01 to 0.1 mg/mL and causes collapse of internal stem tissue on tomato cuttings at 0.1 to 0.2 mg/mL ( 31). Other pathogenic Botryosphaeria species also produce plant toxins.4 It is unclear from these r eports which of the other metabolites are phytotoxins or whether they become phytotoxic act ing synergistically. Fungal Morphology. Identification of D. corticola based on morphology is problematic as is the Diplodia genus generally. There is a substa ntial history of misidentification. Different lighting conditions or temperatures have different effects on development and pathogenicity ( 38) and members of the Botryosphaeriaceae are notoriously similar one to another, with closely related species often inhabiting the same plant (4 2) In some reports, for example, in culture on potato dextrose agar Diplodia corticola when viewed from above, initially appears fluffy white, turning to dark gray after five to seven days, w hile the underside turns olive green, t hen black ( 16, 2). Th e description of abundant gray mycelium with a diurnal zonation that gradually became dark olivaceous was assigned to D. corticolas cousin Neofusicoccum mediterraneaum however, where both pathogenic species were isolated f rom the same sample of grapevines in Spain while Diplodia corticola wa s described as whitish, dense, aerial mycelium [which] remained white up to 10 days on PDA and darkened to gray thereafter ( 35). 4Botryosphaeria obtusa, for example which causes black rot on apple fruit and frogeye leaf spot on apple and other plants produces a veritable smorgasbord of phytotoxins: mullein, tyrosol, 4 hydroxymellein, 5hydroxymellein, 7hydroxymellein, 4,7 dihy droxymellein, and 4 hydroxybenzaldehyde T his production is puzzling on the fruit, which is undefended tissue. B. obtusa also does not require a wound for entry. It enters through lenticels and stomata. ( 4 6 ,1 5 ).
17 Identification of particular structures, generally b y m icroscop y, although also a challenge provides an additional tool to differentiate these species Diplodia corticola is an ascomycete with both sexual and asexual spore stages. The sexual stage, or teleomorph, is rarely found in nature, and in many cas es the teleomorphic characters are too similar to distinguish among species in the Botryosphaeriaceae ( 2 3). Thus, the anamorphic (asexual stage) features, conidia and conidiophores, are used, since size, shape, color, wall thickness, and septation in thes e structures are often distinct. They are described for D. corticola as follows: Conidiomata pycnidial, separate or aggregated, globose, dark brown to black, immersed, multilocular, thick walled; outer wall layers thick walled textura angularis inner lay ers thin walled, hyali ne; ostiole central, papillate. Conidiophores reduced to conidiogenous cells. Conidiogenous cells holoblastic, integrated or discrete, determinate, cylindrical, hyaline, smooth, forming a single, apical conidium, proliferating percur rently to form one or two indistinct annelations, or proliferating at the same level to form periclinal thickenings. Conidia smooth, unicellular, cylindric with broadly rounded ends, some with a large central guttule, smooth, with a thick glassy wall that remains hyaline even after the conidia have been discharged from the pycnidium, rarely becoming light brown and one or two septate after discharge (23.5 )2635( 41) (10)11 17( 18.5) m, average of 250 conidia = 29.9 2.6 13.6 1.4 m. Length/widt h ratios in the range of (1.6)2.22.3( 3.1) with a mean and standard deviation of 2.2 0.3. ( 11). The length width ratio of approximately 2.2 of the conidia of Diplodia corticola is what primarily distinguished it from Neofusicoccum mediterraneum the conidia of which have a length width ratio of about 4.0, in the example above (35).
18 CHAPTER 3 METHODS AND MATERIAL S The impetus for th is host range study was the discovery of two strains of Diplodia corticola P and J, which were isolated from cankers in live oaks characterized molecularly and found to be somewhat different The study, which took place in 2011 and 2012, was to assess the e ffects of the two isolates on different oak species native to, or grown commonly in, the Southeast ern U.S. The e xperiment involved inoculation of (usually) 12 trees per species with one of the two fungal isolates P and J, in the greenhouse and isolation of the fungal strains from the plants after disease developed, pursuant to the design described below Isolation and Characterization of Fung al Strains P and J Samples approximately 4 mm square in size were taken with sterile razor blades f rom the margins of cankers on small branches from declining live oaks ( Quercus virginiana Mill.) in Jacksonville, Florida, betwe en August 8 and 12, 2011. Samples were similarly taken from declining live oaks growing in a grocery store parking lot in Gainesville, F lorida first in 2010 and again for th is host range study on Sept. 20, 2011. The se samples were surfacesterilized in a 50% solution of 8.25% sodium hypochlorite (regular bleach) for 30 sec, the drops shaken off, and t hree to four placed per plate on acidified potato dextrose agar (APDA) Cultures were kept in the dark at approximately 25 C for 710 days, after which ti me each morphologically different fungus on the plates was separately subcultured on APDA. A tentative i dentification as D iplodia corticola was made of those subcultur es matching the published description of this organism in culture Salient features i nclude d initially white m ycelia which turned dark greyoliva ceous on top, with the underside black after 5 10 days ( 2) Molecular c onfirmation that both were D. corticola was obtained by extracting DNA from m ycelia with use of the DNeasy Plant Mini Kit ( Qiagen). PCR (Polymerase Chain Reaction)
19 was run using primers ITS 1 and ITS 4. After confirming the presence of DNA amplicons at approximately 650 bp, 5 L of PCR product was purified with 2 L of Exosap (Affymetrix) and sequenced at the University of F loridas Interdisciplinary Center for Biotechnology Research. The ITS sequences were aligned and edited using Geneious Pro 5.6.6 software. A BLAST n search revealed that the nucleotide sequences from two isolates matched sequences of Diplodia corticola deposited in GenBank to different degrees. T he ITS sequences of these isolates had 99% and 100% homology with the ITS region of Diplodia corticola isolate CBS 112074 (GenBank Accession No. AY268421) which is the type specimen (2) The t ( beta tubulin) regions of the two isolates had 99 and 97% homology with the t sequence s of D. corticola strain UCD2397TX ( Gen Bank Accession No. GU294724), and D. corticola strain CBS 112550 (Gen Bank Accession No. AY259097.1) respectively The isolates were then dubbed J and P. Aligned with each other, the ir ITS regions differ at 10 nucleotide positions. These subcultures, of J and P, were stored on slants of APDA at 5 C New subcultures were prepared from the m and used as the inoculum source in the 2011 study. Despite the similarity of their DNA sequences, t here are differences between the two isolates i n pure culture when they are grown in the dark at 25 C J fills the plate rapidly and begins piling up aerial mycelia around the edges W hile P also has an i nitial rapid growth rate, it drastically slow s before filling the plate entirely. The colony margin is wavy and clearly defined, usually keeping a space between it and the plate edge, and the colony is appressed of dense r texture and darkening much sooner than J (F igures 31 A F ) Plant Materials 2011 S tudy In October 2011, saplings of indeterminate ages in each of 11 species of oaks from commercial seed sources were acquired from two nurseries near Gainesville and Orlando
20 Florida These species, fo llowed by abbreviations used to designate them, were: Quercus virginiana from a Florida source (QVA); Quercus shumardii (QSH); Quercus alba (QAL); Quercus michauxii (QMI); Quercus laurifolia (QLA); Quercus austrina (QAU); Quercus falcata (QFA); Qu e rcus ma rgaretta (QMA); Quercus geminata (QGE); Quercus chapmanii (QCH); and Quer c us myrtifolia (QMY). The trees were acclimated over several weeks, then arranged in two randomized north/south blocks on the floor of a campus greenhouse at the University of Florida in Gaines ville, FL (lat. 29.6514, long. 82.3250, 35.052 a.s.l.) Each block contained eight trees per species, for a total of 16, with the exception of Quercus geminat a (QGE), Quercus margaretta (QMA), and Quercus falcata ( QFA) due to limited avail ability. The re w ere eight QGE, QMA, and QFA trees total and those were distributed randomly within the first block only. All trees were in 3 gal pots, except for QFA, which w as in 1gal pots. Trunk diameter ranged from 7 mm in the 1gal pots to 3 cm in 3gal pots at the inoculation site, which was approximately 5 cm above soil level, except for QFA, which had the smallest stems, so was inoculated approximately 2 cm above soil level. Temperatures ranged from 19 C to 29 C throughout the experiment, and the trees were irrigated daily to field capacity. 2012 S tudy The second study began in August 2012 with 19 different species from the family Fagaceae from commercial seed sources, including a live oak Q. virginiana, from Louisiana, along with four Q. vir giniana cultivars. These were: Castanea pumila (CPU); Quercus acutissima (QAC); Quercus bicolor (QBI); Quercus coccinea (QCO); Quercus hemisphaerica (QHE); Quercus incana (QIN); Quercus inopina (QIO); Quercus palustris (QPA); Quercus laevis (QLV); Quercu s lyrata (QLY); Quercus macrocarpa (QMC); Quercus marilandica
