ANALYSIS OF ERWINIA AMYLOVORA POPULATION DYNAMICS AT FLOWER BLOOM AND FURTHER SYSTEMIC MOVEMENT OF THE PATHOGEN THROUGH HOST TISSUE By Suzanne M. Slack A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Plant Pathology – Doctor of Philosophy 2021 ABSTRACT ANALYSIS OF ERWINIA AMYLOVORA POPULATION DYNAMICS AT FLOWER BLOOM AND FURTHER SYSTEMIC MOVEMENT OF THE PATHOGEN THROUGH HOST TISSUE By Suzanne M. Slack Population dynamics of Erwinia amylovora have been utilized for many applications over the past 50 years. In this work, populations were tracked over the course of apple bloom under naturally occurring environmental conditions. Flower stigmas inoculated on the 1st day of being open can harbor large (107) populations after 3 or 5 days post inoculation, with 100-fold increases in E. amylovora stigma populations observed in atmospheric conditions with daily average temperatures near 14oC. These large 100-fold increases seem to occur at night, indicating that Erwinia amylovora is able to infect flowers in colder field temperatures than previously reported. In tandem, timing of antibiotic application relative to E. amylovora presence on flower stigmas had little impact on population dynamics, with streptomycin and kasugamycin consistently reducing populations while oxytetracycline was more variable. This study also led to the identification that Kasumin is prone to photodegradation. Culturable bacteria, yeast, and fungal populations were assessed over the course of bloom in relation to application of a biological control agent yeast (Aureobasidium pullulans) and a contact sterilant. Though populations fell rapidly directly after application, by 24 hours all populations returned to pre-spray levels. In regards to further systemic spread, a type three secretion system effector was indicted in blocking the abscission of infected flowers. The use of prohexadione calcium (Pro-ca) and acibenzolar-S- methyl (ASM) reduced internal E. amylovora spread in a seemingly synergistic manner on young trees, with rates of 28.3 g Pro-Ca + 28.3 g ASM or 56.6 g Pro-Ca + 28.3 g ASM. The sum of this work gives us a better understanding of E. amylovora population dynamics in field conditions as well as the population response to current treatment options. Copyright by SUZANNE M. SLACK 2021 ACKNOWLEDGMENTS The author would like to thank committee members Dr. George Sundin, Dr. Mary Hausbeck, Dr. Courtney Hollender, and Dr. Kerik Cox; current and past Sundin lab members: Cory Outwater, Dr. Quan Zeng, Dr. Jingyu Peng, Dr. Jeff Schachterle, Dr. Leire Bardaji Goikoetxea, Dr. Xiaochen Yuan, Roshni Kharadi, Emma Sweeney, Megan Botti-Marino, Katherine Olive, Ishani Pandya, and Jerimiah Johnson; Dr. Ray Hammerschmidt, Dr. Kellie Walters and Dr. Sara Villani for their always honest opinions; her family, especially Claire Slack, Jackie Backus, and Robert Slack for their unwavering support; and the following funding sources: the Michigan Apple Committee, an MSU Plant Science Fellowship, the Jordan B. Tatter Scholarship in Horticulture, the MSU Plant Pathology program Paul Taylor travel award, an MSU CNS Dissertation Continuation Fellowship, and the Everett Beneke Endowment Award. v TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES KEY TO ABBREVIATIONS Introduction Impacts of Environment Impact of Biological Control Agents and Microbiomes Future Prospects Conclusions CHAPTER 1 LITERATURE REVIEW OF POPULATION DYNAMICS IN ERWINIA AMYLOVORA CHAPTER 2 IN-ORCHARD POPULATION DYNAMICS OF ERWINIA AMYLOVORA ON APPLE FLOWER STIGMAS Abstract Introduction Materials and Methods Results Discussion CHAPTER 3 EFFECT OF KASUGAMYCIN, OXYTETRACYCLINE, AND STREPTOMYCIN ON IN-ORCHARD POPULATION DYNAMICS OF ERWINIA AMYLOVORA ON APPLE FLOWER STIGMAS Abstract Source CHAPTER 4 EVALUATION OF A CONTACT STERILANT AS A NICHE-CLEARING METHOD TO ENHANCE THE COLONIZATION OF APPLE FLOWERS AND EFFICACY OF AUREOBASIDIUM PULLULANS IN THE BIOLOGICAL CONTROL OF FIRE BLIGHT Abstract Source vi viii x xii 1 1 1 2 4 7 8 9 9 9 10 13 16 34 40 40 40 41 42 42 42 43 Abstract Introduction Materials and Methods Results Discussion CHAPTER 5 DECIPHERING THE ROLE OF A TYPE THREE SECRETION SYSTEM EFFECTOR IN THE PATHOGENICITY OF ERWINIA AMYLOVORA CHAPTER 6 POSSIBLE SYNERGISTIC EFFECT OF PROHEXADIONE-CALCIUM AND ACIBENZOLAR-S-METHYL IN MANAGING SHOOT BLIGHT OF APPLE REFERENCES Abstract Introduction Materials and Methods Results Discussion 44 44 44 45 47 51 61 63 64 64 65 67 72 83 85 vii LIST OF TABLES 21 Table 2-1. Results from field experiments tracking the population dynamics of Erwinia amylovora on apple flower stigmas where experiments were initiated on different-aged flowers. Table 2-2. Analysis of population surge events (daily increase of log10 0.9 or greater) observed in the field for Erwinia amylovora on 1-day and 3-day open flowers including the number of generations and relative growth rates observed in the respective 24-hr periods. Table 2-3. Weather parameters associated with population surges of Erwinia amylovora during growth on 1-day open apple flower stigmas. Table 2-4. Calculated Pearson correlations of weather variables with overnight growth of Erwinia amylovora during days with population surges on apple flower stigmas. Correlations shown are statistically significant (P < 0.05). Positive correlations are depicted with positive value, and negative correlations have a negative value. Values range between -1 and 1; the stronger the correlation, the closer the value will be to 1 or -1, depending on the type of relationship. For this table, if a correlation is positive, then growth increased as the weather variable value increased. Table 2-5. Incidence of blossom blight infection on apple flowers previously inoculated with Erwinia amylovora at the 1-day open age. Weather parameters with the potential to affect growth and infection are also listed. Table 5-1. Results of the 2017 and 2018 field trial observing abscission on flowers inoculated with either Erwinia amylovora Ea110, Pantoea agglomerans E325. Pseudomonas fluorescens A506, a phosphorus buffer, or a water control with 4 replicates and a total of 100 flowers per treatment. Flowers were inoculated on 27 April 2017 and 1 May 2018 respectively. Statistical differences between treatments were compared by a one way ANOVA with means separation by Tukey HSD. * indicates statistical difference between treatments. Table 6-1. Application rates and times of Pro-Ca and ASM for High Density and Young Gala trees. The following treatments were applied at time points 1, 3 and 5: 128.3 g Pro-Ca, 56.6 g Pro-Ca, 84.9g Pro-Ca, 113.4 g Pro-Ca, 226.4 g Pro-Ca, 28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro-Ca+ 28.3 g ASM; and at time points 2, 4 and 6: 28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro-Ca+ 28.3 g ASM. Table 6-2. Total precipitation (mm) during application period for each year. 53 23 27 28 29 69 76 viii Table 6-3. Table 3: Summary of horticultural measurements from 2018-2020 on High Density trees including: average tree height treatment per rootstock; average incidence of winter damage per treatment per root stock; average incidence of systemic spread by E amylovora per treatment per rootstock. 79 ix LIST OF FIGURES 31 20 19 Figure 2-1. Population dynamics of Erwinia amylovora Ea110 from four representative experiments on 1-day, 3-day, and 5-day open flowers of apple cultivars ‘Gala’ and ‘Jonathan’ over a 5-day experimental period. Mean log10 populations ± the standard error of the mean are shown. Daily samplings were conducted at 1000. Figure 2-2. Population dynamics of Erwinia amylovora Ea110 from two representative experiments on 1-day, 3-day, and 5-day open flowers of apple cultivars ‘Gala’ and ‘Jonathan’ over a 5-day experimental period. Mean log10 populations ± the standard error of the mean are shown. Daily samplings were conducted at 1000. Weather parameters, temperature, precipitation, relative humidity, and windspeed are also shown. Figure 2-3. Population dynamics of Erwinia amylovora Ea110 on 1-day open flowers of apple cultivar ‘Gala’ over a 42-hr experimental period with frequent samplings during daylight hours of the second day of the experiment. Mean log10 populations ± the standard error of the mean are shown. Figure 2-4. Population dynamics of Erwinia amylovora Ea110 on 1-day open flowers of apple cultivar ‘Gala’ over a 47-hr experimental period with frequent samplings during overnight hours of the second day of the experiment. Mean log10 populations ± the standard error of the mean are shown. Red asterisks highlight the 2200 and 0200 (+1) sampling time points. Figure 2-5. Temperature and dewpoint readings for the overnight periods of four experiments monitoring the population dynamics of Erwinia amylovora on 1-day open apple flower stigmas with frequent overnight sampling time points. Figure 5-1. Flowers with pollinator excluder bags attached. A: Image of a recently opened apple flower cluster covered by a pollinator excluder bag. B: Image of an apple flower cluster that was inoculated with Pantoea agglomerans E235. Observation was 20 days after inoculation from a field experiment conducted in 2018. Figure 5-2. Rate of abscission elicited by E. amylovora WT and effector and virulence factor mutant strains. Results of two experimental replicates with a total of 6 trees per treatment averaging 120 flowers per treatment total. Error bars represent standard error. 55 Figure 5-3. Comparison of selected flower clusters at three days and seven days after inoculation wither either E. amylovora strain 1189 (WT) or the effector mutant 1 (Ea1189∆HopC). WT at 3 days (A), Ea1189∆HopC at 3 days (B), WT at 7 days (C), and Ea1189∆HopC at 7 days (D). 49 56 32 33 x Figure 5-4. Figure 4: Erwinia amylovora strain populations 24 hr after inoculation into the hypanthium. Flowers were inoculated with 1ul of 1x107 CFU. Error bars indicate the standard error of the population. Figure 5-5. Histology sides observing the abscission zone in apple. A. WT at 3 days post inoculation, B. Un-inoculated control at 3 days post inoculation, C. Ea1189∆HopC at 3 days post inoculation, D. WT at 5 days post inoculation, and E. Un-inoculated control at 5 days post inoculation. Figure 5-6. Effect of Ethephon on flowers inoculated with Erwinia amylovora. Error bars indicate standard error. Treatments that are significantly different have a different letter, means separated one way ANOVA followed by Tukey’s HSD. Figure 6-1. Lay out of the high density orchard block utilized for the experiment. Figure 6-2. Average difference in shoot length of treatments in 8-10 year old Galas from 2018 to 2020. Error bars represent standard error of the mean. Figure 6-3. Average Percent lesion length of treatments in 8-10 year old Galas from 2018, 2019, and 2020. Error bars represent standard error of the mean. Figure 6-4. Average shoot length growth in high density treatments from 2018, 2019, and 2020 per rootstock. Figure 6-5. Average percent branch lesion in the high density Gala orchard in 2019 and 2020 per rootstocks. Error bars represent standard error of the treatment mean Figure 6-6. Average weight of marketable apples in the high density Gala orchard from 2020 per rootstock, comparing inoculated and uninoculated trees within treatment. Error bars represent standard error of the treatment mean. 57 59 60 71 74 75 80 81 82 xi KEY TO ABBREVIATIONS ANOVA ASM Pro-Ca Analysis of variance Acibenzolar-S-methyl Prohexadione calcium Tukey’s HSD Tukey’s honest significant difference test WT Erwinia amylovora strain Ea1189 xii CHAPTER 1 LITERATURE REVIEW OF POPULATION DYNAMICS IN ERWINIA AMYLOVORA Introduction Population dynamics in Erwinia amylovora has historically focused on a single strain’s population size and reaction, either waxing or waning, in response to external stimuli. External stimuli examined for effect on populations includes environmental factors (Billing, 1974; Pusey, 1986; Schouten, 1987). Billing (1974) used temperature to show that E. amylovora growth rates were comparable between 21 and 30°C, and below 18°C the bacterial growth was reduced. Many studies have utilized E. amylovora population dynamics in some capacity for developing disease forecasting models (Steiner, 1990; Smith 1996; Billing, 2011; Smith and Pusey 2011; Turechek and Biggs 2015). Other researchers have further built on these forecasting models based on population dynamics, though mainly for the improvement of timing biological control applications (Thomson and Gouk, 2003; Pusey and Curry, 2004; Johnson et al., 2009). Population dynamics have also been utilized to directly gauge the potential effectiveness of various compounds produced by biological control agents (Pusey, 1997; Giddens et al., 2003; Pusey et al., 2009), and recently, Slack et al. (2021) published the effects of three antibiotics on E. amylovora stigma population dynamics. The other definition of population dynamics focuses on the diversity and spatial distribution of individuals within that population. With the advent of genetic advances in strain characterization, many studies have introduced methods for differentiating E. amylovora strains from a genetic basis (McManus and Jones, 1995; Rezzonico et al., 2011; McGhee and Sundin, 2012; Zeng et al., 2018; Singh and Khan, 2019). In a review published by Puławska and 1 Sobiczewski in 2012, the authors compared the current and historical findings of using techniques such as Restriction fragment length polymorphism (RFLP), Random amplification of polymorphic DNA (RAPD), Amplified fragment length polymorphism (ALFP), Multiple Locus Variable-number Tandem Repeat Analysis (MLVA), and clustered regularly interspaced short palindromic repeats (CRISPR) arrays and found that only MVLA and CRISPR array typing was able to show a high degree of separation between isolates (Puławska and Sobiczewski, 2012). Recently, sequencing whole genomes has become both cheaper and easier. Hence, more studies have been published identifying other repeat regions potentially useful for differentiating strains (Smits et al., 2017) In this mini review, I will examine topics regarding population dynamics of E. amylovora. Specifically, impacts of the environment, biological control agents and microbiomes, strain dynamics and distribution, and possible prospective directions will be discussed in this text. Impacts of Environment The role of environment in fire blight severity has been known since the 1930’s (Rosen, 1933; Thomas and Ark, 1934). Over the past hundred years of fire blight research, there have been many models proposed for predicting blossom blight severity and best timing of management intervention, many utilizing population dynamics. As for which weather factors are the most important for population growth, the only true consensus is that temperature plays a crucial role in growth (Billings, 1974; Taylor et al., 2003; Pusey and Curry, 2004; Pusey and Smith, 2008; Farkas et al., 2012). Though it is known that temperature in general is important, the critical component seems to be the average temperature over the course of bloom and not the 2 minimum or maximum daily temperatures (Pusey and Curry, 2004). In this study, Pusey and Curry tracked the population of E. amylovora over multiple days in with different temperature fluctuations and found no effect of wild temperature swings on the growth trends. Under ideal atmospheric conditions of 24°C with 100% relative humidity, less than 24 hr is needed for E. amylovora populations to build up to 107 on detached flower stigmas (Pusey and Curry, 2004). However, in general, relative humidity is thought to be less crucial, with multiple studies concluding that relative humidity directly is not a major factor in the rapid growth of E. amylovora (Thomson, 1986; Taylor et al., 2003). The other environmental factor heavily studied with population dynamics is flower phenology, mainly through flower age (Thompson and Gouk, 2003; Pusey and Curry, 2004; Dewdney et al., 2007). Younger flowers, open 1-3 days can support large populations of E. amylovora (Thomson and Gouk, 2003). Flowers that are 4-5 days old can also harbor large amounts of bacteria, however not as high as the freshly opened flowers (Thomson and Gouk, 2003; Pusey and Curry, 2004). Thomson and Gouk (2003) also found no difference in saprophytic bacteria populations between 3- to 6-day-old pistils, which led to the conclusion that flower age is more significant than the presence of other microbes. Thomson and Gouk observed flowers in the field and used weather inputs into MaryBlyt, however they did not use any statistical models to correlate weather to populations. For the entire study, EIPs from MaryBlyt indicated that conditions were ideal for infection. The study also rated for disease incidence, however there was no significant difference between flowers with different ages in the United States portion of the study (Thomson and Gouk, 2003). Different floral organs have been inoculated to determine the best growth sites of E. amylovora, with a consensus that the stigma is where the population mainly surges (Thomson, 1986; Rundle and Beer, 1997; Hasler and 3 Mamming, 2001). Not just apple flowers have been evaluated for differences in population dynamics, in 2006 Johnson et al observed E. amylovora populations on stigmas from 11 different plant genera. Using a constant temperature of 15°C, they found all but one genus from the rose family supported epiphytic populations of E. amylovora at rates comparable to apple stigmas (Johnson et al., 2006). Impact of Biological Control Agents and Microbiomes Bacterial antagonists The use of other microbes as a possible biocontrol for fire blight has been examined since at least the 1970s (Riggle and Klos, 1970; Riggle and Klos, 1972; Erskine, 1973). Part of confirming a potential biological control agent is to evaluate if the addition of these candidate microbes changes the population dynamics of the pathogen. Starting in 1970, Riggle and Klos (1970) began evaluating Pantoea agglomerans (previously referred to as Erwinia herbicola) as a potential biological control agent. Using differences in population, they were able to ascertain that P. agglomerans is a potential biological control agent for fire blight. Rundle and Beer (1987) further characterized the population dynamics of E. amylovora and P. agglomerans by observing populations on different organs over the whole flowers. They observed population shifts mainly on stigmas, and found that the growth was not as rapid with the addition of P. agglomerans (Rundle and Beer, 1987). Wilson and Sigee (1992) examined a strain of P. agglomerans isolated from hawthorn leaves against E. amylovora on hawthorn flower stigmas resulting in observing a decrease of growth rate of E. amylovora (Wilson and Sigee, 1992). Other strains of Pantoea agglomerans, such as Eh1087 can compete with E. amylovora if Eh1087 can establish before the pathogen arrives (Giddens et al., 2003). Other bacterial antagonists have been evaluated on the 4 basis of shifts in population dynamics, include Pseudomonas fluorescens and multiple Bacillus spp. (Wilson and Lindow, 1993, Pusey, 1997; Sundin et al., 2009). In 1997, Pusey developed a protocol for using crab apple flowers in growth chambers to evaluate population dynamics between E. amylovora and potential biological controls. This work is widely used to compare the potential efficacy of potential biological control agents (Pusey, 1997). The use of bacterial antagonists in tandem with antibiotics have also been studied, with limited effect. The consensus is still that conventional biocides, such as streptomycin, are superior to antagonistic bacteria at reducing E. amylovora populations, either directly though observing population dynamics or through fire blight incidence (Stockwell et al., 1996; Sundin et al., 2009) With the advent of high throughput sequencing what scientists know about the flower microbiome has changed rapidly over the past decade (Shade et al., 2013; Steven et al., 2018). The diversity of the microflora is immense, and simply applying a bacterial antagonist to flowers that are newly opened may not be fast enough to out complete the native bacteria in the system. A recent study by Cui et al. (2020) analyzed changes in the flower microbiome with the addition of E. amylovora and found that once inoculated E. amylovora grows rapidly on stigmas, outcompeting other species. Bacteriophages Another biocontrol agent that works by manipulating population dynamics are bacteriophages. Eve Billing first demonstrated that exopolysaccharides produced by E. amylovora could impact the ability of bacteriophage to infect cells (Billing, 1960). Other studies have touched on using bacteriophages as a biological control for fire blight (Hendry et al., 1967; Erskine, 1973; Richie and Klos, 1977; Schnabel et al., 1998). 5 In vitro analysis found that a phage cocktail was able to effectively reduce E. amylovora populations in a liquid broth (Schnabel et al., 1998). However, when the scientists took their cocktail to the field, the disease incidence was not as reduced as they had hoped suggesting that developing methodologies to enhance bacteriophage longevity in the field was needed (Schnabel et al., 1998). Roach et al. (2013) further quantified the impact of exopolysaccharides on bacteriophage pathogenesis and found that it was a factor effective for control. A recent study examined the interactions between E. amylovora, P. agglomerans, and bacteriophages that infect and kill both species. The group has been studying the use of P. agglomerans strains as the phage carrier to overcome some of the problems identified with phage application. This study used quantitative real-time PCR to quantify population shifts instead of playting. Using a novel cocktail of multiple bacteriophages, they were able to successfully infect P. agglomerans that was then able to reduce E. amylovora populations by over four log units in 24 hours (Gayder et al., 2020). This study lays the ground work for the group to work on flowers in field or studying population dynamics on stigmas. Antagonistic Yeasts Besides bacterial antagonists and bacteriophages, yeasts have also been screened as potential biological control agents (Pusey et al., 2009). One of most promising candidates is Aureobasidium pullulans, a black yeast-like fungus, a common epiphyte of plants, and has been shown to have antagonistic properties (Lima et al., 2003, Kunz, 2004, Duffy et al., 2006, Pusey et al., 2009). In 2004, a Kunz observed that this yeast reduced incidence of fire blight in apple orchards. This observation was also noted by Duffy et al. (2006), who found that A. pullulans exhibited significant fire blight control on seven cultivars. The study also evaluated application timing, and determined that applying A. pullulans 24–48 hrs prior to E. amylovora inoculation 6 reduced pathogen population growth on flowers (Duffy et al., 2006). Pusey et al., 2009 evaluated the protocol developed in 1997 for testing potential antagonists and found that the yeast colonized detached apple flowers better than other antagonistic yeasts while also reducing E. amylovora populations. Future Prospects Population dynamics related to dissemination and internal movement Slack et al., 2017 found that ooze droplets only contain between 9-14% of the population of internal tissues surrounding the ooze droplet immersion site, suggesting that more cells stay within the host and continue systemic spread rather than disseminate to potential new hosts (Slack et al., 2017). Future studies could be done expanding current knowledge of the internal dynamics of E. amylovora by identifying factors that interact with population movement. Tracking strains through spatial dynamics A study conducted by Wallis et al. (2020), also tracked the distribution and movement of E. amylovora through New York state. They utilized similar techniques presented in this work, however they also suggest developing an online platform plant pathogens in real time, modeled after the GenomeTrakr system developed by NCBI and FDA for tracking human pathogens. This would be a useful tool for tracking disease populations in real time. If strains can be associated with aggressiveness, streptomycin and other antibiotic resistance in real time, this could mean differences in orchard management of fire blight. 7 Conclusions Studies on population dynamics have contributed to our basic understanding of the underlying conditions that lead to pathogen grown on flowers. These studies have developed forecasting models, identified potential biological control agents (some have been commercialized), and other applied applications. Even in the past few years, the hunt for new potential biological control agents is still underway, and with advances in microbiome analysis these potential biologicals may be screened more thoroughly than ever before. The use of molecular techniques for differentiating genomes will enhance knowledge on strain spread, virulence, and potentially change the way we recommend growers treat their orchards. 8 CHAPTER 2 IN-ORCHARD POPULATION DYNAMICS OF ERWINIA AMYLOVORA ON APPLE FLOWER STIGMAS Abstract Populations of the fire blight pathogen Erwinia amylovora on apple flower stigmas were tracked over the course of apple bloom in field studies conducted between 2016 and 2020. Flower stigmas inoculated on the 1st day of opening can harbor large (107 cells / flower) populations of E. amylovora three to five days post-inoculation. However, populations inoculated on stigmas of flowers that were already open for three days only exceeded 107 cells / flower in one of five experiments, and populations inoculated on stigmas of flowers that were already open for five days never exceeded 104 cfu / flower. The incidence of fire blight disease symptoms was also significantly higher for flowers inoculated when open for 1 day. During this study, 10-fold increases in E. amylovora stigma populations in a 24-hr time period (population surges) were observed on 36.2%, 20.0%, and 8.0% of possible days on 1-day, 3-day, and 5-day open flowers, respectively. Population surges were also observed in atmospheric conditions with daily average temperatures below 14oC. The relative growth rate of E. amylovora on 1-day open flowers was as high as 12.8% per hour. Frequent sampling during days and at night revealed that many population surges occurred between 2200 and 0200 during the night. The incidence of fire blight on inoculated flowers that experienced low daily atmospheric temperatures (below an average of 14°C) was between 45-89%. This study refines our knowledge of E. amylovora population dynamics and further indicates that E. amylovora is able to infect flowers during exposure to colder field temperatures than previously reported. 9 Introduction Fire blight, a disease of rosaceous plants caused by Erwinia amylovora (Burrill), first appears in spring in the form of blossom blight. Blossom blight can be quite devastating, as both yield and tree health are directly impacted at this stage; infected flowers are killed and the pathogen cells have a direct route into the host tissue. Once systemic, E. amylovora cells move through the cortical parenchyma or via the vascular system into the trunk of the tree, and migrate to the rootstock:scion junction where cankers are formed that girdle trees and eventually cause host death (Norelli et al., 2003; Billing, 2011). The initial infection period of bloom is thought to be the best time for disease management, as flowers are only open and susceptible to bacterial infection for a short period of time (Norelli et al., 2003; Pusey and Curry, 2004; Pusey and Smith, 2008; Farkas et al., 2012), and E. amylovora cells are located on the external stigma surface. On apple, blossom blight disease symptoms (flower death) of fire blight are typically initiated in the spring on flowers as overwintering inoculum from cankers is disseminated to the surface of stigmas predominantly by insects (Norelli et al., 2003). The surface of the tips of stigmas particularly favors exponential growth and the massive buildup of E. amylovora in the intercellular spaces between columnar papillae cells as observed via scanning electron microscopy (Thomson and Gouk, 2003; Malnoy et al., 2012). Analysis of stigma exudates from apple flowers has shown the presence of approximately 2-3 g of sugars per flower (mostly fructose and glucose) that are exuded from papillae cells (Pusey et al., 2008). Under conducive weather conditions, E. amylovora populations can double rapidly, and populations of 106 to 107 cells per flower are common (Thomson, 1986; Pusey, 2000; Taylor et al., 2003). The availability of free moisture through rain or heavy dew is then required to facilitate the migration of motile 10 E. amylovora cells down the style where these cells infect flowers through natural openings present in the nectaries (Bubán and Orosz-Kovács, 2003; Farkas et al., 2012; Thomson, 1986). The role of environmental conditions in affecting blossom blight severity has been described since the 1930s (Rosen, 1933; Thomas and Ark, 1934). In an assessment of the growth rate of E. amylovora strains measured in vitro as a function of temperature, Billing (1974) showed that the relative growth rate was essentially the same between 21 and 30°C, and a considerable reduction in growth rate was observed below 18°C. Calculation of potential cell doublings per day by Billing (1978), with correction by Schouten (1987), indicated that E. amylovora cell populations could double approximately 1.9, 11.3, and 16.2 times during days with an average temperature of 8, 18, and 24°C, respectively. These calculations resulted in a slightly increased number of doublings for days with the same average temperature but higher maximum temperature, up to 28°C (Schouten, 1987). Cooler temperature effects on E. amylovora growth and infection potential have been less studied, but it was recently shown that E. amylovora can cause infection in vitro at temperatures as low as 4°C (Santander and Biosca, 2017). In a growth chamber experiment, E. amylovora cell populations inoculated to apple flower stigmas increased greater than 100-fold over a three-day period on trees incubated at 10°C, although the population increase was ca. 10,000-fold on trees incubated at 24°C (Dewdney et al., 2008b). Another critical component affecting E. amylovora growth on flower stigmas is flower age. A field inoculation experiment of apple flowers, aged 1-8 days, conducted for one year at two locations, showed that 1-3 day-old flowers could support growth of the inoculated E. amylovora strain, but 5-8 day-old flowers did not support an increase in E. amylovora population (Thomson and Gouk, 2003). Likewise, in a greenhouse experiment, Dewdney et al. (2008a) 11 observed a 3-log increase in the inoculated E. amylovora strain Ea273 on 1 day-old apple flower stigmas and observed a highly variable response of Ea273 populations on 5 day-old flower stigmas. Thomson and Gouk (2003), using culture-based methods, found no difference in saprophytic bacterial populations between 3- to 6-day-old pistils, which led to their conclusion that flower age is more significant than the presence of other microbes in affecting E. amylovora population size. Over the past century, many models have been proposed for predicting blossom blight severity and for identifying the best timings for management practices. Currently, the two mostly commonly-used predictive models for calculating fire blight risk at bloom in the United States are MARYBLYT and CougarBlight (Steiner, 1990; Smith, 1996; Smith and Pusey, 2011; Turechek and Biggs, 2015). Both models input weather data and flower age as factors for determining risk levels for fire blight infection, which, in MARYBLYT, is referred to as Epiphytic Infection Potential (EIP). MARYBLYT is modeled to predict both infection events and symptom development, whereas CougarBlight only predicts flower infection (Dewdney et al., 2007). Historically, threshold maximum temperatures of 18°C and 21-27°C during bloom were suggested to be critical boundaries with the potential for severe infection when these daily maximum temperatures were reached (Mills, 1955; Leupschen et al., 1961; Billing, 1980). Embedded within these studies has been the long-standing assumption that higher daily maximum temperatures (up to 28-30°C) during bloom provide the conditions enabling maximal growth of E. amylovora. Mean daily temperatures have been used in predictive models; for example, Thomson et al. (1982) used critical mean temperatures of 16.7°C for early bloom (early March) infection and 14.4°C for infections occurring during secondary (rattail) bloom (1 May) 12 on pear in California. This seasonally-dependent reduction in threshold temperature was later suggested by Johnson et al. (2006) to reflect the epiphytic buildup and abundance of E. amylovora inoculum increasing over time during bloom. Experimental results from growth chamber experiments have shown that the average daily temperature over the course of bloom and not the minimum or maximum daily temperatures is most important for disease occurrence (Pusey and Curry, 2004). Though there have been field studies focusing on stigma populations in the past, none have spanned the entire bloom event and considered various weather factors over multiple years and how these factors impact E. amylovora growth. The population dynamics of E. amylovora on stigmas under cooler temperatures in the field has also not been documented. We hypothesized that a multi-year field study of E. amylovora population dynamics on apple flower stigmas would enable us to more precisely understand the epiphytic behavior of this pathogen, refine disease prediction, and improve blossom blight disease management. In this study, we examined E. amylovora population dynamics on apple flowers of different ages from multiple cultivars. This enabled the monitoring of E. amylovora populations over time, as the flowers aged, and as weather conditions changed. We also assessed population dynamics at different times of day and night, revealing that the epiphytic growth of E. amylovora on flower stigmas occurs predominantly during the night. Materials and methods Population dynamics of E. amylovora on apple flower stigmas. All experiments were conducted on the apple cultivars ‘Fuji’, ‘Gala’, ‘Ginger Gold’, ‘Jonathan’, and ‘McIntosh’. The trees used for these field experiments are located at the Michigan State University Plant 13 Pathology Research Farm (42.689280, -84.484750) in East Lansing, MI or at the Northwest Michigan Horticultural Research Center (NWMHRC) (44.881996, -85.675251) in Traverse City, MI. Experiments were conducted between 2016 and 2019. Individual flowers were tagged at the pink phenological stage and inoculated with E. amylovora strain Ea110 (Zhao et al., 2005) on either the first day, third day, or fifth day after opening. Ea110 is a spontaneous rifampicin- resistant strain, facilitating re-isolation from field samples. Prior to field inoculation, strain Ea110 was cultured in LB broth overnight at 28°C, and populations were adjusted turbidimetrically (optical density at 600 nm) to 1x103-4 CFU l-1 using a Tecan Spark plate reader (Männedorf, Switzerland). The prepared inoculum was maintained on ice and inoculated within 1 hr onto flower stigmas at a rate of 1 l per flower. For analysis of stigmatic populations, at each sampling time point, 5 flowers were harvested from each replicate for a total of four replicates. With this there were 25 stigmas per replicate (5 stigmas per flower) and 100 per timepoint. Stigmas were harvested in the field with a sterilized razor blade and forceps, and then placed into 5 ml culture tubes containing 1 ml of 0.5 x phosphate-buffered saline solution. An initial re-isolation of inoculated flowers was conducted for a 0 hr time point, and the sampling was conducted every 24 hrs for 5 days. Stigma samples were kept on ice prior to arrival to the lab, then sonicated for 5 min (model 250T; VWR Scientific, Houston, TX), and placed back on ice. Plating methods were as described (Slack et al., 2021) with serial dilutions drop-plated onto LB agar plates amended with 100 g ml-1 rifampicin and 50 µg ml-1 cycloheximide. The addition of both rifampicin and cycloheximide aided in the re-isolation of E. amylovora Ea110 and in reduction of fungal contaminants. Plates were wrapped in aluminum foil and incubated at 24°C for 48 hr before colonies were counted. 