21 (QML); Quercus muehlenbergii (QMU); Quercus nigra (QNI); Quercus nuttallii (QNU); Quercus phellos (QPH); Quercus pumila (QPU); Quercus rubra (QRU); Quercus virginiana (Louisiana seed source) (QV L); Q. virginiana Highrise (QVH); Q. virginiana Cathedral (QVC); Q. virginiana Augustina (QVU); and Q. virginiana Sky Climber (QVS). Again the trees were arranged in two blocks of eight trees per species, except that CPU was short two trees in the second block; QIO was short two trees in each block; QNI was short one tree in the first block; and QNU was short one tree in the second block. The trees were completely randomized within each block. All experimental parameters in 2012 were as in 2011, e xcept that Q. inopina and Q. incana, the smallest trees, were inoculated in the same stem location as Q. falcat a. S ome of the Q. virginiana cultivars and Castanea pumila, and all the Q. laevis, were in Accelerator pots (Nursery Supplies, Inc.) which ha ve holes on the sides, making their rootballs more susceptible to drying than the other pots. These are shown in Table 41. Q. pumila spreads via rhizomes (stems branching from shallow, laterally growing roots) so had multiple smalldiameter s tems rather than a single trunk. T he Q. inopina and Q. incana trees also were small and shrubby, several having multiple stems Pathogenicity Tests In each block, three trees per species were inoculated with isolate J and three with isolate P ,1 with two trees serv ing as controls inoculated with sterile APDA. In the 2011 study, inoculations of the first block, 79 trees, took place Oct. 11, 2011, and inoculations of the second block, also 79 trees, took place Oct. 27, 2011. In the 2012 study, inoculations of the fi rst block, 181 trees, took place Sept. 21, 2012, and of the second block, 179 trees, on Oct. 16, 2012. 1 Exceptions were CPU, which had only t wo of each isolate in the 2nd block; QIO, which had only two of each isolate in each block; QNU, which was short a J in the second block; and QNI, short a P in the first block.
22 In both the 2011 and 2012 studies the inoculum consisted of a n approximately 4 mm2 plug of APDA colonized by mycelia, removed with a razor blade from a two week old subculture of P or J The plug was placed with the mycelium side facing inward inside a flap approximately 2.5 cm long cut in the stem cambial tissue with a sterile single edged blade approximately 5 cm up from soil level. The flap was t hen pressed back against the wood with the inoculum inside, and wrapped with P arafilm to bind it tightly against the stem. Intermediate e xternal disease assessments were done in the 2012 study, on Nov. 12, 2012, for block 1 and Dec. 6, 2012, for block 2, noting the presence of bark cankers fungal fruiting bodies, stem bleeding, whether necrosis extended beyond the flap, and other signs of distress such as branch flagging, leaf desiccation, epicormic shoots or death. Cankers were measured on the bark ve rtically, as well as horizontally (by estimating the percentage of stem girdling). I ntermediate assessments were not done in the 2011study. Final destructive sampling for the 2011 study took place approximately three months after inoculations, and, for the 2012 study, four months after inoculations. Prior to the final assessments, external signs and symptoms as described above, were recorded a nd the internal extent of phloem necrosis was measured In the destructive sampling itself several measures of disease were employed. These included length of necrosis in phloem; percent girdling; presence of pycnidia; stem bleeding; whether or not the flap was dead; and whether necrosis extended beyond the flap. In 2012 xylem discoloration was measured separatel y from phloem necrosis (In 2011, xylem discoloration and phloem necros i s were not distinguished.) Dead trees were retained in the dataset, with an artificial value assigned for lesion length equal to 110% of the longest lesion measured for that speciesisolate pair on a living tree in that block ( 19) Girdling was naturally at 100% in the dead trees.
23 To complete Kochs postulates one symptomatic tree per isolate/species/block was chosen for sampling, to reisolate the pathogen. Sampling was done at ca nker margins as far away from the inoculation site as possible to avoid any inoculum that might have remained within the flap without colonizing the host tissue. No isolations were made from controls, nor from any trees prior to inoculation. The tissue samples were cultured and the morphology in cultures was examined visually to distinguish between isolates P and J DNA extraction and PCR were also performed via the methods described in the first section of this chapter, Isolation and Characterization of Fungal Strains, with an important difference B y the time these isolations were made, Ph.D. student Tyler Dreaden had developed forward primers Dcort 1 and Dcort1A and reverse primer Dcort3, which selectively amplified the J and P strains of Di plodia corticola (and did not amplify other fungi) The amplicons were then trimmed with a restriction enzyme, Msel, which produced different sized restriction products, so they could be differentiated (Dreaden, T., personal communication). Experimental D esign A complete randomized block design was used. There were two blocks in 2011 and two blocks in 2012, a block us ually consisting of eight trees of each species tested in that year T he exceptions were QGE, QMA, and QFA in 2011, which had only eight trees total so four per block; and the 2012 exceptions noted in n.1, above Because equal numbers were not used for all species, the design was not balanced. Statistical Analysis The purpose of the experiment was to evaluate the effects of the two Dipl odia corticola isolates P and J on different oak species. The null hypotheses were that there would be no significant difference in the mean responses of (1) any one tree species to the two fungal isolates; (2) any two tree species to the same fungal isol ate; (3) any one combination of species*isolate
24 with any other combination of species*isolate; and (4) any inoculated tree and a control of the same species. Statistical processing was performed via JMP 8 (SAS Institute, Cary, NC), and data from the two s tudy years were analyzed separately. Blocks were ignored, since the conditions between blocks were very similar. Figure 31. Isolates J (left) and P (right) in culture. A) 5 days old, top; B) 5 days old, bottom; C) 9 days old, top; D) 9 days old, bott om; E) 30 days old, top; F) 30 days old, bottom (appearing browner in photo than it was). Photos and cultures: Adam Black, UF
25 C HAPTER 4 R ESULTS D ifferences in Wound Response between I n fected T rees and C ontrols The re were significant differences in wound response between infected trees and controls. A ll controls formed callus, which is tissue formed by a tree during the healing process ( 1), ringing the wound. In contrast, callus did not form i n 381 of 382 trees inoculated with either fungal isolat e The one inoculated tree that did form callus was a Q. michauxii inoculated wit h isolate P in block 1, in 2011; and the fungus was reisolated from the callus. Fishers Exact Test yielded P <.0001 for each study regardless of isolate, meaning that the re was a statistically significant difference in the ability of trees of all species inoculated with either D. corticola or D. quercivora to heal wounds, compared to controls There were also no lesions on controls, while lesions were produced both vertica lly and horizontally (around the stem between 15 and 100% ) on every tree inoculated with the pathogen. T issue created by the wound in the bark ( the flap) remained alive and green in all controls except in the live oak cultivar Highrise in 2012. In both controls of that cultivar in block 1, the flap died, but the tissue underneath healed. In contrast, in every tree inoculated with either strain of D. corticola, save f ive individuals thus 377 trees -the flap died in its entirety except for two cases where part of it was still alive at the end This difference compared to controls was again significant for both studies, with P < .0001. The exceptions were three Q. shumardii inoculated with J in block 1; a Q. virginiana inoculated with J in block 2 (b oth 2011) ; and a Q. virginiana Augustina inoculated with J in 2012 ( Table 4 1 ) The dead flap on inoculated trees was black the necrosis almost always extending beyond the wound in both directions. The dead flap was also the site where pycnidia were m ost often found.