14 The resulting colony counts were used to determine the population size of E. amylovora Ea110 over the course of the 5 day experimental time periods. Additional experiments were conducted with more frequent samplings. Stigmas of apple cv. Gala flowers that were open for 1 day were inoculated with E. amylovora Ea110 as described above. In one set of experiments, flower stigmas (four replicates) were sampled as described above at 2-3-hr intervals from 0800 to 1900 on the day after inoculation, and then at 48 hr after inoculation. This experiment was conducted in both 2017 and 2018. A second set of experiments were also conducted on flowers that were open for 1 day. In these experiments, flower stigmas were sampled overnight at 4-hr intervals starting on the second evening after inoculation (1800, 2200, 0200 (+1), 0600 (+1), 1000 (+1)). This experiment was conducted four times in 2019. Incidence of blossom blight infection was rated approximately 2-3 weeks after petal fall for all experiments conducted in 2017 to 2019. During the initial inoculation of trees, 20 additional flowers were inoculated and used for the disease incidence rating to determine rates of infection for each treatment. Data calculations and statistical analyses. We used the formula from Johnson et al. (2006) to calculate relative growth rates (r) of E. amylovora from 24-hr periods when population surges of at least 10 fold were observed. Additionally, equations formulated from Billing (1974) were used to determine the number of generations and the generation time of E. amylovora during population surges. All statistical analyses were performed with the program RStudio for statistical computing (R Core Team, 2013). Hourly weather data were retrieved from the Enviroweather monitoring system stationed at the Michigan State University Horticulture farm (HORT) or the Traverse City weather station (NWMHRS). The HORT weather station was located within 2 km 15 of the test orchards; the NWMHRS station was less than 1 km from the inoculated trees. Weather factors measured hourly were downloaded from Enviroweather for temperature, relative humidity, precipitation, and windspeed. The weather variables were analyzed using the R package “Corrplot” to calculate Pearson correlations between the weather variables and E. amylovora populations for 24-hr periods when 10-fold increases in populations were observed (Wei et al., 2017). While calculating Pearson correlations, P-values were adjusted using Bonferroni correction (Napierala, 2012). Results Population dynamics of E. amylovora on apple flower stigmas. Population dynamics series experiments on 1-day, 3-day, and 5-day open flowers were conducted five times over three years, and an additional 21 experiments were conducted on 1-day open flowers over five years. Erwinia amylovora populations on 1-day open flowers reached the carrying capacity of apple flower stigmas (> 107 cfu / flower) in 10 of 25 experiments, and were greater than 106 cfu / flower in 18 of 25 experiments conducted over a five year period (Table 1). In contrast, E. amylovora populations on 3-day open flowers only exceeded 106 cfu / flower in 1 of 5 experiments, and populations only exceeded 104 cfu / flower in 1 of 5 experiments on 5-day open flowers (Table 1). Data from two representative sets of experiments conducted in 2017 on 1-, 3-, and 5-day open flowers are shown in Figure. 1. The 1-day open flowers from cultivars ‘Jonathan’ and ‘Gala’ were inoculated on 26 April and 27 April 2017, respectively. Populations surged quickly on ‘Jonathan” flower stigmas (~ 100-fold increase within 24 hr) and reached the carrying capacity by the fifth day of sampling (Figure 1). On 3-day open ‘Jonathan’ stigmas, the E. 16 amylovora Ea110 population increased about 10-fold over the first two days, and then decreased after that, never reaching the flower carrying capacity, and the population remained constant over the 5-day sampling period on 5-day open flowers (Figure 1). On 1-day open ‘Gala’ flowers, populations rose ca. 100-fold over the first two days of sampling, then were reduced and rebounded, but did not reach the flower carrying capacity. Populations on 3- and 5-day open ‘Gala’ flowers remained constant or were reduced over the five day sampling period (Figure 1). Weather parameters were compiled for the days during which all flower population data sets were generated, and selected weather parameters associated with one specific sampling period are shown in Figure 2. Air temperature was warmest on the first day of the experiment, when only 1-day open ‘Jonathan’ flowers were inoculated. A rain event in the middle of the experiment was associated with declines in populations in several of the treatments. The relative humidity was > 60% throughout the entire sampling period, and wind speeds dropped below 2 m s-1 during each overnight period (Figure 2). Analysis of the 25 experiments conducted on 1-day open flowers revealed that inoculation of 103 to 104 cells of E. amylovora on 1-day open flowers was typically followed by a lag period, as populations similar to the inoculated total were recovered on the next sampling day (24 hr after inoculation) in 21 of these experiments. The exceptions to this observation where growth was observed on the first day were on warm days (24-hr avg. air temperature > 18oC) or days with a precipitation event. Following this lag period, growth occurred over days 2-5 of the sampling period. We defined surges in E. amylovora growth as 24-hr periods in which increases in population size of greater than 10-fold were observed. We observed population surges on at least one day in 24 of the 25 experiments, and in 5 experiments, population surges were observed on three of the five days of the experiment (Table 1). Decreases in population size in a 24-hr 17 period were also observed in 19 of the 25 experiments (27 total events) conducted in 1-day open flowers; these decreases in population were associated with rain events or average daily relative humidity of under 40% in 25 of the 27 events (Table 1). As defined by the relationships from Johnson et al. (2006) and Billing (1974), growth surges and also calculated the number of generations and generation time in hours (Billing, 1974). Growth rates during population surges ranged from 3.8 to 12.8% hr-1, and the number of generations occurring during a 24-hr population surge was as high as 10.2 (generation time of 2.4 hr) (Table 2). 18 Figure 2-1. Population dynamics of Erwinia amylovora Ea110 from four representative experiments on 1-day, 3-day, and 5-day open flowers of apple cultivars ‘Gala’ and ‘Jonathan’ over a 5-day experimental period. Mean log10 populations ± the standard error of the mean are shown. Daily samplings were conducted at 1000. 19 Figure 2-2. Population dynamics of Erwinia amylovora Ea110 from two representative experiments on 1-day, 3-day, and 5-day open flowers of apple cultivars ‘Gala’ and ‘Jonathan’ over a 5-day experimental period. Mean log10 populations ± the standard error of the mean are shown. Daily samplings were conducted at 1000. Weather parameters, temperature, precipitation, relative humidity, and windspeed are also shown. 20 020406080100120Rel. H. (%)01234567Percip. (cm)0123456781:008:0015:0022:005:0012:0019:002:009:0016:0023:006:0013:0020:003:0010:0017:0024:00:007:0014:0021:004:0011:0018:001:008:0015:0022:005:0012:0019:002:009:0016:0023:006:0013:0020:004/26/174/27/174/28/174/29/174/30/175/1/175/2/175/3/175/4/175/5/175/6/17Windspeed (m/s)0246810E. amylovora population (Log10 CFU)Jonthan - 1 DayJonathan - 3 DaysJonathan - 5 DaysGala - 1 DayGala 3 DaysGala 5 Days0102030Atmospheric Temp. (C)Fig. 2-2. Population dynamics of Erwinia amylovora Ea110 from two representative experiments on 1-day, 3-day, and 5-day open flowers of apple cultivars ‘Gala’ and ‘Jonathan’ over a 5-day experimental period. Mean log10populations ±the standard error of the mean are shown. Daily samplings were conducted at 1000. Weather parameters, temperature, precipitation, relative humidity, and windspeed are also shown. Population Decreasing populations in 24-hr period e Final surges d population (no. of days) (no. of days) Rain days RH < 40% days __________________________________ Flower Initial Age b population c Table 2-1. Results from field experiments tracking the population dynamics of Erwinia amylovora on apple flower stigmas where experiments were initiated on different-aged flowers. _______________________________________________________________________________ Expt.a _______________________________________________________________________________ 1 16-F 2.9 ± 0.1 7.2 ± 0.1 3 1 1 0 1 3.5 ± 0.2 7.4 ± 0.2 2 1 0 0 16-F 1 16-G 2.9 ± 0.5 7.0 ± 0.6 2 1 0 1 16-G 1 2.9 ± 0.3 7.5 ± 1.0 3 0 0 0 16-G 1 3.2 ± 0.2 6.9 ± 0.3 2 1 0 0 16-G 1 4.2 ± 0.1 7.3 ± 0.2 2 0 0 0 1 16-J 3.8 ± 0.1 7.3 ± 0.2 3 1 1 0 17-G 1 5.7 ± 0.1 6.0 ± 0.2 2 2 2 0 1 4.3 ± 0.2 6.0 ± 0.5 2 2 2 0 17-G 1 4.2 ± 0.1 7.3 ± 0.2 2 1 0 1 17-G 1 4.2 ± 0.2 6.0 ± 0.5 1 1 1 0 17-G 17-G 1 3.0 ± 0.6 7.1 ± 0.1 1 0 0 0 1 5.9 ± 0.2 7.2 ± 0.2 1 2 2 0 17-G 1 17-J 4.6 ± 0.1 8.0 ± 0.5 2 2 2 0 17-J 1 2.2 ± 0.2 6.4 ± 0.5 3 1 1 0 1 2.3 ± 0.2 7.1 ± 0.1 3 0 0 0 17-J 1 4.3 ± 0.1 5.4 ± 0.3 2 2 2 0 17-M 1 3.8 ± 0.1 5.7 ± 0.4 2 2 2 0 17-M 17-M 1 3.8 ± 0.1 5.5 ± 0.2 2 2 2 0 1 4.3 ± 0.1 4.5 ± 0.4 2 2 2 0 17-M 1 2.8 ± 0.1 5.7 ± 0.5 1 0 0 0 18-G 18-G 1 4.0 ± 0.2 6.1 ± 0.1 1 0 0 0 18-G 1 4.1 ± 0.1 6.1 ± 0.3 1 0 0 0 19-GG 1 4.4 ± 0.1 4.3 ± 0.3 0 1 1 0 19-GG 1 3.6 ± 0.1 5.0 ± 0.2 1 0 0 0 19-J 1 4.1 ± 0.1 3.5 ± 0.4 0 1 1 0 1 4.3 ± 0.1 6.4 ± 0.1 2 0 0 0 19-M 20-F 1 1.8 ± 0.6 5.7 ± 0.3 2 1 0 1 16-F 16-G 16-J 17-G 17-J 16-F 16-G 16-J 3 3.4 ± 0.1 5.8 ± 0.2 1 1 1 0 3 3.6 ± 0.2 5.6 ± 0.8 2 2 2 0 3 3.2 ± 0.2 7.9 ± 0.3 2 0 0 0 3 4.6 ± 0.3 2.1 ± 0.1 0 2 2 0 3 4.1 ± 0.1 4.8 ± 0.6 0 2 2 0 5 3.8 ± 0.2 4.0 ± 0.8 0 1 2 0 5 4.1 ± 0.1 3.4 ± 0.8 1 4 2 0 5 3.3 ± 0.4 3.7 ± 0.6 1 2 1 0 21 5 4.1 ± 0.1 3.3 ± 0.5 0 3 2 0 5 2.1 ± 0.1 2.1 ± 0.1 0 0 0 0 Table 2-1 (cont’d). 17-G 17-J _____________________________________________________________________________ a Experiment. 16-20 refer to years the specific experiment was conducted (2016-2020). F, ‘Fuji’; G, ‘Gala’; GG, ‘Ginger Gold’; J, ‘Jonathan’; M, ‘McIntosh’. b 3, 3-day open; 5, 5-day open. c stigmas per flower. d increased by 10-fold or more. e was observed compared to the previous day, and if these decreases were associated with days with measurable precipitation or days where the average relative humidity was less than 40%. Decreasing populations tabulates the number of days in which a reduction in population Flower age refers to the age of flowers when experiments were initiated. 1, 1-day open; Populations are in log10 colony-forming units per flower, and were isolated from five Population surges tabulates the number of days where the E. amylovora population 22 Flower Initial Final Number of Generation growth rate Population surges __________________________________________ Relative Table 2-2. Analysis of population surge events (daily increase of log10 0.9 or greater) observed in the field for Erwinia amylovora on 1-day and 3-day open flowers including the number of generations and relative growth rates observed in the respective 24-hr periods. ___________________________________________________________________ Expt.a Ageb populationc population generationsd time (hr) (% / hr)e ___________________________________________________________________ 1 16-F 1 16-F 1 16-F 1 16-F 1 16-G 1 16-G 1 16-G 1 16-G 1 16-J 1 16-J 1 17-F 1 17-G 1 17-G 1 17-J 1 17-J 1 17-J 1 17-J 1 17-J 1 17-J 1 17-J 1 17-J 1 17-J 1 17-M 1 17-M 1 17-M 1 17-M 1 18-F 1 18-G 18-G 1 19-GG 1 19-GG 1 1 19-M 20-F 1 16-F 16-G 3.3 ± 0.4 4.7 ± 0.4 4.7 ± 0.4 5.2 ± 0.2 5.8 ± 0.2 4.7 ± 0.4 6.1 ± 0.4 4.7 ± 0.4 5.1 ± 0.2 5.9 ± 0.2 5.7 ± 0.3 7.2 ± 0.1 5.0 ± 0.2 4.8 ± 0.1 6.2 ± 0.1 3.1 ± 0.2 5.1 ± 0.3 6.6 ± 0.2 3.6 ± 0.1 8.3 ± 0.1 2.6 ± 1.6 5.7 ± 0.5 10.2 ± 1.1 2.4 ± 0.5 12.8 ± 0.5 5.6 ± 0.9 7.0 ± 0.2 4.6 ± 0.6 5.2 ± 0.3 5.7 ± 0.1 6.5 ± 0.6 7.5 ± 0.2 3.3 ± 0.4 7.3 ± 0.2 4.1 ± 0.1 3.8 ± 0.3 6.0 ± 0.1 7.2 ± 0.2 3.3 ± 0.1 9.0 ± 0.1 4.6 ± 0.6 5.6 ± 0.2 3.2 ± 0.4 7.4 ± 0.2 4.1 ± 0.1 5.6 ± 0.2 7.7 ± 0.3 7.1 ± 0.2 3.4 ± 0.1 8.9 ± 0.1 3.0 ± 0.6 4.2 ± 0.2 4.2 ± 0.4 5.7 ± 0.2 5.3 ± 0.1 4.2 ± 0.2 5.1 ± 0.1 3.2 ± 0.1 7.5 ± 0.1 4.0 ± 0.1 3.5 ± 0.6 6.2 ± 0.4 5.4 ± 0.5 4.5 ± 0.2 6.8 ± 0.2 4.5 ± 0.1 5.7 ± 0.3 3.9 ± 0.2 6.2 ± 0.1 4.9 ± 0.1 5.1 ± 1.1 6.4 ± 0.2 4.3 ± 0.7 5.6 ± 0.3 5.3 ± 0.1 6.4 ± 0.2 7.4 ± 0.5 3.3 ± 0.3 7.2 ± 0.1 4.2 ± 0.2 2.2 ± 0.2 3.2 ± 0.3 3.4 ± 0.2 7.1 ± 0.1 4.2 ± 0.1 3.2 ± 0.3 4.7 ± 0.6 5.1 ± 0.4 4.7 ± 0.2 6.4 ± 0.3 3.7 ± 0.6 7.0 ± 0.2 7.4 ± 0.4 3.2 ± 0.2 9.3 ± 0.1 2.3 ± 0.2 3.4 ± 0.1 3.6 ± 0.1 6.8 ± 0.1 4.5 ± 0.1 3.4 ± 0.1 5.3 ± 0.3 6.3 ± 0.2 3.8 ± 0.1 8.0 ± 0.1 5.3 ± 0.3 6.8 ± 0.1 4.8 ± 0.2 5.0 ± 0.1 6.0 ± 0.1 3.8 ± 0.1 4.8 ± 0.1 3.3 ± 0.1 7.2 ± 0.1 4.2 ± 0.1 4.2 ± 0.4 5.5 ± 0.2 4.5 ± 0.3 5.3 ± 0.1 5.6 ± 0.1 4.1 ± 0.2 5.7 ± 0.4 5.2 ± 0.2 4.6 ± 0.1 6.5 ± 0.1 3.8 ± 0.3 5.8 ± 0.1 6.5 ± 0.2 3.7 ± 0.1 8.2 ± 0.1 4.2 ± 0.2 7.1 ± 0.1 9.3 ± 0.1 2.6 ± 0.1 11.7 ± 0.1 2.9 ± 0.1 3.8 ± 0.2 3.0 ± 0.1 8.0 ± 0.1 3.8 ± 0.1 3.8 ± 0.2 5.7 ± 0.5 6.3 ± 0.3 3.8 ± 0.1 7.9 ± 0.2 5.4 ± 0.1 6.4 ± 0.1 3.3 ± 0.1 7.2 ± 0.1 4.2 ± 0.1 3.7 ± 0.2 5.0 ± 0.2 4.1 ± 0.2 5.8 ± 0.1 5.2 ± 0.1 4.3 ± 0.1 5.4 ± 0.1 3.5 ± 0.1 6.8 ± 0.1 4.4 ± 0.1 4.4 ± 0.9 6.3 ± 0.2 6.4 ± 0.6 3.7 ± 0.3 8.0 ± 0.1 3 3 3.9 ± 0.2 6.2 ± 0.2 7.6 ± 0.2 3.1 ± 0.1 9.6 ± 0.1 4.0 ± 0.3 6.0 ± 0.3 6.6 ± 0.3 3.6 ± 0.1 8.3 ± 0.1 23 3 3 3 3 Experiment. 16-20 refer to years the specific experiment was conducted Flower age refers to the age of flowers when experiments were initiated. 1, 4.3 ± 1.3 5.6 ± 0.9 4.2 ± 1.0 5.7 ± 0.4 5.3 ± 1.0 3.2 ± 0.2 5.1 ± 0.7 6.4 ± 0.3 3.8 ± 0.1 8.0 ± 0.3 5.1 ± 0.7 7.0 ± 0.1 6.2 ± 0.4 3.9 ± 0.2 7.8 ± 0.1 3.7 ± 0.3 4.8 ± 0.7 3.7 ± 0.4 6.2 ± 0.2 4.7 ± 0.4 Table 2-2 (cont’d). 16-G 16-J 16-J 17-J ___________________________________________________________________ a (2016-2020). F, ‘Fuji’; G, ‘Gala’; GG, ‘Ginger Gold’; J, ‘Jonathan’; M, ‘McIntosh’. b 1-day open; 3, 3-day open; 5, 5-day open. c from five stigmas per flower. d calculated using the formula e (2006) using the formula: r = 1/24 • [ln(y24) – ln(yzero)] where yzero and y24 are observed population means at the beginning and end of a 24-hr period where a surge in population was observed. Populations are in log10 colony-forming units per flower, and were isolated The number of generations per day for each population surge event was The relative growth rate per hour (r) was calculated from Johnson et al. 24 Identification of weather parameters most associated with effects on E. amylovora growth on apple flower stigmas and blossom blight disease incidence. We selected 15 (out of 25) representative 24-hr population surge events on 1-day open flowers, and tabulated daily averages and maximum and minimum values for temperature, relative humidity, and windspeed occurring during these time periods (Table 3). The average temperature recorded during 24-hr periods with population surges ranged from 8.7 to 23.0°C, and average daily relative humidity ranged from 41.0 to 88.2% (Table 3). The minimum wind speed recorded during the 24-hr periods ranged between 0.0 and 1.7 m s-1 (Table 3). The average daily temperature was not consistently high during days when the E. amylovora growth rate was high. For example, we observed a growth rate of 10.3% per hr on ‘McIntosh’ flowers when the average daily temperature was 8.7°C which exceeded the growth rates of 6.4% and 8.0% per hr on ‘Jonathan’ flowers when the average daily temperature was 23.0°C (Table 3). The daily maximum temperature exceeded the reported threshold maximum temperatures of 18°C and 21-27°C associated with the potential for severe infection (Mills 1955; Leupschen et al. 1961; Billing 1980) in only 6 of 15 datasets, and, in 5 of 8 instances, daily growth rates above 5.9% per hr occurred on days with maximum temperatures below 14.8°C (Table 3). A calculated Pearson’s correlation coefficient value was used to numerically discern the relationship between population surges and specific weather variables (Table 4). Coefficient values for weather variables with a P-value < 0.05 were considered significant. Using this calculation, we determined that increases in atmospheric temperature and relative humidity along with decreases in wind speed were the three weather variables that significantly affected E. amylovora population surges (Table 4). Other weather variables assessed, leaf wetness and solar radiation, did not significantly affect population surges. 25 Blossom blight disease incidence was determined for 12 experiments conducted between 2017 and 2019 on 1-day open flowers, and ranged from 30% to 100% of inoculated flowers (Table 5). For these experiments, we also tabulated the ending E. amylovora population five days after inoculation, and then for days 5 to 14 after inoculation, we quantified the number of days with: maximum air temperature > 18oC, average temperature > 16.7°C [critical mean temperature for early bloom infection (Thomson et al. 1982)], rainfall  0.25 mm, average RH > 85%, and overnight air temperature falling within 2.0oC of the dewpoint (Table 5). We observed 64% blossom blight disease incidence in one experiment where there were 0 days (5-14 days after inoculation) with a maximum air temperature > 18°C and average temperature > 16.7°C, and observed blossom blight incidence ranging from 45 to 89% in three other studies in which only one day occurred (5-14 days after inoculation) with a maximum air temperature > 18oC and average temperature > 16.7°C (Table 5). Similarly, days with average RH above 85% did not occur 5-14 days after inoculation in six of the experiments with blossom blight incidence ranging from 40 to 89% (Table 5). We did observe 3-5 days with rainfall greater than 0.25 mm in the period from 5 to 14 days after inoculation in all of the experiments, and also observed at least six overnight conditions that were likely suitable for dew formation (Table 5). 26 Table 2-3. Weather parameters associated with population surges of Erwinia amylovora during growth on 1-day open apple flower stigmas. Cultivar Growth rate in 24 hours Total Pcpn (mm) % RH high % RH low % RH Avg Average temp (C∘ hourly average) High temp C∘ Low temp C∘ Average Windspeed (m/s) Windsp eed High (m/s) Windspe ed Low (m/s) Jonathan 2.6 Fuji Gala Gala Fuji 5.9 12.9 5.2 12.2 McIntosh 10.3 Gala 3.3 McIntosh 8.2 Jonathan 6.4 Jonathan 8.0 Gala 12.0 G. Gold 4.2 McIntosh 4.2 Jonathan 3.1 Fuji 4.5 0 0 0 0 0 0 0 0 0 0 0 0 0 8.1 0 48 82 57.4 61.9 54.1 70.9 76.2 76.2 60 60 81.5 55.4 89.1 95.6 96.3 33.2 37.4 47.7 34.2 33.8 68.9 60.5 60.5 43.5 43.5 48 74.7 68.3 80 61.1 41.0 57.4 53.7 47.5 46.4 70.1 70.4 70.4 53.3 53.3 62.3 80.5 80.9 88.2 76.4 10.8 9.6 12.2 13.4 14.8 8.7 13.3 13.3 23.0 23.0 20.3 8.3 17.7 18.2 19.6 27 13.6 14 11.6 15.7 17.7 8.9 14.2 14.2 25.4 25.4 22.6 12.1 19.8 20.5 23.3 8.7 5.4 11.6 11.3 12.2 8.5 12.5 12.5 21.7 21.7 22.6 6.5 17 15.7 15.7 1.9 3.1 1.8 1.4 1.6 2.2 3.25 3.25 3.3 3.3 3.8 2.5 1.9 2.4 2.2 4.6 4.8 3.5 2.3 3.8 3.8 4.9 4.9 5.6 5.6 6.0 3.1 2.7 4.4 3.5 0.2 1.7 0.4 0.2 0.2 0.2 1.6 1.6 1.5 1.5 1.5 1.2 0.3 0 0.2 Table 2-4. Calculated Pearson correlations of weather variables with overnight growth of Erwinia amylovora during days with population surges on apple flower stigmas. Correlations shown are statistically significant (P < 0.05). Positive correlations are depicted with positive value, and negative correlations have a negative value. Values range between -1 and 1; the stronger the correlation, the closer the value will be to 1 or -1, depending on the type of relationship. For this table, if a correlation is positive, then growth increased as the weather variable value increased. ________________________________________________________________ _________________________________________________________________ Population surges ________________________________________________________________ ______________________________________________ Average atmospheric Average Temperature (oC) windspeed (m s-1) Average %RH Pearson correlations -0.71 0.38 0.22 28 with the following conditionsc Number of days (from 5 to 14 days after inoculation) ________________________________________________ Overnight air temp. temp. temp. Rainfall Avg. RH within 2.0 oC > 18oC > 16.7oC  0.25 mm > 85% of dewpoint E. amylovora Blossom blight Max. Avg. population incidence Table 2-5. Incidence of blossom blight infection on apple flowers previously inoculated with Erwinia amylovora at the 1-day open age. Weather parameters with the potential to affect growth and infection are also listed. _______________________________________________________________________________ Expt.a on day 5b (%) _______________________________________________________________________________ M G M J J GG GG M J F M M _______________________________________________________________________________ a F, ‘Fuji’; G, ‘Gala’; GG, ‘Ginger Gold’; J, ‘Jonathan’; M, ‘McIntosh’. b E. amylovora populations recorded on day 5 after inoculation are in log10 colony-forming units per flower, and were isolated from five stigmas per flower. daily temperature from hourly measurements; Avg. RH is the average daily relative humidity from hourly measurements. 5.4 ± 0.3 89 1 1 4 0 7 5.3 ± 0.5 45 1 1 4 0 7 4.5 ± 0.4 77 1 1 4 0 7 7.4 ± 0.5 64 0 0 4 0 8 7.1 ± 0.1 75 4 4 5 2 11 4.3 ± 0.3 30 3 3 4 2 14 5.0 ± 0.2 35 5 5 4 2 14 6.4 ± 0.1 65 6 6 5 5 14 3.5 ± 0.4 100 6 6 4 5 14 5.7 ± 0.3 57 11 11 5 1 12 5.7 ± 0.4 40 3 3 3 0 6 5.5 ± 0.2 50 3 3 3 0 6 Abbreviations: Max. temp. is the maximum daily temperature; Avg. temp. is the average 29 Population dynamics assessment of E. amylovora on apple flower stigmas with frequent samplings. To better understand the growth dynamics of E. amylovora on apple flower stigmas, we conducted experiments on 1-day open ‘Gala’ flowers in 2017 and 2018 at the NWMHRC, using a sampling frequency of every 2-3 hr during the day with a 13-14 hr gap between the last evening sample and the next sample the following morning. In 2017, we observed a ca. 10-fold increase in population between 1500 and 1900, with an additional increase of over 100-fold between 1900 and 0800 the next morning (Figure 3). In 2018, we observed a ~ 10-fold increase in population between 1400 and 2000, with an additional increase of slightly greater than 10-fold occurring between 2000 and 1000 the next morning (data not shown). These experimental results led us to conduct a further series of four experiments in 2019 using overnight sampling to identify timings during the night in which E. amylovora population size was increasing. For these experiments, 1-day open flowers were utilized, and intensive sampling was initiated on the second overnight period after inoculation to account for the lag phase in growth which was typically observed on the first day after inoculation. In experiment 1, a 10-fold increase in population size was observed in samplings between 1800 and 0200 (Figure 3A), and in the other three experiments, increases in population size of 10-fold or greater were observed in samplings between 2200 and 0600 (Figure 4B - 4D). In experiment 3, a 10-fold increase was observed between 2200 and 0200 (Figure 4C). We also graphed the air temperature and dew point during the overnight hours of these experiments, which indicated that the air temperature was close to the dewpoint during the overnight hours in two of the four experiments (Figure 5A and 5C), and that the air temperature was within 1°C of the dewpoint between 2200 and 0200 in the other two experiments (Figure 5B and 5D). 30 Figure 2-3. Population dynamics of Erwinia amylovora Ea110 on 1-day open flowers of apple cultivar ‘Gala’ over a 42-hr experimental period with frequent samplings during daylight hours of the second day of the experiment. Mean log10 populations ± the standard error of the mean are shown. 31 0.001.002.003.004.005.006.007.008.00Erwiniaamylovoralog10 CFU140018000800100013001500170019000800Fig. 2-3. Population dynamics of ErwiniaamylovoraEa110 on 1-day open flowers of apple cultivar ‘Gala’ over a 42-hr experimental period with frequent samplings during daylight hours of the second day of the experiment. Mean log10populations ±the standard error of the mean are shown. Figure 2-4. Population dynamics of Erwinia amylovora Ea110 on 1-day open flowers of apple cultivar ‘Gala’ over a 47-hr experimental period with frequent samplings during overnight hours of the second day of the experiment. Mean log10 populations ± the standard error of the mean are shown. Red asterisks highlight the 2200 and 0200 (+1) sampling time points. 32 012345618:0019:0021:0023:001:003:005:007:009:0011:0013:0015:0017:005/17/195/18/195/19/19Erwiniaamylovoralog10(CFU/sigma0123456718:0019:0021:0023:001:003:005:007:009:0011:0013:0015:0017:005/15/195/16/195/17/19Erwiniaamylovoralog10(CFU/sigma012345618:0019:0021:0023:001:003:005:007:009:0011:0013:0015:0017:005/11/195/12/195/13/19Erwiniaamylovoralog10(CFU/sigma012345618:0019:0021:0023:001:003:005:007:009:0011:0013:0015:0017:005/9/195/10/195/11/19Erwiniaamylovoralog10(CFU/sigma)ABCD********Fig. 2-4. Population dynamics of ErwiniaamylovoraEa110 on 1-day open flowers of apple cultivar ‘Gala’ over a 47-hr experimental period with frequent samplings during overnight hours of the second day of the experiment. Mean log10populations ±the standard error of the mean are shown. Red asterisks highlight the 2200 and 0200 (+1) sampling time points. Figure 2-5. Temperature and dewpoint readings for the overnight periods of four experiments monitoring the population dynamics of Erwinia amylovora on 1-day open apple flower stigmas with frequent overnight sampling time points. 33 ABDCFig. 2-5. Temperature and dewpoint readings for the overnight periods of four experiments monitoring the population dynamics of Erwiniaamylovoraon 1-day open apple flower stigmas with frequent overnight sampling time points. Discussion Our analysis of the population dynamics of E. amylovora on apple flower stigmas under field conditions has shown that: (i) growth is most robust on younger flowers; (ii) the most important weather parameters affecting growth on stigmas are average daily temperature, average daily relative humidity, and average windspeed; (iii) short-term large increases (> 10- fold) in population size can occur when average daily temperatures are considered suboptimal by widely-used current fire blight disease prediction models; and (iv) surges in population growth occur at night. The factor that most favored population growth of E. amylovora to the flower carrying capacity was flower age and, as expected, the more favorable growth conditions on 1- day open flowers were correlated with an increased chance of infection. We generated a total of 25 datasets tracking E. amylovora population dynamics on apple flower stigmas from 1-day open flowers over 5-day periods, and noted that surges, increases of greater than 10-fold in population over a 24-hr period, and growth to carrying capacity of the flower occurred in 92% and 38.5% of the experiments, respectively. In contrast, growth to carrying capacity was only observed in 1 of 5 experiments using 3-day open flowers and 0 of 5 experiments using 5-day open flowers. Our field results of population dynamics of E. amylovora on apple flower stigmas of different aged flowers conducted over five years are similar to previous field results from Thomson and Gouk (2003), combining one year of results from Utah with one year of results from New Zealand, and greenhouse results from Dewdney et al. (2008a). As flowers age, the quantity of free sugars, mostly glucose and fructose, increased from 1.9 g / flower at early anthesis (1 day open) to 3.2 g / flower at late anthesis (4-5 days open), and that polysaccharide (mostly arabinose and galactose), ranged from 17 to 33 g / flower at late anthesis (Pusey et al. 2008). However, flower aging is also accompanied by the collapse of 34 papillae cells accompanied by extensive stigmatic secretions (Thomson and Gouk 2003). These modifications to the stigma surface appear to affect new colonization by E. amylovora, and may block access to preferred sites for growth. However, our data do show that E. amylovora populations that were established on 1-day open flowers can continue to grow between the fourth and fifth day after inoculation. A recent study by Cui et al. (2020) has shown that the diversity of the apple flower stigma microbiome decreases as flowers age, and predominantly consists of Enterobacteriaceae and Pseudomonadaceae in 5 day open flowers. These results might also indicate that immigrant bacteria in general have difficulty in colonizing older flowers, and that Enterobacteriaceae and Pseudomonadaceae are the most competitive bacteria in the preferred growth sites of the apple flower stigma. Our work tracking the population dynamics of E. amylovora on 1-day open flowers over four years encompassed a wide range of environmental conditions and allowed for identification of optimal timings for growth and the critical environmental parameters that significantly affect growth when the physical status of the stigma habitat was optimal. Historically, most studies of E. amylovora populations on flowers have focused on temperature as the most important environmental factor affecting growth and/or the potential for blossom blight infection (Mills, 1955; Leupschen et al., 1961; Billing, 1980; Thomson et al., 1982; Thomson, 1986; Thomson and Gouk, 2003). While these previous studies focused on the importance of warmer temperatures driving blossom blight infection, there have been a few studies indicating that E. amylovora growth (Dewdney et al., 2008b) and infection (Santander and Biosca, 2017) can occur at colder temperatures (ex. 4°C and 10°C) under controlled conditions. While we did identify temperature as a significant factor driving E. amylovora growth, particularly affecting 24-hr 35 surges in population size, these surges occurred many times when the average daily temperatures were below previously-identified thresholds. In most situations where population surges occurred under colder conditions, blossom blight disease was also observed. This result suggests that critical temperature minimums or average daily temperature thresholds utilized by fire blight disease prediction programs may be too high and therefore underestimate the disease risk at lower temperatures. It should be noted, however, that we were studying flowers that were inoculated with 103-4 cfu of E. amylovora cells. Normally, under cool to cold conditions (average daily temperature < 10-14°C), ooze droplets may not be present on overwintered cankers, and insect activity would be much reduced, significantly decreasing the chances for delivery of the pathogen to flowers. Thus, the important consideration from the stand point of disease prediction would be growth potential on cool to cold days that follow a warmer period such that E. amylovora populations on flowers would already be established. We found that rain events were mostly associated with 24-hr decreases in E. amylovora populations on apple flower stigmas. This is not surprising as rain splash can disperse bacterial cells to other flowers, and also because rain facilitates the movement of E. amylovora cells down the style to the hypanthium. Thus, even though reductions in population were observed, we recorded a comparable level of blossom blight incidence in experiments impacted by rain. In a small number of cases, we also observed reductions in E. amylovora population size occurring on days where the average RH was below 40%. Pusey (2000) has shown the relationship between RH and E. amylovora population size on apple flowers maintained in a growth chamber with results indicating that E. amylovora reached the highest population densities when the RH was 80-100%. 36 Our observation of E. amylovora growth on flower stigmas occurring at night, specifically between 10 PM and 2 AM is interesting because the optimal growth temperature for this organism observed in vitro is close to 28°C (Billing, 1974). Growth at night does validate the finding of Pusey and Curry (2004) from growth chamber studies that showed the importance of the contribution of the average daily temperature to E. amylovora growth rather than the daily temperature range. But how do environmental conditions positively impact E. amylovora growth, at times of day when the atmospheric temperature is declining? A major clue could come from our determination that declining average wind speeds and rising relative humidity are significantly associated with increasing populations of E. amylovora (Table 4). We hypothesize that dew formation on apple flower stigmas occurring during the night provides free moisture that stimulates bacterial growth. Rosaceae family stigmas are characterized as “wet”, with “abundant, free-flowing surface secretions” that contain polysaccharide slimes and mucilages along with lipids and lipoproteins that have evolved to function in pollen adhesion (Clarke et al., 1979; Heslop-Harrison and Heslop-Harrison, 1985). The intercellular spaces between the columnar papillae cells of apple flower stigmas have also been postulated to provide a “stigmatic hydration factor”, functioning to rehydrate pollen grains (Ferrari et al., 1981); this factor would ensure a high humidity microenvironment on the stigma (Wilson et al., 1989). It is possible that the “wet” apple stigma surface does not provide enough free moisture during daylight hours in the absence of dew, such that moisture is limiting for E. amylovora growth during the day, unless it is provided by atmospheric precipitation. However, in the context of E. amylovora growth, an optimal growth rate may only be achieved in the presence of additional free moisture. 37 Dew consists of water droplets formed by the condensation of water vapor on a cold surface. Dew forms when the temperature of the surface of an object falls below the dew point temperature. Plant surfaces, for example, exhibit radiative cooling during the night, and can become colder than the ambient air. Dew will form under conditions with relative humidity > 85% and light winds (0.5 m s-1 to no greater than 4.5 m s-1), and dew forms optimally under clear skies (Beysens et al., 2005; Muselli et al., 2002; Shaw, 1973; Xiao et al., 2013). The formation of dew requires condensation nuclei, usually small particles that water droplets will form around; these nuclei simultaneously increase the size of the water droplet and decrease the probability of evaporation of the droplet (Hussein et al., 2018). Recently, Steiner et al. (2015) showed that subpollen particles (SPP), pollen grains that have ruptured under high humidity conditions, can function as atmospheric cloud condensation nuclei. Since the “wet” surface of the apple stigma is a nonselective adhesive surface for pollen, it is possible that the occurrence of SPPs on stigmas could serve as condensation nuclei for dew formation. Dew formation on apple flower stigmas would serve as a source of free moisture to stimulate E. amylovora growth. It should also be noted that during the process of dew formation, the condensation of water from the atmosphere into the liquid phase generates heat or the latent heat flux (Xiao et al., 2013). The importance of this heat generation to E. amylovora growth is currently unknown, although the perception of heat input into the microenvironment of the stigma may also stimulate E. amylovora growth. Recently, Rougerie-Durocher et al. (2020) used thermocouples to measure apple flower stigma temperatures, and to document the difference between these temperatures and the ambient air temperature. Their comparative measurements of both temperatures during the overnight hours indicated that stigma temperatures were consistently lower than the ambient air, by sometimes as much as 1-2°C (Rougerie-Durocher et al., 2020). These observations are 38 consistent with the knowledge that plant surfaces exhibit radiative cooling, and suggest that stigmas would serve as suitable locations for dew formation. The growth of E. amylovora on apple flower stigmas is critical for pathogenesis, as establishment of large populations of 105 to > 107 per flower are associated with an increased frequency of blossom blight. Our data have shown that population surges of greater than 10-fold increases per day can occur when temperatures are below previously reported thresholds and well below known optimal growth temperatures in vitro. Our demonstration of growth occurring at night also suggests that evening applications of antibiotics during bloom might enhance population reduction since orchards should harbor net smaller populations of E. amylovora on flowers in evenings compared to the following morning. Evening applications of antibiotics, such as kasugamycin and oxytetracycline that are sensitive to photodegradation (Slack et al., 2021), may also increase efficacy against blossom blight. 39 CHAPTER 3 EFFECT OF KASUGAMYCIN, OXYTETRACYCLINE, AND STREPTOMYCIN ON IN- ORCHARD POPULATION DYNAMICS OF ERWINIA AMYLOVORA ON APPLE FLOWER STIGMAS Abstract We assessed the effect of three antibiotics (streptomycin, oxytetracycline, and kasugamycin) on populations of the fire blight pathogen Erwinia amylovora on apple flower stigmas during three field seasons. Timing of application relative to E. amylovora presence on flower stigmas had little impact on both population dynamics and subsequent disease incidence. While E. amylovora populations on water-treated flowers increased to 106-7 cfu flower-1 after 4- 5 days in each experiment, the antibiotics streptomycin and kasugamycin caused statistically- significant reductions of stigma populations by as many as 4-5 logs over a 4-5 day period in two of the three experiments. In contrast, the effect of oxytetracycline on E. amylovora populations on stigmas was more variable, with reductions in E. amylovora populations only observed in one of the three experiments. In agreement with the population data, disease incidence was significantly higher on oxytetracycline-treated flowers compared to the other antibiotic treatments in two of three years. Statistical analyses of effects of weather parameters on antibiotic activity revealed that solar radiation and temperature negatively impacted the activity of both kasugamycin and oxytetracycline. We further assessed the potential for photodegradation of formulated kasugamycin (Kasumin 2L, Arysta LifeSciences), and found that Kasumin 2L was susceptible to degradation in vitro after exposure to a 16-hr photoperiod of daily light integrals (DLIs) varying from 6 to 35 mol·m–2·d–1. We further determined that exposure to three consecutive 16-hr photoperiods of DLIs of 23 or 35 mol·m–2·d–1 reduced the available concentration of Kasumin 40 2L (assessed using a bioassay) from 100 g ml-1 to 10-20 g ml-1. Our results correlate the superior blossom blight control efficacy of kasugamycin and streptomycin to significant population reductions of E. amylovora on apple flower stigmas but indicate that, similar to oxytetracycline, kasugamycin is vulnerable to photodegradation which would suggest that further considerations should be made when applying this antibiotic. Source For a full text of this work go to: Slack, S.M., Walters, K. J., Outwater, C., & Sundin, G. W. (2020). Effect of kasugamycin, oxytetracycline, and streptomycin on in-orchard population dynamics of Erwinia amylovora on apple flower stigmas. Plant Disease. Accepted . https://apsjournals.apsnet.org/doi/10.1094/PDIS-07-20-1469-RE 41 CHAPTER 4 EVALUATION OF A CONTACT STERILANT AS A NICHE-CLEARING METHOD TO ENHANCE THE COLONIZATION OF APPLE FLOWERS AND EFFICACY OF AUREOBASIDIUM PULLULANS IN THE BIOLOGICAL CONTROL OF FIRE BLIGHT Abstract Due to rapid expansion of organic apple production and issues with antibiotic resistance in conventional production, there is a need for novel strategies to protect orchards from outbreaks of fire blight caused by the bacterial pathogen Erwinia amylovora. The biological control material, Blossom Protect (consisting of two strains of the fungus Aureobasidium pullulans), has shown promising results in some apple growing regions but has shown reduced efficacy in humid apple-growing regions such as the Midwestern and Eastern United States. In an attempt to increase both flower colonization by A. pullulans and disease control efficacy, we applied a contact sterilant to clear the flower niche of native microbial populations 4 hr prior to application of Blossom Protect. Results from four years of field testing showed that application of the contact sterilant resulted in a significant reduction in populations of bacteria, fungi, and native yeast populations; however, all of these microbial populations recovered after 24 h. Examination of fire blight incidence revealed that use of the contact sterilant did not significantly impact disease control. Two applications of Blossom Protect, at 70–80% bloom and full bloom, exhibited the best efficacy in disease control over four years. In addition, we observed marginal differences in the marketability of the final fruit between treatments, each with less than 8% deemed unmarketable due to russet. Our results indicate that niche-clearing of apple flowers did not enhance flower colonization by A. pullulans, nor produce an increase in biological control efficacy. However, our current experimental results (2014–2018) do indicate consistency and an increase in efficacy of Blossom Protect compared to results from previous years. 