26 St em bleeding (sap exudation) occurred on none of the controls with the exception of the two Quercus l yrat a controls in b lock 1 in 2012, which bled briefly, then healed Stem bleeding-of a translucent amber colored exudate -did occur profusely, on all but one tree inoculated with isolate P and one with J, in 2012, and all but 18 trees inoculated in 2011, as shown in Table 41. These results were again statistically significant, with no difference in the ability of J or P to induce s tem bleeding. Controls were not sampled for the presence of D. corticola, so it is unknown whether the organism could be found naturally in any of these species, as an endophyte. If present, it did not cause disease, even though these trees were wounded. Finally, fruiting bodies or mycelia of fungi which were not Diplodia corticola appeared at the wound sites of several inoculated trees, but never on controls. Only in the case of Q. phellos was the fungus cultured and s equenced. It was bright orange with a rubbery texture initially, fading to salmon, then buff, which appear ed on two of the Q. phellos tree s inoculated with J, and four inoculated with P finally disappearing altogether. It was cultured subcultured, and sequenced. There were two products, having 100% homology with Pestalotiopsis photinia and an unidentified Fusarium species. Relative Susceptibility of Oak Species to the Two Isolates Categorical Variables of Disease Assessment (Visual Observations) All host species and cultivars challenged with either isolate of D. corticola developed stem lesions extending both upwards and downwards from the inoculation site Representative photographs a re presented in Figures 41 43. The specific strain of Diplodia corticola with which a tree was inoculated (P or J) was reisolated from one symptomatic tree per isolate/ species/block at the end of the experiment in both 2011 and 2012, with two exceptions : QLAJ which did not show up, and QMI J, for which a culture was not made.
27 In 2012, symptoms appear ed within two weeks of inoculation on most species. Certain northern species in the redoak group, such as Quercus palustris (pin oak) and Quercus rubra (red oak), exhibited rapidly progressing external symptoms. By the end of the study one red oak tree, a Quercus phellos (willow oak), had produced a 46cm long phloem necrosis with a 46cm long xylem discoloration, and another had a 49 cm long xylem discoloration (in connection with a 2.4cm phloem necrosis) both in reaction to isolate J. The mean canke r length of red oaks was significantly longer than it was in white oaks (more detail under Statistical Analysis, below). The frequency of disease indications at the final destructive sampling, along with mean lesion lengths and mean percent girdling (the metrics used for statistical analysis) for all species are set forth in Table 4 1. Stem bleeding cooccurred with flap death in 266 of 268 inoculated trees (99%), in 2012; the cooc currence of these symptoms was 96 of 114 (82.5%) in 2011. Phloem necros is extended beyond the flap in most cases. No foliar symptoms ( such as leaf yellowing or flagging) nor twig dieback or branch canker s were observed in any of the trees. Xylem streaking, measured only in 2012, was extensive in every tree much longer tha n phloem necrosis. In all cases, the extent of xylem discoloration above the inoculation site was at least equal to and often much longer than that below the site, since it stopped at the soil line (5 cm below the inoculation point ) S t aining was thick est at the inoculation site, tapering down to hair thin streaks (Figure 4 3). Results of Statistical Analysis Relationship between lesion length and stem diameter in 2012 In 2012, canker vertical measurement (CVM) combining results from both isolates, w as negatively correlated with stem diameter, with P < .0001. The statistical test employed was a one way ANOVA, confirmed via a multivariate analysis followed by Spearmans correlation.
28 The correlation is shown graphically in Figure 4 4. N o significant re lationship was found between these parameters in the 2011 results. Relationship between oak section and lesion length Section is a subgeneric taxonomic division. All but one of the oak species tested fell into Section Lobatae (red oaks) or Section Quercus (white oaks), the exception being Q. acutissima from Asia (which is in Section Cerris) ( Castanea pumila is in a different genus.) In the 2012 study, 12 red oak and nine white oak species were tested. T he differences in mean canker vertical m easurement (CVM) and mean girdling scaled by [ (%HC d) / ] between the red oaks (6.83 cm; 0.67 cm ) and white oaks (4.10 cm; 0.52 cm ) were both statistically significant, with P <.0001. In 2011, four red oak and seven white oak species were tested. In c ontrast with 2012, where the species were more northerly, the means in 2011, where the species were native to Florida, were almost the same (3.58 cm and 0.48 cm for red vs. 3.28 cm and 0.46 cm for white ) Table 4 2 show s the section to which each species belongs. Statistical significance of the pathogenicity tests Different response variables were developed and tested to try to remove the small stem effect from the data; however, at least in 2012, as discussed, th is effect wa s real: species with small stem diameter we re more susceptible to both pathogens not only becoming girdled more quickly, but having disproportionately longer cankers Most informative were simply the raw data mean girdling (the percentage of circumference covered by the lesion, al so referred to here as percent horizontal coverage or %HC), and mean canker vertical measurement (CVM) -which are graphed and reported first, below. Because the cankered area is a portion of a cylinder, a n approximat e lesion volume ( CVM %HC r2) (a volume again scal ed by ), where r is stem diameter and CVM is the length of the cylinder, was also used as a response variable and the outcome of the statistics program for it is reported for both 2012 and 2011.