42 Source For a full text of this work go to: Slack, S.M., Outwater, C. A. Grieshop, M. J., & Sundin, G. W. (2019). Evaluation of a contact sterilant as a niche-clearing method to enhance the colonization of apple flowers and efficacy of Aureobasidium pullulans in the biological control of fire blight. Biological Control, 139, 104073. https://doi.org/10.1016/j.biocontrol.2019.104073 43 CHAPTER 5 DECIPHERING THE ROLE OF A TYPE THREE SECRETION SYSTEM EFFECTOR IN THE PATHOGENICITY OF ERWINIA AMYLOVORA Abstract Flowers are ephemeral in nature, may only be open for 1-2 weeks, and if a flower is not pollinated, it may quickly abscise. When apple flowers are infected with Erwinia amylovora, it can take weeks for the necrotic symptom development to occur. However, these infected flowers showing blossom blight symptoms do not abscise throughout the entire growing season. We conducted a field study showing that flowers excluded from pollinators and inoculated with E. amylovora did not abscise, whereas flowers inoculated with nonpathogenic bacteria abscised at the same rate as an untreated control. The E. amylovora HopPtoCEa (HopC) effector, which is part of the Hrp Type III secretion system (T3SS), was implicated as functioning to delay flower abscission, as a mutant (Ea1189∆HopC) without this effector elicited a response of early flower abscission under growth chamber conditions. Further investigation into the peduncle abscission zone indicated that zone separation was occurring at 3 days post inoculation of the effector mutant. This study provides the foundation for more in-depth investigations into the functional role of E. amylovora HopC in host-pathogen interactions. 44 Introduction Abscission occurs in plants as the shedding of unwanted organs, including leaves, flowers, individual floral organs, and fruits (Patharkar & Walker, 2018). Abscission takes place at anatomically distinct cell layers termed abscission zones (AZ) and is mediated by several phytohormones, such as ethylene, jasmonic acid, abscisic acid, auxins, gibberellins, and brassinosteroids (Roberts et al., 2002; Estornell et al., 2013). Separation at the AZ is a natural process typically triggered by developmental or environmental cues. For example, shortening photoperiods and cooler temperatures trigger leaf abscission in the fall and pollination or fertilization, or the lack thereof, modulates floral organ or whole flower abscission in a wide range of plant species (van Doorn & Stead, 1997). Abscission has been heavily studied in apple, particularly at the AZ located at the base of the flower pedicel (Berüter and Droz, 1991; Kolarič, 2000; Sun et al., 2009; Botton et al., 2011). Flowers that are not pollinated abscise within a week after petal fall (Wertheim, 1999). There have been many studies attempting to decipher the role of plant hormones in apple abscission. The overall consensus is that auxins block abscission while ethylene stimulates it, though their relationship is not that straightforward. During apple anthesis, the timing where flowers are most susceptible to E. amylovora infection, the flower ovaries have little growth and low auxin levels (Gruber and Bangerth, 1990). The low auxin levels create an opportunity for flowers to abscise, and ethylene (ethephon) has been used in commercial orchards as a chemical thinner at this stage with ease (Byers, 1993; Meland and Kaiser, 2011). After pollination however, the seeds create their own source of auxin leading to a rapid decrease in natural abscission rates (Wertheim, 1999). 45 Other than a natural response to environmental and internally-controlled hormone regulation, abscission of plant organs also can occur following the infection of plants by pathogens, as an active plant immune response. Bacterial pathogens have been long documented to cause infected host organ abscission as a plant defense response (Ben-David et al., 1986; Patharkar et al., 2017). Though there have many studies demonstrating this abscission phenotype, a more recent study utilizing Pseudomonas syringae pv. tomato DC3000 demonstrated a correlation between bacterial infection and the abscission of Arabidopsis cauline leaves (Patharkar et al., 2017). This pathogen-triggered leaf abscission required a functional type III secretion system (T3SS), a well- studied protein secretion/translocation system in bacteria that triggers effector-triggered immunity (ETI) in plants (Hueck, 1998; Deng et al., 2017; Patharkar et al., 2017). The T3SS enables bacterial pathogens to translocate effector proteins directly into the host cell cytoplasm through a needle-like pilus. These effectors are thought to interfere in the regulation of host cell functions that ultimately benefit the pathogen’s survival and multiplication (Coburn et al., 2007; Büttner & He, 2009). In E. amylovora, the T3SS is a pathogenicity factor; as T3SS-mutant strains are unable to cause disease symptoms (Oh et al., 2005, Oh & Beer, 2005). In E. amylovora, genomic and secretomic studies have revealed that there are at least five effectors (DspA/E, Eop1, Eop3, Eop4 (AvrRpt2Ea), and HopPtoCEa) that are secreted by the T3SS (Bogdanove et al., 1998a; Zhao et al., 2005; Zhao et al., 2006; Nissinen et al., 2007). DspA/E is currently the only of the five effectors that is a pathogenicity factor; it contributes to blocking defense from the host that would otherwise cause callose deposition (Gaudriault et al., 1997; Bogdanove et al., 1998b; DebRoy et al., 2004; Yuan et al., 2020). The other effectors, Eop1, Eop3, AvrRpt2Ea, and HopPtoCEa are considered virulence or avirulence factors (Yuan et al., 2020). Eop1, Eop3, and HopPtoCEa are relatively understudied in the literature, partially 46 because deletion of these effectors does not affect virulence (Zhao et al., 2005; Asselin et al., 2006; Bocsanczy et al., 2012). Of relevance to abscission, there has been recent evidence that the T3SS is active during the first stages of flower infection (Cui et al., 2020). The potential role of flower abscission in apple in response to pathogen infections have not been documented. However, many studies have utilized tomato and pepper as a model for abscission, and recent studies have determined that many of the host genes required in those systems are present in apple (Nakano et al., 2019). Even with gene homology, we observed a lack of abscission in the Erwinia-apple phytosystem, suggesting that E. amylovora is using a T3SS effector to manipulate the apple host response to block infected flowers from abscising. Materials and Methods Field Studies In 2017 and 2018, 600 apple flower clusters at pink stage from the cultivar ‘Gala’ were covered with translucent silk bags to exclude pollinators (Figure 5-1A; Native Seeds, Tucson, AZ). Upon opening, king bloom flowers were treated with either E. amylovora strain Ea110 (Zhao et al., 2005), Pantoea agglomerans strain E325 (Pusey et al., 2011), Pseudomonas fluorescens strain A506 (Stockwell et al., 2010), a phosphorus buffer, or a water control. E. amylovora strain Ea110 has been widely utilized in pathogenicity field trials at Michigan State University (Slack et al., 2017), and both P. agglomerans strain E325 (Pusey et al., 2011) and P. fluorescens strain A506 (Stockwell et al., 2010) have been used as biological controls for fire blight and are nonpathogenic on apple flowers. Prior to inoculation, all bacteria were grown overnight, and populations were adjusted turbidimetrically to 1x106 CFU/ml; a total of 5 µl of the cell suspensions was directly pipetted into the hypanthium of each flower. Each experimental 47 treatment consisted of 4 replicates of 25 flowers each, for a total of 100 flowers per treatment. Every few days, flowers were checked for abscission after inoculation on 27 April 2017 and 1 May 2018, respectively. 48 Figure 5-1. Flowers with pollinator excluder bags attached. A: Image of a recently opened apple flower cluster covered by a pollinator excluder bag. B: Image of an apple flower cluster that was inoculated with Pantoea agglomerans E235. Observation was 20 days after inoculation from a field experiment conducted in 2018. 49 In 2020, two rates of ethephon (375 and 600 ppm; Meland and Kaiser, 2011) were used to determine if treatment with this compound would cause flower abscission on flowers inoculated with E. amylovora strain Ea110. Unlike in the previous field experiments, whole ‘Gala’ trees were sprayed with strain Ea110 at a rate of 1x106 CFU/ml, not inoculated directly, and pollinator excluders were not used as well. Four replicate trees per treatment were used, and this experiment was repeated twice. Immediately prior to ethephon application, all open flowers were tagged, and only flowers that had been open during inoculation and treatment were used to rate the results. Growth chamber experiments Feathered ‘Gala’ trees on EMLA-26 rootstock were planted into 18.9 L plastic pots, and trees were moved into a growth chamber where they broke bud and began to flower. Flower cluster development was relatively uniform among the trees. Treatments were either bacterial suspensions or a water check: E. amylovora strain Ea1189 (Ea1189 WT), effector mutant 1 (Ea1189∆HopC), effector mutant 2 (Ea1189∆DspE), effector mutant 3 (Ea1189∆AvrPpt2), a non-T3SS virulence factor mutant (Ea1189∆V1), P. agglomerans strain E325, and a non- bacterial water control. All E. amylovora mutants were previously characterized (Zhao et al., 2005). All bacterial strains were directly pipette inoculated into the flower hypanthium at a rate of 1x104 CFU/µl with 5 µl per hypanthium; we also applied 5µl sterile distilled water for the control treatment. Flowers were checked daily for abscission and trials were conducted twice in June and July 2018. Similar to the field experiments, ‘Gala’ flowers from 24 trees were inoculated with either WT, Ea1189∆HopC, or Ea1189∆DspE (8 trees each) at rates of 1x107 CFU/µl per hypanthium. Populations of E. amylovora were tracked over a time course in separate flower structures: the 50 flower base, the pedicel, and the flower cluster base. Due to the destructive nature of this sampling and limited availability of flowers, populations were sampled at 0, 12, and 48 hours. Histology of abscission In 2019, ‘Gala’ flowers from 24 trees were inoculated with either WT, Ea1189∆HopC, or a water control (8 trees each) at rates of 1x104 CFU/µl with 5 µl per hypanthium. At days 0, 3, and 5, whole flower clusters were carefully collected to avoid causing premature flower detachment and taken to the Michigan State University Histopathology Services Center (https://humanpathology.natsci.msu.edu/). Samples were carefully cut to size and inserted into small embedding cassettes (40.1 x 28.5 mm; Fisher Scientific, Waltham, MA). Samples were then processed with 3% glutaraldehyde and in 0.02 M phosphate buffer and mounted in paraffin wax. Slides were constructed and dyed with Toluidine Blue O as previously described (Sakai, 1973). Statistical analysis All statistical analyses were completed using the open access program “R” (R Core Team, 2013) utilizing a one way ANOVA to check for statistical differences, and then separating means with Tukey’s HSD test. Results In 2017 and 2018 in the field, flowers treated with any bacterial strain or compound other than E. amylovora abscised at similar rates, albeit significantly slower in 2018 compared to 2017 (Table 5-1). Flowers abscised at the AZ located at the base of the pedicel, and there was no other observed AZ or artificial breakage caused by the bags (Figure 5-1B). By 15 or 20 days after inoculation irrespective of year, all flowers abscised besides flowers inoculated with E. 51 amylovora. The first sign of infection was ooze production, and this typically occurred between 5-10 days after inoculation, with necrosis on floral organs starting to occur at day 10 in both years; flower tissue was completely dead with systemic spread of the E. amylovora pathogen apparent on some flowers by Day 20 for each year (Table 5-1). 52 Table 5-1. Results of the 2017 and 2018 field trial observing abscission on flowers inoculated with either Erwinia amylovora Ea110, Pantoea agglomerans E325. Pseudomonas fluorescens A506, a phosphorus buffer, or a water control with 4 replicates and a total of 100 flowers per treatment. Flowers were inoculated on 27 April 2017 and 1 May 2018 respectively. Statistical differences between treatments were compared by a one way ANOVA with means separation by Tukey HSD. * indicates statistical difference between treatments. Year Days after inoculation Erwinia amylovora strain 110 Percent flower abscission and standard error Phosphorus Buffer Pseudomonas fluorescens strain A506 Pantoea agglomerans strain E325 Water Control 2017 5 days 0.0 ± 0 0.0 ± 0 0.0 ± 0.0 0.03 ± 0.01 0.0 ± 0.0 10 days 0.0 ± 0* 0.17 ± 0.04 0.13 ± 0.05 0.11 ± 0.02 0.11 ± 0.03 15 days 0.0 ± 0* 1.0 ± 0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 2018 5 days 0.0 ± 0 0.0 ± 0 0.0 ± 0.0 0.02 ± 0.02 0.0 ± 0.0.0 10 days 0.0 ± 0* 0.07 ± 0.01 0.07 ± 0.01 0.05 ± 0.1 0.06 ± 0.02 15 days 0.0 ± 0* 0.86 ± 0.04 0.90 ± 0.04 0.82 ± 0.07 0.79 ± 0.1 20 days 0.0 ± 0.0* 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 53 For the growth chamber experiments, early abscission elicited by Ea1189∆HopC was observed on day 3 of the experiment for an average of 29.5% of flowers; the incidence of abscission in all other treatments was below 5% abscission. This trend continued until day 5, when abscission observed for other virulence factor mutant strains as well as the uninoculated and P. agglomerans-inoculated flowers reached levels similar to Ea1189∆HopC (Figure 5-2). Flowers from day 3 show the differences in flower response to the missing effector (Figure 5-3A and 5-3B), and that the host cells are still alive in the wild type at day 3, and dead by day 7 in growth chamber conditions (Figure 5-3C). We observed that 1 out of 6 flower clusters that pathogen cells were still able to systemically spread from the flower base to the flower cluster base before abscission (Figure 5-3D). Populations of E. amylovora (WT, Ea1189∆HopC, and Ea1189∆DspE) were observed in multiple flower organs 24 hr after inoculation into the hypanthium (Figure 5-4). The mutant strains were able to spread internally to the cluster base, though with not as a large population as the WT. A smaller number of cells of the WT was also observed in the inoculation site, indicating that there may be an advantage for the WT strain for systemic spread (Figure 5-4). 54 ) % ( d e s i c s b a s r e w o l f t n e c r e P 120 100 80 60 40 20 0 3 days 4 days 5 days 6 days 7 days WT ∆V1 ∆E1 ∆E2 ∆E3 non-inoculated P. agglomerans E325 Figure 5-2: Rate of abscission elicited by E. amylovora WT and effector and virulence factor mutant strains from growth chamber experiments. Results of two experimental replicates with a total of 6 trees per treatment averaging 120 flowers per treatment total. Error bars represent standard error of the mean percentages. 55 Figure 5-3. Comparison from growth chamber experiments of selected flower clusters at three days and seven days after inoculation wither either E. amylovora strain 1189 (WT) or the effector mutant 1 (Ea1189∆HopC). WT at 3 days (A), Ea1189∆HopC at 3 days (B), WT at 7 days (C), and Ea1189∆HopC at 7 days (D). 56 U F C 0 1 g o L n o i t a l u p o P a r o v o l y m a a i n i w r E 8 7 6 5 4 3 2 1 0 Flower base Petiole Cluster base Flower base Petiole Cluster base Flower base Petiole Cluster base WT ∆E1 ∆E2 Figure 5-4. Erwinia amylovora strain populations 24 hr after inoculation into the hypanthium. Flowers were inoculated with 1ul of 1x107 CFU. Error bars indicate the standard error of the population. 57 Observations at the AZ shows distinct separation on flowers inoculated with Ea1189∆HopC after three days (Figure 5-5C) while the WT and uninoculated control flowers do not show that phenotype at the AZ (Figure 5-5A and 5-5B). By 5 days post flower inoculation, the AZ on the uninoculated control flowers has spilt, where the WT remind attached (Figure 5D and 5E). Ethephon-treated uninoculated flowers abscised at rates comparable to flowers not treated with the plant hormone (Figure 6). When comparing inoculated to uninoculated trees, there was significant differences in the rates of abscission. However inoculated flowers treated with 600 ppm Ethephon was statistically the same as uninoculated flowers treated with either 375 ppm Ethephon or no ethephon treatment. A higher rate of flower abscission in inoculated trees was observed in 2020; however a frost event damaged many flowers in Michigan in this year (Figure 6). 58 Figure 5-5. Histology sides observing the abscission zone in apple. A. WT at 3 days post inoculation, B. Un-inoculated control at 3 days post inoculation, C. Ea1189∆HopC at 3 days post inoculation, D. WT at 5 days post inoculation, and E. Un-inoculated control at 5 days post inoculation. 59 n o i s i c s b A r e w o l F t n e c r e P 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% A AB AB BC CD D un-inoculated un-inoculated un-inoculated E. amylovora E. amylovora E. amylovora flowers (375 ppm flowers (600 ppm flowers inoculated flowers inoculated flowers inoculated flowers Ethephon) Ethephon) (Untreated Control) (375 ppm Ethephon) (600 ppm Ethephon) (Untreated Control) Figure 5-6: Effect of Ethephon on flowers inoculated with Erwinia amylovora. Error bars indicate standard error of mean percentages. Treatments that are significantly different have a different letter, means separated one way ANOVA followed by Tukey HSD. Use Non- inoculated instead of un-inoculated; italicize E. amylovora 60 Discussion Abscission, potentially as a defense response, was observed in apple flowers inoculated with an effector mutant of E. amylovora. This study found that flower abscission was not rapid enough to inhibit the bacteria from internally spreading past the flower AZ. However, it may be that artificial inoculation was partially responsible for this slowed abscission. In orchard settings, it typically takes 2-3 days for E. amylovora populations to develop on flower stigmas, and for cells to migrate to the hypanthium (Bubán and Orosz-Kovács, 2003); we inoculated the bacteria directly into the hypanthium, therefore bypassing this step. Also, it has been noted that E. amylovora is translocating effectors into cells on the stigma (Cui et al., 2020), meaning that the effectors used in this study are functionally utilized by the pathogen before the bacteria begin to spread systemically, but in this instance we created a situation where the bacteria and effectors are moving and translocating concurrently. While may E. amylovora spread too quickly for abscission to inhibit spread completely, the use of an effector to disable abscission in hosts where spread is slower would be advantageous for a disease-causing organism. In this study, we have verified that abscission in apple flowers is delayed by an effector protein secreted by the T3SS system in E. amylovora. The Ea1189∆HopC strain elicited an early abscission response in flowers, and, when complemented (not shown in this chapter), the lack of flower abscission was restored. The phenotype of abscission was not present on leaves (data not shown), suggesting that the target of the effector is flower specific. The lack of abscission caused by application of ethephon suggests that the ethylene response is potentially being blocked by E. amylovora. Thus, in commercial orchards where ethephon is applied for flower thinning, growers may be inadvertently selecting for infected flowers. 61 Aside from the findings demonstrated here, it would be important to determine the host pathways that impacted by the effector as well as locating the effector targets. In addition, the identification of potential genes and the effector target could help with breeding resistant apple lines. If there were breading lines that could recognize effector at the initiation of infection, i.e. during the buildup of populations on flower stigmas, there would be enough time for flowers to abscise prior to infection by E. amylovora. Such abscission at the early stages of flowering could also be advantageous for crop thinning as well, since growers already induce abscission in flowers. 62 CHAPTER 6 POSSIBLE SYNERGISTIC EFFECT OF PROHEXADIONE-CALCIUM AND ACIBENZOLAR-S-METHYL IN MANAGING SHOOT BLIGHT OF APPLE Abstract The secondary cycle of fire blight is evident in pomaceous shoots, with symptoms of necrosis, shepherd’s crooks dieback, and development of dark sunken lesions. With limited management options, low rates of prohexadione calcium and acibenzolar-S-methyl were evaluated for three years on ‘Gala’ apple trees of various ages, including newly planted trees. Following inoculation of trees with Erwinia amylovora, lesions developed more slowly in shoots treated with both prohexadione calcium (Pro-Ca) and acibenzolar-S-methyl (ASM) as a tandem treatment compared to these materials used alone. We found the rates of 28.3 g Pro-Ca + 28.3 g ASM and 56.6 g Pro-Ca + 28.3 g ASM per hectare significantly (P < 0.5) reduced shoot blight lesion incidence without significantly inhibiting growth on trees less than 5 years old. This rate did not impact tree yield, caused minimal winter injury defined as dead wood or injured shoots, and had statistically less incidence of E. amylovora systemic spread than control trees. On apple trees between 5 to 10 years old, these same rates of Pro-Ca + ASM significantly reduced lesion size in 2018-2020, while only reducing shoot length in 2018. Comparable rates of either treatment alone were not as effective in reducing lesion length as compared with the combination of Pro-Ca + ASM. 63 Introduction Fire blight is a polycyclic disease of rosaceous plants include pome fruit, and is caused by the bacterium Erwinia amylovora (Burill). This disease can cause sporadic but devastating losses for both pear and apple production (Vanneste, 2000). Recently, concerns over losses due to fire blight have increased, both as a result of climate change and modern orchard design. Increased severity and frequently of epidemics possibly resulting from climate change and new planting systems have led to increased concerns from apple stakeholders. The interest in high density, ‘spindle tree trellised” plantings has greatly increased the risk of fire blight, as trees are grown close together, and rapid growth is greatly encouraged (Robinson et al., 2004). One of the major features of high-density plantings is the lack of permanent scaffolding branches as most growth is less than 5 years old and directly off the central leader of the tree. Previously, the cultivation of long-lived scaffolding branches was a pivotal strategy for traditional stand-alone trees (Robinson et al., 2004), and such longer-lived scaffolds were more recalcitrant to fire blight infection. The first line of defense against fire blight is cultural control methods including pruning cankers that formed the previous season, and sanitation via removal of pruned wood from the orchard (van der Zwet et al., 2012). The goal of removing cankers is to reduce inoculum in the orchard, as E. amylovora cells emerge from the cankers as ooze droplets in the early spring (Slack et al., 2017). However, many disease management tactics focus on the site primary infection which are flowers(Wise, 2021). Flowers have a short window of susceptibility for E. amylovora colonization and subsequent infection, making them prime for precise chemical management (Pusey, 2000; Dewdney et al., 2007, van der Zwet et al., 2012; Slack et al., 2021). Currently, three antibiotics and numerous biopesticides and copper products egistered in the 64 United States for control at bloom (Wise, 2021). There are disease forecasting programs freely available for grower commercial enterprises to precisely time applications antibiotics (Steiner, 1990; Smith 1996; Dewdney et al 2007; Billing, 2011; Smith and Pusey 2011; Turechek and Biggs 2015). Fire blight is a polycyclic disease and secondary infections occur in the form of shoot blight. New shoot growth is highly susceptible to infection, and ooze emerging from cankers, infected flowers, or other infected shoots is easily spread by wind, rain, and insects through the orchard leading to new infections (Slack et al., 2017; Boucher et al., 2019). Although there have been some efforts have been made to select for resistant cultivars, the most popular commercial cultivars are susceptible to fire blight being principally bred for color, taste and storability as breeding for these traits comes first, which is why there few fire blight resistant cultivars (Kellerhals, et al., 2017). Unlike blossom blight, both curative and preventive options are limited for shoot blight control (Wise, 2021). One of the limited options for control is prohexadione calcium (Pro-Ca), which is plant growth regulator registered for use in apple since the late 1990s (Evans et al., 1999). Pro-Ca is a gibberellin inhibitor that decreases the length of internodes on woody shoots. Though originally marketed as a product to reduce shoot growth for pruning and fruit color management, researchers quickly realized the potential of Pro-Ca for shoot blight management (Breth et al., 1998; Jones et al., 1999; Yoder et al., 1999). These early studies recommended a shoot blight management program consisting of two applications one at 125 mg/liter or one application of 250 mg/liter before shoots reached 10 cm in length, as this compound acts. Pro-Ca is used as a preventative measure as there is little evidence of curative action (McGrath et al 2009; Adwinkle et al., 2002; Yoder et al., 1999). The mode of action of Pro-Ca is thought to be 65 by thickening the cell wall of host cells, preventing infection by E. amylovora, as the width of the cell walls was greater than the length of the type three secretion pilus, which is required to translocate effector proteins that drive disease occurrence (McGrath et al., 2009). Another promising option for shoot blight management is acibenzolar-S-methyl (ASM). In 2000, Brisset et al., found that ASM applied to apple trees at rates of 100 and 200 mg/L active ingredient reduced incidence of fire blight. Recommendations rates of ASM have ranged from biweekly applications of 75 mg/L to once every 7-10 days at 200 mg/L (Brisset et al., 2000; Maxon-Stein et al., 2002; Johnson and Temple 2016; Johnson et al., 2016). ASM elicits a systemic acquired resistance (SAR) response by the plant through the salicylic acid pathway and leads to the expression of pathogenesis-related (PR) genes (Maxon-Stein et al., 2002). Specifically, defense-related enzymes, peroxidases, and β-1,3-glucanases, which last up to 17 days after application were found important for fire blight symptom reduction (Brisset et al., 2000). Salicylic acid is key to the SAR response and can be elicited by both endogenous production and exogenous application of ASM. While Maxson-Stein et al., (2002) found that ASM alone was not enough to control fire blight in the field, they did note that, when integrated weekly with applications of streptomycin at flower bloom, the incidence of fire blight was reduced (Maxon-Stein et al., 2002). Similar findings were later reported by Johnson et al., (2016), where the addition of ASM to bloom-applied antibiotics reduced fire blight incidence by 6-11%. Both Pro-Ca and ASM also have drawbacks that could limit their use in commercial apple production. Pro-Ca reduces shoot growth which may be important in establishment years, and ASM is cost prohibitive at amounts recommended. Recently, Wallis and Cox (2020) found that applications of Pro-Ca at bloom reduced both blossom and shoot blight with a minimal 66 growth reduction on 15 year old apple trees, which were well into maturity. Presently, little is known about the impacts of Pro-Ca on young (1-3 years) high density plantings were trees encouraged to grow rapidly and are most susceptible to fire blight. Since young high-density trees are smaller in size than traditional trees, testing lower rates of Pro-Ca for efficacy is needed, as an increase in rate causes greater reductions in shoot growth (McGrath et al., 2009). Low volume rate Pro-Ca recommendations are needed for growers that may not necessarily want to stop shoot growth, but need to protect trees during a severe infection year. Similarly, updated standard recommended application rates are needed for ASM to maximize cost effective-ness. We hypothesized that we could use lower rates of Pro-Ca with ASM to promote shoot blight defense without negatively impacting rapid shoot growth in young apple trees. On both apple trees less than 5 years old and 6-10 years old, applications of low rates of Pro-Ca and ASM, both alone and combined, were applied on calendar schedules with 7-14 day intervals depending on treatment over a three year period. We found that rates of 28.3 g Pro-Ca + 28.3 g ASM and 56.6 g Pro-Ca + 28.3 g ASM ha-1 significantly reduced shoot blight lesion severity without inhibiting growth. These rates had little impact on yield, caused minimal winter damage, and had less incidence of E. amylovora systemic spread. Materials and Methods Established Gala Planting In 2018, 8-year old apple cv. ‘Gala’ trees on semi-dwarfing rootstocks were assigned treatments in a randomized block design with repeated measures and four single-tree replicates. Treatments were repeated per year on these same trees from 2018-2020. There were a total of 10 treatments (ha-1): 28.3 g Pro-Ca, 56.6 g Pro-Ca, 84.9g Pro-Ca, 113.4 g Pro-Ca, 226.4 g Pro-Ca, 67 28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro- Ca+ 28.3 g ASM, and an untreated control. Treatments were applied using a 15 L backpack sprayer (Solo, Newport News, VA). Treatments were either applied biweekly applications (28.3 g Pro-Ca, 56.6 g Pro-Ca, 84.9g Pro-Ca, 113.4 g Pro-Ca, 226.4 g Pro-Ca, 28.3 g ASM, 56.6 g) or weekly (28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro-Ca+ 28.3 g ASM) for a total of 3 to 6 applications based on timing schedule (Table 1). After the third application, 5 shoots per tree were inoculated with Erwinia amylovora strain Ea110 (Zhao et al., 2005) for a total of 20 inoculated shoots per treatment. Strain Ea110 was cultured in LB broth overnight at 28oC in a shaking incubator, and populations were adjusted turbidometrically (OD600 nm) to 1 x 108 CFU ml-1 using a Tecan Spark plate reader (Männedorf, Switzerland). The prepared inoculum was maintained on ice while taken to the field, and inoculations were done within 1 hr of inoculum preparation. Inoculation of apple shoots was done by cutting leaves located at the tip of rapidly growing shoots with sterilized scissors that had been dipped in the E. amylovora inoculum (Slack et al., 2017). Total shoot length (new growth) and total lesion length on tagged shoots were measured weekly starting within 48 hrs of the first treatment applications. Weekly measurements ceased when either lesions reached the bud scar from the previous year or shoots had formed terminal buds on untreated trees. 68 Table 6-1. Application rates and times of Pro-Ca and ASM for High Density and Young Gala trees. The following treatments were applied at time points 1, 3 and 5: 128.3 g Pro-Ca, 56.6 g Pro-Ca, 84.9g Pro-Ca, 113.4 g Pro-Ca, 226.4 g Pro-Ca, 28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro-Ca+ 28.3 g ASM; and at time points 2, 4 and 6: 28.3 g ASM, 56.6 g ASM, 28.3 g Pro-Ca + 28.3 g ASM, 28.3 g Pro-Ca + 56.6 g ASM, 56.6 Pro-Ca+ 28.3 g ASM. Time points 1 2 3 4 5 6 Application schedule Pro-Ca treatments Pro-Ca + ASM treatments Pro-Ca + ASM treatments Pro-Ca treatments Pro-Ca + ASM treatments Pro-Ca + ASM treatments Pro-Ca treatments Pro-Ca + ASM treatments Pro-Ca + ASM treatments 69 Young high-density planting. In 2017, ‘Gala’ apple trees were planted with 3 ft spacing utilizing a trellis system, with 36 ft spacing between poles. Trees were planted to, ‘B.9’ and ‘EMLA-26’, were utilized in this planting alternating every 6 trees, allowing for sets of 12 trees with 6 trees per rootstock between poles as replicates (Figure 6-1). Flowers were removed from trees in years 2017-2019 on cool days to ensure that no blossom blight developed and confounded shoot blight inoculations. From 2018 and 2020, treatments were assigned to 12 trees, 6 of each rootstock with four replicates using a randomized block design with repeated measures, as treatments were applied to the same trees each year (Figure 6-1). A total of seven rates of varying levels of fertilizer, ASM (Actigard 50 WG, Syngenta: Basel, Switzerland), and Pro-Ca (Apogee, BASF: Ludwigshafen, Germany) were applied weekly or biweekly starting at petal fall. Treatments were as follows (ha-1): untreated control with no fertilizer, 28.3 g ASM, 113.4 g Pro-Ca, 56.6 g Pro-Ca + 28.3 ASM, 28.3 Pro-Ca + 28.3 ASM, 113.4 g Pro-Ca + 28.3 ASM and a untreated control + fertilizer. In 2019 and 2020, three of the six trees per treatment were inoculated with E. amylovora Ea110 as described above with the modification of two branches were inoculated per tree, with six inoculated shoots per rootstock, per treatment replicate resulting in a total of 24 shoots per treatment. The same trees inoculated in 2019 were used in 2020. Hence, only half of the treatment replicates were inoculated . Shoot length was measured on from 2018-2020, and lesion growth was recorded weekly until terminal bud set or shoot blight extended from the tip to the bud scar. 70 Figure 6-1. Lay out of the high density orchard block utilized for the experiment. 71 High Density Horticultural Measurements: Tree productivity was determined by measuring the height from the top of the rootstock junction to the top of the central leader at end of the growing season from 2017-2020. In spring 2019 and summer 2020, both the severity of winter injury and systemic lesion development were qualitatively evaluated. Winter injury was categorized as blind wood and evidence of shoots not hardening off before winter and subsequent tip dieback. Systemic spread was defined as occurring if the lesions for six shoots spread beyond the bud scar and nearby shoots started to display symptoms. In 2020, on the fourth year, apples were harvested and weighed; apples from both inoculated and control trees were harvested to determine if the inoculated trees had reduced yields. Statistical analysis All statistical analyses were performed with the program RStudio for statistical computing (R Core Team, 2013). Differences between treatments were compared using ratios of shoot growth to lesion length. Total rainfall from first to last measurement was retrieved from the Enviroweather monitoring system stationed at the Michigan State University Horticulture farm (HORT) located within 1 km of the test orchards. Results Established Gala Planting In 2018, seven of ten treatments exhibited significantly less growth once the onset of treatments began. However, there were no significant differences in shoot length between treatments in both 2019 and 2020 (Figure 6-2). Again in 2018, seven of the ten treatments had statistically smaller fire blight lesion sizes than the control by the end of the growing season 72 (Figure 6-3). 2019, the only treatment that did not have significantly smaller lesion size was 28.3 g ASM, indicating that even though there was no difference in shoot length size, lesion size was reduced in nine of the ten treatments. In 2020, lesion size was less than that of the untreated control for the following treatments: 56.6 g ASM, 28.3 g Pro-Ca + 28.3 ASM, 28.3 g Pro-Ca + 56.6 g ASM, and 56.6 g Pro-Ca + 28.3 ASM (Figure 6-3). These 4 treatments had consistently lower lesion lengths than the control each year, however in 2018, 56.6 g ASM and 56.6 g Pro-Ca + 28.3 ASM exhibited a growth reduction (Figure 6-2). Overall shoot growth in the orchard was highest in years receiving more precipitation. 2018 was the driest of the test years, followed by 2019 and 2020 recorded 117 mm of precipitation over the 6 weeks of treatment applications (Table 6-2). 73 25 20 15 10 5 0 ) m c ( t n e m e r u s a e m t s a l d n a t s r i f n e e w t e b e c n e r e f f i D a C - o r P g 3 . 8 2 a C - o r P g 6 . 6 5 a C - o r P g 9 . 4 8 M S A g 3 . 8 2 a C - o r P g 4 . 3 1 1 a C - o r P g 4 . 6 2 2 M S A g 6 6 5 . … + a C - o r P g 3 . 8 2 … + a C - o r P g 3 . 8 2 … g 3 . 8 2 + a C - o r P 6 . 6 5 a C - o r P g 3 . 8 2 a C - o r P g 6 . 6 5 a C - o r P g 9 . 4 8 a C - o r P g 4 . 