29 This lesion volume is a n imperfect a pproximation, since it assumes that the lesion extends in all cases to the center of the stem. In 2012, using mean CVM as the response variable, there was a statistically significant ( species*isolate ) interaction, P < .0399. The statistics program output is presented in Figure A 1, showing where significant differences occurred via the different letters test (Least Square Means Differences, Tukeys HSD) A g raph showing the mean CVM for each isolate, across species, appear s as Figure 4 5. Using mea n girdling as the response variable did not produce a statistically significant (species*isolate) interaction, although the species effect was significant at P <.0001. The species effect means that the difference between mean girdling of at least two sp ecies caused by the same isolate is not likely due to chance. A graph showing the mean girdling for each isolate (scaled by ) across species, is at Figure 4 7, for comparison purposes with Figure 45, which show s the same species mean canker length s Using lesion volume as the response variable produced no significant result for the ( species*isolate ) interaction in 2012. S pecies had a significant effect however (P < .00 12). The statistics program output for the species test using th is variable is at Figure A 2. A graph of the lesion volume approximation for each isolate, by species, appears as Figure 46, so that it too, can be compared directl y with the graph s of CVM at Figure 45 and girdling at Figure 47. In 2011, neither girdling CVM, nor volume yielded a significant species*isolate interaction Both girdling and CVM did yield a significant species effect P<.0001 ( Figure A 5) and P < .0112 (Figure A 4), respectively, in 2011. Q. laurifolia (QLA) had the longest mean lesions in 2011. O ther significant differences are as shown by the different letters test in Figure s A 4 and A 5.
30 U sing the approximation of lesion volume (%HC CVM r2) resulted in a significant interaction for both species (P < .0002) and isolate (P < .00 64 ) in 2011. The statistics program outputs are in Figures A 6 and A 7 and the graph is at Figure 410.
31 Figure 41. QMU J1 1: Pycnidia ar ound flap. Photo: Adam Black F igure 42. QMU J1 1 Phloem necrosis (canker length = 3.3 cm) Photo: Adam Black.
32 Figure 43. QMU J1 1. Xylem streaking = 16.2 cm Photo: Adam Black. Figure 44. Negative correlation between canker length a nd stem diameter in 2012. (Stem diameters were estimated based on photographs, not measured.) 0 10 20 30 40 50 Canker Vertical Measurement (cm) Phloem tissue only 1 1.5 2 2.5 3 3.5 stem diam# (cm)
33 Figure 45. Mean canker l ength for i solates P and J, by species, 2012. Figure 46. Mean lesion volume (r2CVM%HC) for i solates P and J, by s pecies, 2012 (r = 1/2 estimated stem diameter) 0.00 2.00 4.00 6.00 8.00 10.00 12.00CPU QAC QBI QCO QHE QIN QIO QPA QLV QLY QMC QML QMU QNI QNU QPH QPU QRU QVL QVH QVC QVU QVS Mean volume of lesion, divided by pi, in cc 2012 Species mean volume of lesion (%HC x CVM x r*2) for P mean volume of lesion (% HC x CVM x r*2) for J
34 Figure 47. Mean c m, for i solates P and J, by species, 2012. Figure 48. Mean c anker l ength for i solates P and J, by species, 2011 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6CPU QAC QBI QCO QHE QIN QIO QPA QLV QLY QMC QML QMU QNI QNU QPH QPU QRU QVL QVH QVC QVU QVS Mean girdling (divided by pi), cm 2012 Species Mean %HC d/ ( P) Mean %HC d/ ( J) 0 1 2 3 4 5 6 QVA QSH QAL QMI QLA QAU QFA QMA QGE QCH QMYAverage Canker Length (cm) Average canker length per species for P Average canker length (cm) per species for J
35 Figure 49. Mean girdling (cm) for i solates P and J, by species, 2011 Figure 410. Mean lesion volume (r2CVM%HC) for isolates P and J, by species, 2011 (r = 1/2 estimated stem diameter) 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00 QVA QSH QAL QMI QLA QAU QFA QMA QGE QCH QMY(%HC d)/ cm Species 2011 Mean girdling, cm (P) Mean girdling, cm (J) 0.00 0.50 1.00 1.50 2.00 2.50 QVA QSH QAL QMI QLA QAU QFA QMA QGE QCH QMYMean lesion volume (%HC CVM r2) divided by (cm3) Species Mean lesion volume due to P Mean lesion volume due to J
36 Table 41. Frequencies of signs and symptoms ; mean canker lengtha,b; and mean per cent girdling in trees i noculated with Diplodia corticola strains P and J, by isolate. Dead tree frequencies denoted by asterisk. Hos t No. of Plants Inoculated Dead Plants in Accelerator Pot? c Pycnidia Bleeding Flap Death Necrosis Extent Beyond Flap Mean Canker Length a, b (cm) Mean Percent Girdling P J P J P J P J P J P J P J P J P J 2012 C. pumila 5 5 2* 2* 2 3 3 2 5 4 5 5 5 5 5.71 6.23 59.17 56.25 Q. acutissima 6 6 1 2 6 6 6 6 4 5 4.43 2.93 34.17 25.00 Q. bicolor 6 6 1 6 5 6 6 5 3.45 2.57 25.00 20.00 Q. coccinea 6 6 1 1 6 6 6 6 6 6 3.23 3.82 31.67 45.83 Q. hemisphaerica 6 6 1 4 6 6 6 6 5 6 5.17 8.82 31.67 35.00 Q. incana 6 6 1 6 6 6 6 6 6 10.45 8.38 55.83 45.00 Q. inopina 4 4 2* 4 4 4 4 4 4 4.05 7.50 63.75 82.50 Q. palustris 6 6 4 3 6 6 6 6 5 6 4.20 6.28 24.17 31.67 Q. laevis 6 6 1* 2* 2 5 6 6 6 6 6 6 5.178 12.60 48.33 67.50 Q. lyrata 6 6 2 1 6 6 6 6 6 1 6.45 2.80 26.67 20.83 Q. macrocarpa 6 6 2 1 6 6 6 6 6 3 4.20 3.22 31.67 20.00 Q. marilandica 6 6 6 6 6 6 6 6 3.25 3.92 21.67 21.67 Q. muehlenbergii 6 6 4 1 6 6 6 6 6 6 5.60 5.22 21.67 20.00 Q. nigra 5 6 3 4 5 6 5 6 5 6 5.57 7.85 28.33 40.00 Q. nuttallii 6 5 1 3 6 5 6 5 5 5 4.85 7.40 26.67 27.00 Q. phellos 6 6 1 5 6 6 6 6 6 6 5.02 13.08 24.17 45.00 Q. pumila 6 6 2* 3* 1 3 5 6 6 6 6 6 6.93 13.90 77.50 78.33 Q. rubra 6 6 1 6 6 6 6 4 5 3.37 3.68 21.67 24.17 Q. virginiana (LA seed source) 6 6 -5 2 6 6 6 6 5 4 5.23 3.52 25.83 21.67 Q. virginiana Highrise 6 6 1 1 6 6 6 6 6 4 3.18 2.93 28.33 21.67 Q. virginiana Cathedral 6 6 6 6 1 6 6 6 6 6 4 4.43 3.10 25.00 21.67 Q. virginiana Augustina 6 6 6 6 -6 6 6 5 6 5 4.10 3. 08 25.83 20.83 Q. virginiana Sky Climber 6 6 6 6 1 6 6 6 6 6 5 4.92 5.82 29.17 32.50 2011 Q. virginiana (FL seed source) 6 6 1 5 1 6 5 4 4.71 3.13 24.17 16.67 Q. shumardii 6 6 6 5 6 3 2 3 3.04 2.83 20.83 17.5 0 Q. alba 6 6 6 6 6 6 4 2 4.38 2.67 24.17 20.83 Q. michauxii 6 6 6 6 6 6 5 1 3.42 2.71 23.33 19.17 Q. laurifolia 6 6 1 1 6 6 6 6 5 5 5.04 4.63 25.00 27.50 Q. austrina 6 6 1 5 1 6 6 5 3.00 2.42 21.67 20.00 Q. falcata 3 3 1* 3 3 3 3 2 1 2.50 2.42 20.00 46.67 Q. margaretta 3 3 3 3 3 3 3 2.83 2.17 20.00 20.00 Q. geminata 3 3 1 2 1 3 3 3 1 4.83 2.08 23.33 20.00 Q. chapmanii 6 6 3 -5 3 6 6 5 3 4.67 2.33 42.50 26.67 Q. myrtifolia 6 6 2 6 6 6 6 5 5 3.83 3.25 35.83 25.83 a Dead plants were included in this mean by assigning an artificial lesion length equal to 110% of the longest lesion measured in that species/isolate/block. b Necroses measured in phloem (cm). c Accelerator pots were used only in 2012, for plants referring to this note ( c ). Of the dead C. pumila one P and one J were in Accelerator pots.
37 Table 42. Speciesa, s howing s ection (W = Quercus; R = Lobatae) Species Specie s codes Section (white or red) 2012 Q. bicolor QBI W Q. coccinea QCO R Q. hemisphaerica QHE R Q. incana QIN R Q. inopina QIO R Q. palustris QPA R Q. laevis QLV R Q. lyrata QLY W Q. macrocarpa QMC W Q. marilandica QML R Q. muehlenbergii QMU W Q. nigra QNI R Q. nuttallii QNU R Q. phellos QPH R Q. pumila QPU R Q. rubra QRU R Q. virginiana (LA seed source) QVL W Q. virginiana Highrise QVH W Q. virginiana Cathedral QVC W Q. virginiana Augustina QVU W Q. virginiana Sky Climber QVS W 2011 Q. virginiana QVA W Q. sh umardii QSH R Q. alba QAL W Q. michauxii QMI W Q. laurifolia QLA R Q. austrina QAU W Q. falcata QFA R Q. margaretta QMA W Q. geminata QGE W Q. chapmanii QCH W Q. myrtifolia QMY R a. QAC ( Quercus acutissima ) and CPU (Castanea pumila ) were omitted, Q AC being a member of the group Cerris, in China, and CPU belonging to a different genus
38 CHAPTER 5 DISCUSSION A singular finding in this experiment is that, on the wounded stem s inoculated with Diplodia corticola or Diplodia quercivora, callus tissue never formed while it did form on all controls. Equally important is that all tree s tested were highly susceptible to both pathogens. Absence of callus was observed by Sanchez et al ( 38) in wintertime field inoculations of oak species with D corticola in Mediterranean regions, while callus was formed in plants inoculated with the other pathogenic fungi they tested, Diplodia sarmentorum and Botryosphaeria dothidea The l esions produced by D. corticola were also significantly longer. I n the stem inocula tions done by Luque et al in 2000 (2 8) again callus formed only on controls and the lesions produced by D. corticola were much longer ( 21 cm in adequately irrigated plants and 18 cm in water stressed plants) than th os e produced by any of the other 31 pa thogens tested of which measured 2.5 cm and the longest of which was 9 cm Moreover, almost all the plants inoculated with D. corticola died within two weeks. Callus would limit the spread of the pathogen by sealing off the opening to the outside so that air c ould not dry out plant tis sues, allowing oxygen to reach the fungus ( 34) as well as by virtue of antimicrobial properties Its absence, at least under the conditions of this experiment, indicates that both Diplodia corticola and Diplodia quercivora interfere with callus formation The alterations in cell walls which normally occur to compartmentalize the decay or disease agent by walling it off inside the stem, per the Compartmentalization of Decay in Trees (CODIT) model (3 4), are apparently also not occurring, s ince nothing is limiting the vertical growth of the lesion in the bark, its expansion around the cambium its radial entry into the xylem or the extensive discoloration in the xylem, particularly above the inoculation point.
39 While there may be many reasons for the fai lure of wound healing in the plant, which would involve both lignification and suberization, the interference with suberin production, or the enzymatic degradation of suberin, seems likely since suberin has been shown to have a more important role than lignification in host/pathogen interactions in peach bark (5). Biggs and coworkers concluded that [i]n trees, the rapid production of high levels of suberin may be a determining factor in resistance against fungal pathogens (5). Suberin is a component of all four walls in the CODIT model (34, 5 ) and is formed de novo upon wounding, extensively in Quercus (3 4 33). A small amount also pre exist s in some cells. Wall 1 (which provides a conical cap across the xylem above and below the lesio n) consists of suberized vessel occlusions (tyloses) while wall 4 is formed in the cambium and would limit girdling. Suberized fibers are found beneath newly forming callus ( 3). T yloses are outgrowths of the parenchyma cells which protrude through pits i nto xylem vessels in response to infection ( 3, 33 34) W hite oaks overproduc e tyloses, the reason w hite oak is exclusively used for wine casks. Suberin, a primary component of tyloses, is hydrophobic and makes a waterproof seal; it also has anti microbial pr operties (3 4, 19, 3). The difference in disease response between white oaks and red oaks may be due to this difference in tylos i s production. L esion lengths and girdling in response to D. corticola infection were significantly less in white oaks than in red oaks in this study in 2012. Ironically, the disease so far has been observed in Florida almost exclusively in a white oak, Quercus virginiana but this could be a result of the greater market share this species commands A great deal of suberin is produced in cork cells, as reflected in the name Quercus suber explaining why the removal of cork may leave Q. suber particularly vulnerable to Diplodia corticola Accumulation rates of suberin at wound sites in bark tissues were positively
40 correlated w ith disease resistance in peach cultivars ( 4 ). In the present study, no resistance was seen in any of the inoculated trees, save the single Q. michauxii tree that did form callus, despite the presence of fungus in that tissue. S mall stem diameter is a ris k factor for disease. There was a significant negative correlation between small stem size and not simply girdling, which would seem obvious, but canker length, in 2012. Figure 44. The species most girdled in 2012 were C PU and QLV, both of which are co nsidered shrubby. C ankers were longest in QPU, QPH, and QLV all in response to isolate J, and QPU and QLV are also considered shrubby. Figs. 4 6 and 410. The t wo QPH ( Q. phellos ) individuals with 46and 49cm long cankers were not particularly small stemmed. It is a red oak, however, so more susceptible as a rule. The conclusion that small stem size is a risk factor is consistent with the twig dieback seen in natural infections caused by both pathogens, but more consistently in Florida by isolate P, which has now been identified as Diplodia quercivora. Interference with either lignification or suberization might explain why D. corticola and D. quercivora in nature, without any obvious wounding, kill the smallest twigs. Although no express author ity was found for this proposition, t hese twigs are the youngest, and presumably still in the process of forming bark. In fact, because no twig dieback was observed on the artificially infected trees in this study, where the trees were inoculated in the trunk --nor is twig dieback reported as a symptom in the cork oak infections, where the wound is on the trunk this symptom must occur because the pathogen has entered through (or become active in) the outermost twigs initially. It is unknown how the fungus gains entry where there are no obvious wounds. T here could be physiological factors governing entry, such as gas exchange rates or lenticel size ; environmental influences ; and ( probable ) differences in infective capability between the two fungal isolates. Some of the