3 1 1 l o r t n o c d e t a e r t n U a C - o r P g 4 6 2 2 . M S A g 3 . 8 2 M S A g 6 . 6 5 … + a C - o r P g 3 . 8 2 … + a C - o r P g 3 . 8 2 … g 3 . 8 2 + a C - o r P 6 . 6 5 M S A g 3 . 8 2 M S A g 6 . 6 5 a C - o r P g 4 . 3 1 1 a C - o r P g 4 . 6 2 2 … + a C - o r P g 3 . 8 2 … + a C - o r P g 3 . 8 2 a C - o r P g 9 4 8 . a C - o r P g 3 . 8 2 a C - o r P g 6 . 6 5 l o r t n o c d e t a e r t n U … g 3 . 8 2 + a C - o r P 6 . 6 5 l o r t n o c d e t a e r t n U 2018 Figure 6-2. Average difference in shoot length of treatments in 8-10 year old Galas from 2018 to 2020. Error bars represent standard error of the mean. 2019 2020 74 120% 100% 80% 60% 40% 20% 0% t m e m e r u s a e m l a n i f t a d a e r p S n o i s e L h t i w h c n a r b f o t n e c r e P 2018 Final Measurement 2019 Final Measurement 2020 Final Measurement Figure 6-3. Average Percent lesion length of treatments in 8-10 year old Galas from 2018, 2019, and 2020. Error bars represent standard error of the mean. 75 Table 6-2. Total precipitation (mm) during application period for each year. Year Precipitation (mm) 2018 2019 2020 57.7 117 73.7 76 Young high-density planting: B.9 Rootstock In 2018, the first year of treatment applications and tree establishment, the only difference in growth was in the control with no fertilizer application compared to the fertilizer control. None of the Pro-Ca and ASM treatments had reductions in tree growth (Table 3). With the exception of the control with no fertilizer treatment, trees in the 113.4 g Pro-Ca + 28.3 ASM and 113.4 g Pro-Ca treatments had reduced growth from 2019 and 2020 compared to other treatments . By the end of the study in 2020, 113.4 g Pro-Ca + 28.3 ASM and 113.4 g Pro-Ca treated trees had only gained an average of 67.3 ± 18.4 and 70.4 ± 21.4 cm in height respectively, compared to 117.1 ± 22.5 by the fertilized control. The other 3 treatments were not significantly different than the fertilized control (Table 6-3). Trees treated with 113.4 g Pro-Ca + 28.3 g ASM and 113.4 g Pro-Ca also had statistically higher than average incidence of winter injury compared to the other treatments in 2019 and 2020 (Table 6-3). In both years, all treatments had lower systemic lesion development than the fertilized control with 28.3 Pro-Ca + 28.3 g ASM having the shortest lesion development in 2019 and 56.6 g Pro-Ca + 28.3 g ASM having the shortest lesion development in 2020 (Table 6-3). In 2018 and 2020, there were no differences between treatments in shoot length. However, in 2019 all treatments excepted for the 56.6 g Pro-Ca + 28.3 g ASM treatment had statistically lower growth than the control (Figure 6-4). Comparing percent of shoots with lesion development, all treatments had significantly reduced lesion elongation than the fertilizer control. In 2020, where the same trees had been inoculated as prior, only the 56.6 g + 28.3 g ASM and 113.4 g Pro-Ca treatments had shorter lesions (Figure 6-5). Fire blight symptom development only hindered apple production on 113.4 g Pro-Ca treated trees; however, 77 regardless of E. amylovora inoculation, trees treated with 113.4 g Pro-Ca and 113.4 g Pro-Ca +28.3 g ASM had fewer apples at the end of year 4 (Figure 6-6). EMLA-26 rootstock The trends in treatment impacts on growth of trees on EMLA-26 were similar to those of apples on B.9, with 113.4 g Pro-Ca + 28.3 g ASM and 113.4 g Pro-Ca treated trees having the least amount of growth. Tree uniformity was unfortunately not as similar, with the treatment of 28.3 g Pro-Ca + 28.3 g ASM having greatest variability in tree heights, making it difficult to resolve significant differences between treatments in the experience (Table 6-3). Winter injury was more pronounced in scions grafted to EMLA-26, with only 28.3 g Pro-Ca + 28.3 g ASM having statistically less injury than the other treatments in 2019 and 2020. Systemic lesion development was also higher than B9 in EMLA-26, with 28.3 Pro-Ca + 28.3 g ASM, 56.6 Pro- Ca + 28.3 g ASM, and 113.4 g Pro-Ca consistently having statistically less symptom development (Table 6-3). In 2018, 28.3 Pro-Ca + 28.3 g ASM and 113.4 g Pro-Ca + 28.3 g ASM treated trees had reduced height compared to all other treatments (Figure 6-4). In 2019, only 28.3 Pro-Ca + 28.3 g ASM treatment had reduced height and in 2020 no treatments had any reductions in height compared to the control (Figure 6-4). All treatments had considerably lower lesion development than the fertilized control in 2019 (Figure 6-5). In 2020 however, there was no differences in lesion ratios between any treatments (Figure 6-5). Unlike B9, trees inoculated with E. amylovora had numerically fewer marketable fruit, but no statistical differences among treatments. However, the effect of treatment compounded with inoculation decreased marketable fruit production in all treatments except 28.3 g ASM with only 133.4 g Pro-Ca having less marketable fruit when not inoculated (Figure 6-6). 78 Table 6-3 Summary of horticultural measurements from 2018-2020 on High Density trees including: average tree height treatment per rootstock; average incidence of winter damage per treatment per root stock; average incidence of systemic spread by E amylovora per treatment per rootstock. 28.3 Pro-Ca + 28.3 g ASM 29.1 ± 4.6 38.8 ± 24.3 101.3 ± 23.9 28.3 g ASM 34.1 ± 7.2 49.5 ± 16.5 117.7 ± 17.1 2018 2019 - 2020 3 year Growth 2019 12.4 ± 2.2 * 6.6 ± 0.6 2020 29.2 ± 7.9 27.2 ± 5.8 2019 37.5 ± 4.1 25.0 ± 1.6 Treatment 56.6 g Pro-Ca + 28.3 g ASM 113.4 g Pro-Ca + 28.3 g ASM 113.4 g Pro-Ca Control Control - + no Fertilizer fertilizer 24.3 ± 6.6 47.5 ± 29.3 95.9 ± 10.7 12.0 ± 1.4 * 20.8 ± 12.5 41.7 ± 2.8 * 31.2 ± 7.2 9.1 ± 3.3 67.3 ± 18.4 11.0 ± 4.0 * 46.7 ± 11.8 * 29.2 ± 1.4 34.1 ± 12.9 20.2 ± 14.6 70.4 ± 21.4 16.8 ± 7.6 * 50.0 ± 15.2 * 58.3 ± 4.8 * 36.4 ± 3.4* 71.2 ± 23.4 117.1 ± 22.5 22.9 ± 6.8 12.5 ± 13.7 85.4 ± 21.0 6.0 ± 1.6 6.0 ± 0.3 33.3 ± 5.8 75.0 ± 3.6 * 29.6 ± 14.2 33.3 ± 3.9 Average Horticultural measurements Year Rs B9 EMLA 26 Tree Height Growth (cm) Incidence of winter damage (Percent of trees) Incidence of systemic spread (Percent of trees) Tree Height Growth (cm) Incidence of winter damage (Percent of trees) Incidence of systemic spread (Percent of trees) 2020 54.2 ± 10.5 55.0 ± 8.2 * 25.0 ± 14.4 * 45.8 ± 7.9 * 47.5 ± 10.4 * 95.8 ± 4.0 * 79.2 ± 7.9 2018 30.4 ± 12.2 2019 - 2020 43.8 ± 11.6 3 year Growth 91.8 ± 33.1 26.8 ± 8.7 30.9 ± 15.5 54.6 ± 47.9 40.8 ± 10.4 55.8 ± 34.2 107.1 ± 11.2 21.3 ±5.8 25.6 ± 22.7 86.5 ± 13.9 2019 12.7 ± 1.7 * 3.2 ± 0.5 11.6 ± 1.3 * 9.3 ± 2.1 * 2020 33.3 ± 9.6 20.8 ± 4.0 2019 50.0 ± 3.2 * 29.2 ± 1.4 41.7 ± 11.8 * 29.2 ± 3.5 62.5 ± 10.5 * 45.8 ± 4.7 * 19.4 ± 5.9 10.0 ± 7.1 88.2 ± 15.3 12.5 ± 1.7 * 60.0 ± 9.5 * 33.3 ± 3.2 28.3 ± 9.4 46.4 ± 21.8 103.1 ± 16.5 11.7 ± 0.8 * 25.0 ± 8.3 66.7 ± 3.9 * 29.7 ± 12.9 16.2 ± 12.07 71.2 ± 27.3 2.5 ± 0.5 20.8 ± 12.5 29.2 ± 4.2 2020 79.2 ± 15.7 45.8 ± 14.2 26.7 ± 11.3 20.8 ± 7.9 50.0 ± 15.2 100.0 ± 0.0 83.3 ± 6.8 79 t n e m t a e r t g n i r u d h t w o r t o o h s e g a r e v A ) m c ( 18 16 14 12 10 8 6 4 2 0 -2 Control - no 28.3g ASM 113.4g Pro-Ca 56.7g Pro-Ca + 28.3 Pro-Ca + fertlizer 28.3 ASM 28.3 ASM 113.4g Pro-Ca + 28.3 ASM Control + Fertilizer B9 2018 B9 2019 B9 2020 EMLA-26 2018 EMLA-26 2019 EMLA-26 2020 Figure 6-4. Average shoot length growth in high density treatments from 2018, 2019, and 2020 per rootstock. 80 n o i s e l h t i w t o o h s f o t n e c r e P ) % ( t n e m p o l e v e d 120% 100% 80% 60% 40% 20% 0% 28.3g ASM 28.3 Pro-Ca + 56.7g Pro-Ca + 28.3 ASM 28.3 ASM 113.4g Pro-Ca + 28.3 ASM 113.4g Pro-Ca Control + Fertilizer Control - no fertlizer 2019 -B9 2020 -B9 2019 - EMLA 26 2020 - EMLA 26 Figure 6-5. Average percent branch lesion in the high density Gala orchard in 2019 and 2020 per rootstocks. Error bars represent standard error of the treatment mean. 81 9 8 7 6 5 4 3 2 1 0 ) g k ( s e l p p a e l b a t e k r a m f o t h g i e w e g a r e v A 28.3g ASM 28.3g Pro-Ca + 56.7g Pro-Ca + 113g Pro-Ca + 113g Pro-Ca 28.3g ASM 28.3 ASM 28.3 ASM Fertilizer control No fertilizer control B-9 inoculated B-9 non-inoculated EMLA-26 inoculated EMLA -26 non-inoculated Figure 6-6. Average weight of marketable apples in the high density Gala orchard from 2020 per rootstock, comparing inoculated and uninoculated trees within treatment. Error bars represent standard error of the treatment mean. 82 Discussion Overall, trees treated with a combination of Pro-Ca and ASM had slower lesion development. Though no treatment was able to completely prevent lesion development, the slowing of pathogen systemic spread as evidence by lesion elongation would enable the implementation of cultural management before the pathogen is able to spread to the point where the tree would be lost. The slower lesion spread in Pro-Ca and ASM treatments was observed in years with high rates of shoot growth, either from high precipitation (2019) or a small crop load from frost (2020). Trees beginning to come into their profitable years (Young Gala trees) benefited from the Pro-Ca and ASM combined treatments with little growth reduction rates. Trees under 5 years old would stand to benefit from yearly repeated applications of low volumes of Pro-Ca and ASM combined, with either 28.3 g Pro-Ca + 28.3 g ASM or 56.6 g Pro-Ca + 28.3 g ASM having little effect on horticultural characteristics, such as tree height and yield. EMLA-26 is more susceptible to fire blight infection than B9, and the Gala scions had fewer marketable apples in year 4 on trees that had been inoculated compared to trees that had never had fire blight (Ferree et al., 2002; Figure 6-6). Whereas trees grafted on to B9, a more resistant root stock, only had fewer marketable apples on trees continuously treated with the highest amounts of Pro-ca (Ferree et al., 2002; Figure 6-6). The first year of harvest is important for growers to make back the initial investment in the trees as well as the trellis system, and any treatment that reduced profitability will not be adopted by growersadapted. Rootstocks are selected by growers not only for disease resistance, but dwarfing (tree size), soil quality, irrigation capabilities, and other topographical considerations. From this experiment, if the trees are exposed to E. amylovora on EMLA26, regardless of treatment, they will have fewer marketable fruit than trees exposed to fire blight on B9. If a grower in a high fire blight pressure 83 area had to choose between the two, the results from this study suggest that a resistant rootstock will help with initial returns in yield. The higher application rates of Pro-ca had high incidence of winter damage, mainly through buds breaking dormancy late in the season after applications had stopped. These tender shoots were subjected to harsh Michigan winter cold, frost, and freezes, not surprisingly dying in the process. Another option would be to apply these low rates at the timing suggestions from Wallis and Cox (2020) to protect early shoots from injury. These low rates may be able to prevent flower infection as well as protectshoots under 5 cm in young trees. Although the goal was to find ways of reducing systemic spread later in the growing season, typically when growers start to notice shoot infection, the use of these products earlier in the season may be advantageous on newly planted or young non-bearing trees. In summary, treatments of Pro-ca combined with ASM reduced shoot blight over three years in trees between 5-10 years old and over 3 years with trees 0-5 years old. When applied alone at the same rates, the same level of control was not consistently observed. The possibility of a synergism with both products potentially thickening cell walls either more consistently or having higher callous deposition without shortening shoots is ideal for young orchards in high fire blight risk areas. 84 REFERENCES 85 REFERENCES Asselin, J., Oh, C., Nissinen, R., Beer, S. and Nissinen, R. (2006) The secretion of EopB from Erwinia amylovora. In: Acta Horticulturae 409-416. Ben-David, A, Bashan Y, Okon Y. (1986). Ethylene production in pepper (Capsicum annuum) leaves infected with Xanthomonas campestris pv. vesicatoria. Physiol Mol Plant Pathol. 29: 305–316. https://doi. org/10.1016/S0048-4059(86)80047-9 Berüter, J., & Droz, P. (1991). Studies on locating the signal for fruit abscission in the apple tree. Scientia Horticulturae, 46(3-4), 201-214. Beysens, D., Muselli, M., Nikolayev, V., Narhe, R., and Milimouk, I. (2005). Measurement and modeling of dew in island, coastal and alpine areas. Atmos. Res. 73:1-22. Billing, E. (1960). An association between capsulation and phage sensitivity in Erwinia amylovora. Nature, Lond. 186, 819. Billing, E. (1974). The effect of temperature on the growth of the fireblight pathogen, Erwinia amylovora. Journal of Applied Bacteriology, 37(4), 643-648. Billing, E. (1978). Developments in fire blight prediction in south-east England. pp. 159-166 In Scott, P.R., and Bainbridge, A. (ed.) Plant Disease Epidemiology. London, Blackwell. Billing, E. (1980). Fireblight in Kent, England in relation to weather (1955-1976). Ann. Appl. Biol. 95:341-364. Billing, E. (2011). Fire blight. Why do views on host invasion by Erwinia amylovora differ? Plant Pathol. 60:178-189. Bocsanczy, A., Schneider, D., DeClerck, G., Cartinhour, S. and Beer, S. (2012). HopX1 in Erwinia amylovora functions as an avirulence protein in apple and is regulated by HrpL. Journal of bacteriology, 194, 553-560. Bogdanove, A. J., Bauer, D. W. and Beer, S. V. (1998a) Erwinia amylovora secretes DspE, a pathogenicity factor and functional AvrE homolog, through the Hrp (type III secretion) pathway. Journal of Bacteriology, 180, 2244-2247. Bogdanove, A. J., Kim, J. F., Wei, Z., Kolchinsky, P., Charkowski, A. O., Conlin, A. K., et al. (1998b) Homology and functional similarity of an hrp-linked pathogenicity locus, dspEF, of Erwinia amylovora and the avirulence locus avrE of Pseudomonas syringae pathovar tomato. Proceedings of the National Academy of Sciences, 95, 1325-1330. 86 Botton, A., Eccher, G., Forcato, C., Ferrarini, A., Begheldo, M., Zermiani, M., ... & Ramina, A. (2011). Signaling pathways mediating the induction of apple fruitlet abscission. Plant physiology, 155(1), 185-208. Boucher, M., Collins, R., Cox, K., & Loeb, G. (2019). Effects of exposure time and biological state on acquisition and accumulation of Erwinia amylovora by Drosophila melanogaster. Applied and Environmental Microbiology, 85(15), e00726-19. Breth, D. L., Momol, M.T., and Aldwinkle, H. S. (1998). Evaluation of BAS for the control of fire blight in apple shoots. Fungicide Nematic Tests. 53:4. Brisset, M. N., Cesbron, S., Thomson, S. V., & Paulin, J. P. (2000). Acibenzolar-S-methyl induces the accumulation of defense-related enzymes in apple and protects from fire blight. European Journal of Plant Pathology, 106(6), 529-536. Bubán, T., & Orosz-Kovács, Z. (2003). The nectary as the primary site of infection by Erwinia amylovora (Burr.) Winslow et al.: a mini review. Plant Systematics and Evolution, 238(1-4), 183-194. Büttner, D. and He, S. Y. (2009). Type III protein secretion in plant pathogenic bacteria. Plant physiology, 150, 1656-1664.\ Byers, R. E. (1993). Controlling growth of bearing apple trees with ethephon. HortScience, 28(11), 1103-1105. Clarke, A., Gleeson, P., Harrison, S., and Knox, R.B. (1979). Pollen-stigma interactions: identification and characterization of surface components with recognition potential. Proc. Natl. Acad. Sci. USA 76:3358-3362. Coburn, B., Sekirov, I. and Finlay, B. B. (2007). Type III secretion systems and disease. Clinical microbiology reviews, 20, 535-549. Cui, Z., Huntley, R. B., Schultes, N. P., Kakar, K. U., & Zeng, Q. (2020). Expression of the type III secretion system genes in epiphytic Erwinia amylovora cells on apple stigmas benefits endophytic infection at the hypanthium. bioRxiv. Cui, Z., Huntley, R. B., Schultes, N., Steven, B., & Zeng, Q. (2020). Inoculation of stigma colonizing microbes to apple stigmas alters microbiome structure and reduces the occurrence of fire blight disease. Phytobiomes Journal, (ja). Dagher, F., Olishevska, S., Philion, V., Zheng, J., & Déziel, E. (2020). Development of a novel biological control agent targeting the phytopathogen Erwinia amylovora. Heliyon, 6(10), e05222. DebRoy, S., Thilmony, R., Kwack, Y.-B., Nomura, K. and He, S. Y. (2004). A family of conserved bacterial effectors inhibits salicylic acid-mediated basal immunity and promotes disease necrosis in plants. Proceedings of the National Academy of Sciences, 101, 9927-9932. 87 Deng, W., Marshall, N. C., Rowland, J. L., McCoy, J. M., Worrall, L. J., Santos, A. S., et al. (2017). Assembly, structure, function and regulation of type III secretion systems. Nature Reviews Microbiology, 15, 323. Dewdney, M. M., Biggs, A. R., & Turechek, W. W. (2007). A statistical comparison of the blossom blight forecasts of MARYBLYT and Cougarblight with receiver operating characteristic curve analysis. Phytopathology, 97(9), 1164-1176. Dewdney, M.M., Seem, R.C., Aldwinckle, H.S., Svircev, A.M., and Kim, W.-S. (2008b). How cool temperatures affect Erwinia amylovora populations on the apple stigma surface. Acta Hortic. 793:33-38. Dewdney, M.M., Seem, R.C., and Aldwinckle, H.S. (2008a). The effect of apple blossom age on populations of Erwinia amylovora on the stigma surface. Acta Hortic. 793:99-104. Doolotkeldieva, T., Bobushova, S., Schuster, C., Konurbaeva, M., & Leclerque, A. (2019). Isolation and genetic characterization of Erwinia amylovora bacteria from Kyrgyzstan. European Journal of Plant Pathology, 155(2), 677-686. Erskine, J. M. (1973). "Characteristics of Erwinia amylovora bacteriophage and its possible role in the epidemiology of fire blight." Canadian Journal of Microbiology 19, no. 7 837-845. Estornell, L. H., Agustí, J., Merelo, P., Talón, M. and Tadeo, F. R. (2013). Elucidating mechanisms underlying organ abscission. Plant Science, 199, 48-60. Evans, J. R., Evans, R. R., Regusci, C. L., & Rademacher, W. (1999). Mode of action, metabolism, and uptake of BAS 125W, prohexadione-calcium. HortScience, 34(7), 1200-1201. Farkas, Á., and Orosz-Kovács, Z. (2009). Spatial and temporal relations of reproductive organs and traits of self-fertility in the flowers of various Pyrus communis L. cultivars. Int. J. Plant Reprod. Biol. 1:141–146. Farkas, Á., Mihalik, E., Dorgai, L., and Bubán, T. (2012). Floral traits affecting fire blight infection and management. Trees 26:47-66. Farkas, Á., Orosz-Kovács, Z., Déri, H., and Chauhan, S.V.S. (2007). Floral nectaries in some apple and pear cultivars with special reference to bacterial fire blight. Curr. Sci. 92:1286-1289. Ferree, D. C., Schmid, J. C., & Bishop, B. L. (2002). Survival of apple rootstocks to natural infections of fire blight. HortTechnology, 12(2), 239-241. Gaudriault, S., Malandrin, L., Paulin, J. P. and Barny, M. A. (1997). DspA, an essential pathogenicity factor of Erwinia amylovora showing homology with AvrE of Pseudomonas syringae, is secreted via the Hrp secretion pathway in a DspB‐dependent way. Molecular Microbiology, 26, 1057-1069. 