41 hosts which reacted to the pathogen in trunk inoculations in this study may never be infected naturally, therefore. Possible pathogen entry through lenticels in twigs might be investigated by spraying a spore suspension on them directly, wi thout wounding. The death of young twigs due to girdling by cankers may be due, as Agrios said, to their inability to produce defensive tissues faster than the fungus grows (1) However, i n the present study, trunks much larger ( 13.25 cm in diameter ) a nd older ( 1 3 years ) than small twigs were girdled. This fact, along with the stem bleeding and long xylem streaking are diagnostic not simply of a fast growing fungus, but of a phytotoxin readily translocated through the xylem particularly since the fu ngus itself was not isolated away from the phloem necrosis boundary. Investigations of canker formation by Hypoxylon mammatum in quaking aspen ( Populus tremuloides) similarly found that in addition to bark necrosis and collapse the failure to form callu s was due to a pathotoxin which interfered with wound healing, and was produced as a normal metabolite by H. mammatum (39). A s mentioned in the Introduction, d iplopyrone the phytotoxin produced by Diplodia corticola caused the collapse of internal stem t issue in tomato plants ( 31) and was toxic to cork oak in tiny concentrations In a nother fungus which causes stem lesions, Botrytis cinerea phytopathogenicity is caused by a suite of factors which are believed to act synergistically, including cell wall degrading enzymes, phytotoxins, re active oxygen species, or membrane transporters for secretion of plant defence compounds ( 37). Diplopyrone, along with other D. corticola / D. quercivora metabolites, may be involved in degrading suberin or interfering with cell signaling necessary for its production, and s ynergistic activity among them is likely as with B otrytis cinerea It should further be remarked that Quercus has an impressive arsenal of phytoalexins, defensive chemicals such as tannins and ter penes, that many other
42 pathogens cannot tolerate (3 4) Diplodia corticola and Diplodia quercivora are obviously not deterred The pathogens may be interfering with these induc ible oak defenses, as well, therefore. Respecting the use of approximated lesion volume as a r esponse variable: ideally, rather than %HC r2h ( percent girdling times cylinder volume, with h = lesion length, and r = stem diameter), as was used here (scaled by ) the actual thickness of the lesion (difference between stem radius with and witho ut lesion) should be used for a better approximation. That measurement was not available in this study. If it had been, the formula would be (R2 r2) h %HC, where r is the radius of the cankered (depressed) area on the stem and R is the radius of the uncankere d stem This would describe a wedge the shape described by Urbez T orres for cankers formed by D. corticola on grapevine (42). The appearance of nonDiplodia fungal fr uiting bodies on many inoculated trees remains a mystery worthy of more study. They could possibly be endophytes forced out by the invader, or mycoparasites. They were not observed on controls, and the same fungi appeared on the same oak species. In sum under the conditions of this experiment all species were susceptible to both pathogens, and the difference in their responses to the two strains was in the main not statistically significant, as noted. T hese members of the Fagaceae have some common genet ic susceptibility to these closely related species of Diplodia. This paper concludes that the high susceptibility of oaks is due to the impairment of wound repair processes by the phytotoxin (or suite of phytotoxins) produced by Diplodia corticola and Dip lodia quercivora.
43 APPENDIX A RESULTS FROM STATISTIC S PROGRAM Level Least Sq Mean QPU,J A 13.896667 QPH,J A B 13.083333 QLV,J A B C 12.598333 QIN,P A B C D 10.445000 QHE,J A B C D 8.816667 QIN,J A B C D 8.383333 QNI,J A B C D 7.8500 00 QIO,J A B C D 7.500000 QNU,J A B C D 7.400000 QPU,P A B C D 6.933333 QNI,P A B C D 6.680000 QLY,P A B C D 6.450000 QPA,J A B C D 6.283333 CPU,J A B C D 6.232000 QVS,J A B C D 5.816667 CPU,P A B C D 5.710000 QMU,P A B C D 5.600000 QVL,P A B C D 5.233333 QMU,J A B C D 5.216667 QLV,P A B C D 5.171667 QHE,P A B C D 5.166667 QPH,P A B C D 5.016667 QVS,P A B C D 4.916667 QNU,P A B C D 4.850000 QVC,P A B C D 4.433333 QAC,P A B C D 4.433333 QPA,P B C D 4.200000 QMC,P B C D 4.200000 QVU, P B C D 4.100000 QIO,P A B C D 4.050000 QML,J B C D 3.916667 QCO,J B C D 3.816667 QRU,J B C D 3.683333 QVL,J B C D 3.516667 QBI,P C D 3.450000 QRU,P C D 3.366667 QML,P C D 3.250000 QCO,P C D 3.233333 QMC,J C D 3.2166 67 QVH,P C D 3.183333 QVC,J C D 3.100000 QVU,J C D 3.083333 QVH,J D 2.933333 QAC,J D 2.933333 QLY,J D 2.800000 QBI,J D 2.566667 Levels not connected by same letter are significantly different. Figure A 1. 2012, species*isolate interaction using mean canker length as response variable, LS Means Differences Tukeys HSD
44 Level Least Sq Mean QLV A 5.9787500 QPH A B 5.4079167 CPU A B C 4.2540000 QNI A B C 2.9857500 QMU A B C 2.9707031 QHE A B C 2.554583 3 QMC A B C 2.2640625 QPU A B C 2.2329167 QVS A B C 2.1279167 QNU A B C 1.7933333 QPA A B C 1.7383333 QCO A B C 1.4362500 QAC A B C 1.3520833 QIN A B C 1.3478125 QIO A B C 1.1989063 QLY A B C 1.1558333 QVL A B C 1.0754167 QVC B C 0.9125000 Q VU B C 0.8645833 QML B C 0.8133333 QRU B C 0.8129167 QBI B C 0.6987500 QVH C 0.1913542 Levels not connected by same letter are significantly different. Figure A 2. 2012: Lesion volume (%HC CVM r2) LS Means Differences, Tukeys HSD for species
45 Level Least Sq Mean QLV,J A 1.3500000 CPU,P A B 1.3200000 CPU,J A B C 1.3000000 QLV,P A B C D 0.9666667 QMC,P A B C D 0.9500000 QPU,P A B C D 0.9416667 QCO,J A B C D 0.9166667 QPH,J A B C D 0.9000000 QIO,J A B C D E 0.8250000 QNI,J A B C D E 0.8000000 QPU,J A B C D E 0.7833333 QMU,P A B C D E 0.7041667 QHE,J B C D E 0.7000000 QAC,P B C D E 0.6833333 QNI,P A B C D E 0.6800000 QMU,J B C D E 0.6500000 QVS,J B C D E 0.6500000 QIO,P A B C D E 0.6375000 QPA,J C D E 0.6333333 QCO,P C D E 0.6333333 QHE,P C D E 0.6333333 QMC,J D E 0.6000000 QVS,P D E 0.5833333 QIN,P D E 0.5583333 QNU,J D E 0.5400000 QLY,P D E 0.5333333 QNU,P D E 0.533333 3 QVL,P D E 0.5166667 QVU,P D E 0.5166667 QBI,P D E 0.5000000 QAC,J D E 0.5000000 QRU,J D E 0.4833333 QPA,P D E 0.4833333 QPH,P D E 0.4833333 QVC,P D E 0.4500000 QIN,J D E 0.4500000 QVC,J D E 0.4333333 QVL,J D E 0.4333333 QML,P D E 0.4333333 QML,J D E 0.4333333 QLY,J D E 0.4166667 QVU,J D E 0.4166667 QRU,P D E 0.4000000 QBI,J D E 0.4000000 QVH,P D E 0.3250000 QVH,J E 0.21666 67 Levels not connected by same letter are significantly different. Figure A 3. 2012: Mean girdling, LS Means Difference Tukeys HSD, for species.