88 Gayder, S., Parcey, M., Nesbitt, D., Castle, A. J., & Svircev, A. M. (2020). Population Dynamics between Erwinia amylovora, Pantoea agglomerans and Bacteriophages: Exploiting Synergy and Competition to Improve Phage Cocktail Efficacy. Microorganisms, 8(9), 1449. Giddens, S. R., Houliston, G. J., & Mahanty, H. K. (2003). The influence of antibiotic production and pre‐emptive colonization on the population dynamics of Pantoea agglomerans (Erwinia herbicola) Eh1087 and Erwinia amylovora in planta. Environmental microbiology, 5(10), 1016-1021. Gouk, S.C., and Thomson, S.V. (1999). Influence of age of apple flowers on growth of Erwinia amylovora. Acta Hortic. 489:525–528. Gruber J. and Bangerth F. (1990). Diffusible IAA and dominance phenomena in fruit of apple and tomato. Physiol Plant 79: 354–358 Hasler, T., & Mamming, L. (2001). Population dynamics of Erwinia amylovora on different blossom elements of pear and apple. In IX International Workshop on Fire Blight 590 (pp. 181- 184). Hasler, T., and Mamming, L. (2002). Population dynamics of Erwinia amylovora on different blossom elements of pear and apple. Acta Hortic. 590:181–184. Hendry, A. T., J. A. Carpenter, and E. H. Gerrard (1967). Bacteriophage studies of isolates from fire blight sources. Can. J. Microbiol. 13:1357-1364. Heo, S. J. H. Hwang, J. H. Jun & H. J. Lee (2016). Abscission-related genes revealed by RNA- Seq analysis using self-abscising apple (Malus × domestica), The Journal of Horticultural Science and Biotechnology, 91:3, 271-278, DOI: 10.1080/14620316.2016.1155315 Hernandez-Escarcega, G., Rezzonico, F., Blom, J., Goesmann, A., Duffy, B., Stockwell, V. O., ... & Guerrero-Prieto, V. M. (2013, July). Comparative Genomics of Erwinia amylovora Isolates from Mexico. In XIII International Workshop on Fire Blight 1056 (pp. 173-177). Heslop-Harrison, J., and Heslop-Harrison, Y. (1985). Surfaces and secretions in the pollen- stigma interaction: a brief review. J. Cell Sci. Suppl. 2:287-300. Hueck, C. J. (1998). Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiology and molecular biology reviews, 62, 379-433. Johnson, K. B., & Temple, T. N. (2016). Comparison of methods of acibenzolar-S-methyl application for post-infection fire blight suppression in pear and apple. Plant Disease, 100(6), 1125-1131. Johnson, K. B., Sawyer, T. L., & Temple, T. N. (2006). Rates of epiphytic growth of Erwinia amylovora on flowers common in the landscape. Plant disease, 90(10), 1331-1336. 89 Johnson, K. B., Sawyer, T. L., Stockwell, V. O., & Temple, T. N. (2009). Implications of pathogenesis by Erwinia amylovora on rosaceous stigmas to biological control of fire blight. Phytopathology, 99(2), 128-138. Johnson, K. B., Smith, T. J., Temple, T. N., Gutierrez, E., Elkins, R. B., & Castagnoli, S. P. (2016). Integration of acibenzolar-S-methyl with antibiotics for protection of pear and apple from fire blight caused by Erwinia amylovora. Crop Protection, 88, 149-154. Johnson, K.B., Stockwell, V.O., Sawyer, T.L., and Sugar, D. (2000). Assessment of environmental factors influencing growth and spread of Pantoea agglomerans on and among blossoms of pear and apple. Phytopathology 90:1285–1294. Jones, A. L., Fernando, W. G. D. and Ehret, G. R. (1999). Controlling secondary spread of fire blight with prohexadione calcium. Phytopathology, 89, S37. Kellerhals, M., Schütz, S., & Patocchi, A. (2017). Breeding for host resistance to fire blight. Journal of Plant Pathology, 99, 37-43. Kolarič, J. (2010). Abscission of young apple fruits (Malus domestica Borkh.): a review. Agricultura 7: 31-36. Luepschen, N.S., Parker, K.G., and Mills, W.D. (1961). Five year study of fireblight blossom infection and its control in New York. Bull. Cornell Agric. Expt. Stn., Ithaca 963:1-19. Mann, R. A., Smits, T. H., Bühlmann, A., Blom, J., Goesmann, A., Frey, J. E., ... & Rodoni, B. (2013). Comparative genomics of 12 strains of Erwinia amylovora identifies a pan-genome with a large conserved core. PloS one, 8(2), e55644. Maxson-Stein, K., He, S.-Y., Hammerschmidt, R., and Jones, A. L. (2002). Effect of treating apple trees with acibenzolar-S-methyl on fire blight and expression of pathogenesis-related protein genes. Plant Dis. 86:785-790. McGrath, M. J., Koczan, J. M., Kennelly, M. M., and Sundin, G. W. (2009). Evidence that prohexadione-calcium induces structural resistance to fire blight infection. Phytopathology 99:591-596. McManus P.S., Jones A.L., (1995). Genetic fingerprinting of Erwinia amylovora strains isolated from tree-fruit crops and Rubus spp. Phytopathology 85: 1547-1553 McManus, P.S. and Jones, A.L. (1994). Role of wind-driven rain, aerosols, and contaminated budwood in incidence and special pattern of fire blight in an apple nursery. Plant Dis. 87:1059- 106. Meland, M. And C. Kaiser (2011). Ethephon as a Blossom and Fruitlet Thinner Affects Crop Load, Fruit Weight, Fruit Quality, and Return Bloom of ‘Summerred’ Apple (Malus ·domestica) Borkh. HORTSCIENCE 46(3):432–438. 2011. 90 Mihalik, E., Radvánszky, A., Dorgai, L., Bubán, T. (2004). Study of Erwinia amylovora colonization and migration on blossoms of susceptible and tolerant apple cultivars. Int. J. Hortic. Sci. 10:15–19 Mills, W.D. (1955). Fireblight development on apple in western New York. Plant Dis. Rep. 39:206-207. Muselli, M., Beysens, D., Marcillat, J., Milimouk, I., Nilsson, T., and Louche, A. (2002). Dew water collector for potable water in Ajaccio (Corsica Island, France). Atmos. Res. 64:297-312. Nakano, T., Kato, H., Shima, Y., & Ito, Y. (2015). Apple SVP family MADS-box proteins and the tomato pedicel abscission zone regulator JOINTLESS have similar molecular activities. Plant and Cell Physiology, 56(6), 1097-1106. Nissinen, R. M., Ytterberg, A. J., Bogdanove, A. J., Van Wijk, K. J. and Beer, S. V. (2007). Analyses of the secretomes of Erwinia amylovora and selected hrp mutants reveal novel type III secreted proteins and an effect of HrpJ on extracellular harpin levels. Molecular plant pathology, 8, 55-67. Norelli, J.L., Jones, A.L., and Aldwinckle, H.S. (2003). Fire blight management in the twenty- first century: using new technologies that enhance host resistance in apple. Plant Dis. 87:756- 765. Oh, C. S., Kim, J. F. and Beer, S. V. (2005). The Hrp pathogenicity island of Erwinia amylovora and identification of three novel genes required for systemic infection. Molecular Plant Pathology, 6, 125-138. Oh, C.-S. and Beer, S. V. (2005). Molecular genetics of Erwinia amylovora involved in the development of fire blight. FEMS microbiology letters, 253, 185-192. Park, J., Lee, G. M., Kim, D., Park, D. H., & Oh, C. S. (2018). Characterization of the lytic bacteriophage phiEaP-8 effective against both Erwinia amylovora and Erwinia pyrifoliae causing severe diseases in apple and pear. The plant pathology journal, 34(5), 445. Patharkar, O. R. and Walker, J. C. (2018). Advances in abscission signaling. Journal of Experimental Botany, 69, 733-740. Patharkar, O. R., Gassmann, W. and Walker, J. C. (2017). Leaf shedding as an anti-bacterial defense in Arabidopsis cauline leaves. PLoS genetics, 13, HopC007132. Philion, V., and Trapman, M. (2010). Description and preliminary validation of RIMPRO- Erwinia, a new model for fire blight forecast. Acta Hortic. 896:307-317. Puławska, J., & Sobiczewski, P. (2012). Phenotypic and genetic diversity of Erwinia amylovora: the causal agent of fire blight. Trees, 26(1), 3-12. 91 Pusey, P. L., & Curry, E. A. (2004). Temperature and pomaceous flower age related to colonization by Erwinia amylovora and antagonists. Phytopathology, 94(8), 901-911. Pusey, P. L., Stockwell, V. O., & Mazzola, M. (2009). Epiphytic bacteria and yeasts on apple blossoms and their potential as antagonists of Erwinia amylovora. Phytopathology, 99(5), 571- 581. P usey, P. L., Stockwell, V. O., Reardon, C. L., Smits, T. H. M., & Duffy, B. (2011). Antibiosis activity of Pantoea agglomerans biocontrol strain E325 against Erwinia amylovora on apple flower stigmas. Phytopathology, 101(10), 1234-1241. Pusey, P.L. (1997). Crab apple blossoms as a model for research on biological control of fire blight. Phytopathology 87: 1096–1102. Pusey, P.L. (2000). The role of water in epiphytic colonization and infection of pomaceous flowers by Erwinia amylovora. Phytopathology 90:1352-1357. Pusey, P.L., and Curry, E.A. (2004). Temperature and pomaceous flower age related to colonization by Erwinia amylovora and antagonists. Phytopathology 94:901-911. Pusey, P.L., and Smith T.J. (2008). Relation of apple flower age to infection of hypanthium by Erwinia amylovora. Plant Dis. 92:137–142. Pusey, P.L., Rudell, D.R., Curry, E.A., and Mattheis, J.P. (2008). Characterization of stigma exudates in aqueous extracts from apple and pear flowers. HortScience 43:1471-1478. R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org Rezzonico F., Smits T.H.M,. Duffy B. (2011). Diversity, evolution and functionality of clustered regularly interspaced short palindromic repeat (CRISPR) regions in the fire blight pathogen Erwinia amylovora. Appl Environ Microbiol 77:3819–3829 Riggle, J. H., & Klos, E. J. (1972). Relationship of Erwinia herbicola to Erwinia amylovora. Canadian Journal of Botany, 50(5), 1077-1083. Riggle, J. H., and E. J. Klos (1970). "Inhibition of Erwinia amylovora by Erwinia herbicola." Phytopathology 60 Ritchie, D. F., & Klos, E. J. (1977). Isolation of Erwinia amylovora bacteriophage from aerial parts of apple trees. Phytopathology, 67(1), 101-4. Roach, D. R., Sjaarda, D. R., Castle, A. J., & Svircev, A. M. (2013). Host exopolysaccharide quantity and composition impact Erwinia amylovora bacteriophage pathogenesis. Applied and environmental microbiology, 79(10), 3249-3256. 92 Roberts, J. A., Elliott, K. A. and Gonzalez-Carranza, Z. H. (2002). Abscission, dehiscence, and other cell separation processes. Annual review of plant biology, 53, 131-158. Robinson, T. L., DeMarree, A. M., & Hoying, S. A. (2004). An economic comparison of five high density apple planting systems. In VIII International Symposium on Canopy, Rootstocks and Environmental Physiology in Orchard Systems 732 (pp. 481-489). Rosen, H.R. (1933). Further studies on the overwintering and dissemination of the fire-blight pathogen. Ark. Agric. Exp. Stn. Bull. 283:1–102. Rougerie-Durocher, S., Philion, V., and Szalatnay, D. (2020). Measuring and modelling of apple flower stigma temperature as a step towards improved fire blight prediction. Agric. For. Metereol. 295:108171. Rundle, J. R., & Beer, S. V. (1986). Population dynamics of Erwinia amylovora and a biological control agent, Erwinia herbicola, on apple blossom parts. In IV International Workshop on Fire Blight 217 (pp. 221-222). Sakai, W. S. (1973). Simple method for differential staining of paraffin embedded plant material using toluidine blue O. Stain technology, 48(5), 247-249. Santander, R.D., and E.G. Biosca. (2017). Erwinia amylovora psychrotrophic adaptations: evidence of pathogenic potential and survival at temperate and low environmental temperatures. PeerJ 5:e3931. Santander, R.D., Meredith, C.L., and Aćimović, S.G. (2019). Development of a viability digital PCR protocol for the selective detection and quantification of live Erwinia amylovora cells in cankers. Sci. Rep. 9:1-17. Schnabel, E. L., Fernando, W. G. D., Meyer, M. P., Jones, A. L., & Jackson, L. E. (1998). Bacteriophage of Erwinia amylovora and their potential for biocontrol. In VIII International Workshop on Fire Blight 489 (pp. 649-654). Schouten, H.J. (1987). A revision of Billing’s potential doublings table for fire blight prediction. Neth. J. Plant Pathol. 93:55-60. Shade, A., McManus, P. S., & Handelsman, J. (2013). Unexpected diversity during community succession in the apple flower microbiome. MBio, 4(2). Sharifazizi, M., Harighi, B., & Sadeghi, A. (2017). Evaluation of biological control of Erwinia amylovora, causal agent of fire blight disease of pear by antagonistic bacteria. Biological Control, 104, 28-34. Shaw, R.H. (1973). Dew duration in central Iowa. Iowa State J. Res. 47:291-227. 93 Singh, J., and A. Khan. (2019). Distinct patterns of natural selection determine sub-population structure in the fire blight pathogen, Erwinia amylovora. Scientific reports 9: 1-13. Slack, S. M., Zeng, Q., Outwater, C. A., & Sundin, G. W. (2017). Microbiological examination of Erwinia amylovora exopolysaccharide ooze. Phytopathology, 107(4), 403-411. Slack, S., Walters, K. J., Outwater, C., & Sundin, G. W. (2020). Effect of kasugamycin, oxytetracycline, and streptomycin on in-orchard population dynamics of Erwinia amylovora on apple flower stigmas. Plant Disease, (ja). Smith T.J. (1999). Report on the development and use of Cougarblight 98C – a situation‐specific fire blight risk assessment model for apple and pear. Acta Hortic. 489:429-436. Smith, T. J., & Pusey, P. L. (2010). Cougarblight 2010, a significant update of the Cougarblight fire blight infection risk model. In XII International Workshop on Fire Blight 896 (pp. 331-336). Smith, T.J. (1996). A risk assessment model for fire blight of apple and pear. Acta Hortic. 411:97-104. Smits, T., Duffy, B., Sundin, G., Zhao, Y., & Rezzonico, F. (2017). Erwinia amylovora in the genomics era: from genomes to pathogen virulence, regulation, and disease control strategies. Journal of Plant Pathology, 99(Special issue), 7-23. Soltesz, M., Nyeki, J., and Szabo, Z. (1996). Receptivity of sexual organs. Pages 132-152 in: Floral Biology of Temperate Zone Fruit Trees and Small Fruits. Song, J. Y., Yun, Y. H., Kim, G. D., Kim, S. H., Lee, S. J., & Kim, J. F. (2020). Genome analysis of Erwinia amylovora strains responsible for a fire blight outbreak in Korea. Plant Disease, (ja). Steiner, A.L., Brooks, S.D., Deng, C., Thornton, D.C.O., Pendleton, M.W., and Bryant, V. (2015). Pollen as atmospheric cloud condensation nuclei. Geophys. Res. Lett. 42:3596-3602. Steiner, P.W. (1990). Predicting apple blossom infections by Erwinia amylovora using the MARYBLYT model. Acta Hortic. 273:139-148. Steven, B., Huntley, R. B., & Zeng, Q. (2018). The influence of flower anatomy and apple cultivar on the apple flower phytobiome. Phytobiomes, 2(3), 171-179. stigmas. Phytopathology, 101(10), 1234-1241. Stockwell, V. O., Johnson, K. B., & Loper, J. E. (1996). Compatibility of bacterial antagonists of Erwinia amylovora with antibiotics used to control fire blight. Phytopathology, 86(8), 834-840. Stockwell, V. O., Johnson, K. B., Sugar, D., & Loper, J. E. (2010). Control of fire blight by Pseudomonas fluorescens A506 and Pantoea vagans C9-1 applied as single strains and mixed inocula. Phytopathology, 100(12), 1330-1339. 94 Sun, L., Bukovac, M. J., Forsline, P. L., & van Nocker, S. (2009). Natural variation in fruit abscission-related traits in apple (Malus). Euphytica, 165(1), 55. Tancos, K. A., & Cox, K. D. (2016). Exploring diversity and origins of streptomycin-resistant Erwinia amylovora isolates in New York through CRISPR spacer arrays. Plant Disease, 100(7), 1307-1313. Taylor, R.K., Hale, C.N., Henshall, W.R., Armstrong, J.L., and Marshall, J.W. (2003). Effect of inoculum dose on infection of apple (Malus domestica) flowers by Erwinia amylovora. N. Z. J. Crop Hortic. Sci. 31:325–333. Thomas, H.E., and Ark, P.A. (1934). Nectar and rain in relation to fire blight. Phytopathology 24:682–685. Thomson, S. V., & Gouk, S. C. (2003). Influence of age of apple flowers on growth of Erwinia amylovora and biological control agents. Plant Disease, 87(5), 502-509. Thomson, S.V. (1986). The role of the stigma in fire blight infections. Phytopathology 76:476– 482. Thomson, S.V., Schroth, M.N., Moller, W.J., and Reil, W.O. (1982). A forecasting model for fire blight of pear. Plant Dis. 66:576-579. Thomson, S.V., Wagner, A.C., and Gouk, S.C. (1999). Rapid epiphytic colonization of apple flowers and the role of insects and rain. Acta Hortic. 489:459–464. Turechek, W. W., & Biggs, A. R. (2015). Maryblyt v. 7.1 for Windows: An improved fire blight forecasting program for apples and pears. Plant Health Progress, 16(1), 16-22. Van de Zwet, T.B and Beer, S.V. (1995). Fire blight- its nature, prevention, and control. Agriculture Information Bulletin No. 613. Van der Zwet T, Orolaza-Halbrendt N, and Zeller W. (2012). Spread and current distribution of fire blight, p 15–36. Fire blight: history, biology, and management. American Phytopathological Society, St. Paul, MN. van Doorn, W. G. and Stead, A. D. (1997). Abscission of flowers and floral parts. Journal of Experimental Botany, 48, 821-837. Vanneste, J. L. (Ed.). (2000). Fire blight: the disease and its causative agent, Erwinia amylovora. CABI. Wallis, A. E., & Cox, K. D. (2020). Management of Fire Blight Using Pre-bloom Application of Prohexadione-Calcium. Plant Disease, 104(4), 1048-1054. 95 Wertheim, S. (2000). Developments in the chemical thinning of apple and pear. Plant Growth Regulation 31, 85–100 https://doi.org/10.1023/A:1006383504133 Wilson, M., & Lindow, S. E. (1993). Interactions between the biological control agent Pseudomonas fluorescens A506 and Erwinia amylovora in pear blossoms. Phytopathology, 83(1), 117-123. Wilson, M., Epton, H. A. S., & Sigee, D. C. (1992). Interactions between Erwinia herbicola and E. amylovora on the stigma of hawthorn blossoms. Phytopathology, 82(9), 914-918. Wise, J. 2021 Michigan 2021 Fruit Management Guide. Michigan State University, BULLETIN- E0154 Wolfe, D. W., A.T. DeGaetano, G. M. Peck, M. Carey, L. H. Ziska, J. Lea-Cox, A. R. Kemanian, M. P. Hoffmann, and D. Y. Hollinger. 2018. Unique challenges and opportunities for northeastern US crop production in a changing climate. Climatic Change 146: 231-245. Xiao, H., Meissner, R., Seeger, J., Rupp, H., Borg, H., and Zhang, Y. (2013). Analysis of the effect of meteorological factors on dewfall. Sci. Total Environ. 452-453:384-393. Yoder, K. S., Miller, S. S., & Byers, R. E. (1999). Suppression of fireblight in apple shoots by prohexadione-calcium following experimental and natural inoculation. HortScience, 34(7), 1202- 1204. Yuan, X., Hulin, M. T., & Sundin, G. W. (2020). Effectors, chaperones, and harpins of the Type III secretion system in the fire blight pathogen Erwinia amylovora: a review. Journal of Plant Pathology, 1-15. Zeng, Q., Cui, Z., Wang, J., Childs, K. L., Sundin, G. W., Cooley, D. R., ... & Yuan, X. (2018). Comparative genomics of Spiraeoideae‐infecting Erwinia amylovora strains provides novel insight to genetic diversity and identifies the genetic basis of a low‐virulence strain. Molecular plant pathology, 19(7), 1652-1666. Zhao, Y., Blumer, S. E. and Sundin, G. W. (2005). Identification of Erwinia amylovora genes induced during infection of immature pear tissue. Journal of Bacteriology, 187, 8088-8103. Zhao, Y., He, S.-Y. and Sundin, G. W. (2006). The Erwinia amylovora avrRpt2EA gene contributes to virulence on pear and AvrRpt2EA is recognized by Arabidopsis RPS2 when expressed in Pseudomonas syringae. Molecular Plant-Microbe interactions, 19, 644-654. Zibordi, M., Domingos, S., & Corelli Grappadelli, L. (2009). Thinning apples via shading: an appraisal under field conditions. The Journal of Horticultural Science and Biotechnology, 84(6), 138-144. 96