46 Level Least Sq Mean QLA A 4.8333333 QVA A B 3.9166667 QMY A B 3.5416667 QAL A B 3.5208333 QC H A B 3.5000000 QGE A B 3.4583333 QMI A B 3.0625000 QSH B 2.9375000 QAU B 2.7083333 QMA B 2.5000000 QFA B 2.4583333 Levels not connected by same letter are significantly different. Figure A 4. 2011: Canker vertical measurement, LS Means D ifference Tukeys HSD, for species. Level Least Sq Mean QCH A 0.69166667 QMY A B 0.61666667 QAU A B C 0.57291667 QLA A B C 0.52500000 QAL B C D 0.45000000 QMI B C D 0.42500000 QVA B C D 0.40833333 QSH C D 0.38333333 QFA C D 0.33333333 QGE D 0.21666667 QMA D 0.20000000 Levels not connected by same letter are significantly different. Figure A 5. 2011: Mean girdling (cm), LS Means Difference, Tukeys HSD, for species. Level Least Sq Mean QCH A 1. 4750000 QLA A 1.3812500 QMY A B 1.1885417 QAU A B 1.0772624 QVA A B 0.8489583 QAL A B 0.8166667 QMI A B 0.6552083 QSH A B 0.5906250 QGE B 0.1968750 QFA B 0.1895833 QMA B 0.1250000 Levels not connected by same letter are significantly di fferent. Figure A 6. 2011: Approximation of lesion volume (%HC CVM r 2) LS Means Difference, Tukeys HSD, for species
47 Level Least Sq Mean P A 0.96904297 J B 0.58458807 Levels not connected by same letter are significantly different. Figure A 7. 2011: Approximation of lesion volume (%HC CVM r 2), LS Means Differences, Tukeys HSD, for isolate
48 LIST OF REFERENCES 1. Agrios G.N 2005. Plant Pathology (2d ed.) Elsevier Academic Press: New York, pp. 473, 890. 2. Alves, A. A. Correia J. Luque, a nd A. Phillips. 2004. Botryosphaeria corticola, sp. nov. on Q uercus species, with notes and description of Botryosphaeria stevensii and its anamorph, Diplodia mutila Mycologia 96(3), 598613. 3. Biggs, A.R. 1987. Occurrence and location of suberin in wound reaction zones in xylem of 17 tree species. Phytopathology 77(5), 71825. 4. Biggs, A.R., and N.W. Miles. 1988. Association of suberin formation in uninoculated wounds with susceptibility to Leucostoma cincta and L. persoonii in various peach cultivars. Phytopathology 78(8), 107074. 5. Blanchette, R.A. and A.R. Biggs (eds). 1992. Defense Mechanisms of Woody Plants Against Fungi pp. 35, 26. Springer Verlag: New York. 6. Boyce, J.S., Jr., and C.F. Speers. 1960. Oak dieback in Virginia in 1959. Plant Disease Reporter 44(5), 351. 7. Bush, E.A. Botryosphaeria c anker and dieback of t rees and s hrubs in the l andscape. 2009. Virginia Cooperative Extension Report 450726 : http://pubs.ext.vt.edu/450/450726/450726.html (accessed July 25, 2013) 8. Campanile, G., A. Ruscelli and N. Luisi. 2007. Antagonistic activity of endophytic fungi towards Diplodia corticola assessed by in vitro and in planta tests. Eur. J. Plant Pathol 117, 23746. 9. Candolfi Arballo, O., C. ValenzuelaSolano, W.D. Gubler and R. Hernandez Martinez. 2010. Botryosphaeriaceae species associated with grapevine decline in Mexico. Phytopathologia Mediterranea 49, 10506. 10. Coleman, T. N. Grulke, M. Daly, C. Godine z, S. Schilling, P. Riggan, and S. Seybold. 2011. Coast live oak, Q. agrifolia, susceptibility and response to gold spotted oak borer, Agrilus auroguttatus injury in southern California. Forest Ecol. Manage. 261, 185265. 11. CREM http://www.crem.fct.unl.pt/botryosphaeria_site/botryosphaeria_corticola_2.htm? &lang=en_us&output=json&sessionid=047dec826ae4544b72ddbe01909cdb17 (accessed 752013) 12. Crous, P.W. B. Slippers M.J. Wingfield J. Rheeder, W.F.O. Marasas, A.J.L. Philips A. Alves T. Burgess, P. Barber and J.Z. Groenewald. 2006. Phylogenetic lineages in the Botryosphaeriaceae. Stud. My col 55 23553.
49 13. Denman, S., P.W. Crous, J.E. Taylor J. Kang, I. Pascoe, and M.J. Wingfield. 2000. An overview of the taxonomic history of Botryosphaeria, and a re evaluation of its anamorphs based on morphology and ITS rDNA phylogeny. Studies in Myc ology 45, 12940. 14. DeWet, J. B. Slippers O. Preisig B.D. Wingfi eld, and M.J. Wingfield. 2007. Phylogeny of the Botryosphaeriaceae reveals patterns of host association. 2008. Molecular Phylogenetics and Evolution 46, 11626. 15. Djoukeng, J.D., S. Polli P. Larignon, and E. AbouMansour. 2009. Identification of phytotoxins from Botrysphaeria obtusa, a pathogen of black dead arm disease of grapevine. European Journal of Plant Pathology 124, 30308. 16. Dreaden, T.J. K. Shin, and J.A. Smith. 2011. First re port of Diplodia corticola causing branch cankers on live oak ( Q. virginiana) in Florida. Plant Disease 95(8), 1027 (Aug. 2011). 17. Farr, D.F., and Rossman, A.Y. Fungal Databases. Systematic Mycology and Microbiology Laboratory. http://nt.ars rin.gov/fungaldatabases/new_allView.cfm?whichone=FungusHost &thisName=Diplodia%20corticola&organismtype=Fungus&fromAllCount=yes (accessed 72213) 18. Franceschini, A. B.T. Linaldeddu, and F. Marras. 2005. Natural infection periods of Diplodia corticola in a declining cork oak forest. Journal of Plant Pathology 87(4) (suppl.), 294. 19. Garrod, B., B.G. Lewis M.J. B rittain and W.P. Davies. 1982. Studies on the contribution of lignin and suberin to the impedance of wounded carrot root tissue to fungal invasion. New Phytologist 90, 99108. 20. House, E.S. 2007. Diplodia pinea, Diplodia scrobiculata and Lasiodiplodia theobromae in South Georgia pine plantations and their respective roles in causing tip dieback in Southern pine species. M.S. thesis: University of Georgia. 21. Inderbitzin, P., R.M. Bostock, F.P. Trouillas and T.J. Michailides. 2010. A sixlocus phylogeny reveals high species diversity in Botryosphaeriaceae from California almond. Mycologia 102(6), 135068. 22. Ingram, D.E. 1914. A twig blight of Quercus prinus and related species. Journal of Agricultural Research 1(4), 33947. 23. Jacobs, K.A., and S.A. Re hner. 1998. Comparison of cultural and morphological characters and ITS sequences in anamorphs of Botryosphaeria and related taxa. Mycologia 90(4), 601. 24. Linaldeddu, B.T., Franceschini, A., Alves, A., and Phillips, A.J.L. 2013. Diplodia quercivora : a new oak pathogen. Mycologia (preliminary version published May 25, 2013, as doi: 10.3852/12/370).
50 25. Linaldeddu, B.T., C. Sirca D. Spano, and A. Franceschini. 2009. Physiological responses of cork oak and holm oak to infection by fungal pathogens involve d in oak decline. Forest Pathology 39, 23238. 26. Luque, J., M. Cohen, R. Save C. Biel and I.F. Alvarez. 1999. Effects of three fungal pathogens on water relations, chlorophyll fluorescence and growth of Q. suber L. Annals of Forest Science 56: 1926. 27. Luque, J., and J. Girbal. 1989. Dieback of cork oak ( Q. suber ) in Catalonia (NE Spain) caused by Botryosphaeria stevensii European Journal of Forest Pathology 19, 713. 28. Luque, J., J. Parlade, and J. Pera. 2000. Pathogenicity of fungi isolated from Q. suber in Catalonia (NE Spain). Forest Pathology 30, 24763. 29. Lynch, S.C., A. Eskalen P.J. Zambino, J.S. Mayorquin, and D.H. Wang. 2013. Identification and pathogenicity of Botryosphaeriaceae species associated with coast live oak ( Q. agrifolia ) decline in southern California. Mycologia 105(1), 12540. 30. Lynch, S.C., A. Eskalen P. Zambino, and T. Scott. 2010. First r eport of Bot Canker c aused by Diplodia corticola on c oast l ive oak ( Q. agrifolia) in California. Plant Disease 94( 12) 1510. 31. Maddau, L., E. Spanu, B.T. Linalde d du, A. Franceschini and A. Evidente. 2008. Phytotoxic m e t abolites produced by f ungi i nvolved in c ork oak decline. SardiniaChem 2008. Giornato di Studio Dedicata Alla Chimica Organica Delle Molecole Biologicamente Attive 30 Maggio 2008, at 7071. 32. McPherson, B.A., N. Erbilgin P. Bonello, and D.L. Wood. 2013. Fungal species assemblages associated with Phytophthora ramorum infected coast live oaks following bark and ambrosia beetle colonization in northern California. Forest Ec ology and Management 291, 3042. 33. Pearce, R.B. 1984. Suberin in the sapwood of oak ( Quercus robur L.): its composition from a compartmentalization barrier and its occurrence in tyloses in undecayed wood. Physiological Plant Pathology 24, 7181. 34. Pearce, R.B. 1996. Tansley Review No. 87: Antimicrobial defences in the wood of living trees. New Phytologist 132, 20333. 35. Pintos Varela C., V. Redondo Fernndez O. Agun Casal and J.P. Mansilla Vzquez 2011. First report of cankers and dieback caused by Neofusicoccum mediterraneum and Diplodia corticola on grapevine in Spain. Plant Disease 95(10), 1315. 36. Ragazzi, A., I.D. Fedi, and L. Mesturino. 1989. The oak decline: a new problem in Italy. Eur. J. For. Pathol. 19: 105110.
51 37. Reino, J.L., R. Hernande z Galan, R. Duran Patron and I.G. Collado. 2004. Virulence toxin production relationship in isolates of the plant pathogenic fungus Botrytis cinerea Journal of Phytopathology 152, 53366. 38. Sanchez, M.E. J. Venegas, and M.A. Romero. 2003. Botryosphae ria and related taxa causing oak canker in Southwestern Spain. Plant Disease 87(12), 151521. 39. Schipper, A.L., Jr. A Hypoxylon mammatum pathotoxin responsible for canker formation in quaking aspen. 1978. Phytopathology 68, 86672. 40. Schmidt, R.A., and C.L Fergus. 1965. Branch canker and dieback of Quercus prinus L., caused by a species of Botryodiplodia. Canadian Journal of Botany 43: 73137. 41. Slippers, B., W.A. Smit P.W. Crous T.A. Coutinho, B.D. Wingfield, and M.J. Wingfield. 2007. Taxonomy, phylogeny and identification of Botryosphaeriaceae associated with pome and stone fruit trees in South Africa and other regions of the world. Plant Pathology 56, 12839. 42. Urbez Torres, J.R. 2011. The status of Botryosphaeriaceae species infecting grapevines. Phytopathol. Mediterr 50 (Supp.), S5S45. 43. Urbez Torres, J.R., M. Battany, L.J. Bettiga, C. Gispert, G. McGourty, J. Roncoroni, R.J. Smith, P. Verdegaal, and W.D. Gubler. 2010. Botryosphaeriaceae species sporetrapping studies in California vineyards. Plant Disease 94(6), 71724. 44. Urbez Torres, J.R., F. Peduto, S. RooneyLatham and W.D. Gubler. 2010. First r eport of Diplodia corticola c ausing grapevine ( Vitis vinifera ) c ankers and t runk c ankers and d ieback of c anyon l ive oak ( Q. chrysolepis) in California. Plant Disease 94 ( 6) 785. 45. Vajna, L. 1986. Branch canker and dieback of sessile oak ( Q. petraea ) in Hungary caused by Diplodia mutila I. Identification of pathogen. Eur. J. For. Path. 16, 22329. 46. Venkatasubbaiah, P., T.B. Sutton, and W.S. Chi lton. 1991. Effect of phytotoxins produced by Botryosphaeria obtusa, the c ause of black r ot of a pple f ruit and f rogeye l eaf s pot. Phytopathology 81: 3, 24347.
52 BIOGRAPHICAL SKETCH Sonja Mullerin, who is called Sunny by her family and friends, is the name legally adopted by Alison Maynard in 2011, after 24 years as a lawyer in Colorado. In her law practice, Sunny specialized in water, land use, and civil rights law, litigating on behalf of ci tizens and environmental groups often pro bono or on a contingent basis against powerful and well funded developers and governmental officials. She published several law review articles and was the 2002 Green Party candidate for Colorado Attorney General, campaigning against both the Democrat and the Republic an on an anti corruption platform rooted in facts uncovered in her litigation. In 2005, Sunny began to take biology and chemistry courses one or two at a time at the University of Colorado Denver, and to realize she had been diverted from her childhood pas sion, which was biology. She began to work towards a career switch the first step of which was taken when she moved to Gainesville in June 2011 to begin graduate school in Forest Resources and Conservation at the University of Flor ida, working as a research assistant for Asst Professor Jason A. Smith. Her original project was to explore whether a gene for gene interaction was at play between fusiform rust and its oak host. That led to work on the Botryosphaeria ceae, including th e Diplodia corticola host range study which is subject of this thesis. Aside from her degrees from UF and UCD, Sunny obtained her B achelor of Arts in physics in 1976 from Cornell University, College of Arts and Sciences, where she went on a National Merit Scholarship, and her J uris Doctor from the University of Denver in 1986. Between those two educational experiences she traveled and worked in various countries in Europe and Africa, as well as in the geophysical industry in California Colorado, Wyoming, and North Dakota Sunnys father, Robert G. Maynard, was an oil company geologist who had been an Army major in World War II awarded the Silver Star for bravery He died in 2007. Sunnys
53 mother, Emy, who was schooled in Germany and enrolled in UCLA at t he tender age of 16 as a geology major, instilled in Sunny a love of books as well as science. She died in 1972. The name Sonja Mullerin is an invention. Sonja was Sunnys name in Russian class at Cornell, as well as the name of her best friend in 4th grade (with whom she lost contact long ago). It means wisdom in Greek. Mullerin seemed to follow Sonja naturally on the tongue as the name of a Schubert song cycle (die Schoene Mullerin).