TY - JOUR
AU - Loeb, Greg
AB - Abstract The Vector Manipulation Hypothesis (VMH) posits that phytopathogens develop strategies to enhance dissemination by mediating behavior change in insect vectors. The VMH is poorly studied in phytopathogenic bacteria, especially in systems with numerous, occasional vectors. Erwinia amylovora is a bacterial pathogen of pome fruit that produces a bacterial ooze and is mechanically vectored by insects after they feed on ooze. The blossom blight phase of the disease exhibits manipulation of honeybees, leading to enhanced transmission, but whether the same occurs during the shoot blight phase of the disease is unknown. The goal of this study was to evaluate the effect of E. amylovora on the behavior of Delia platura, a fly with a worldwide endemic presence that may transmit E. amylovora. We show that D. platura prefer infected, oozing fruit to uninfected fruit in choice tests and that preference subsides when bacterial ooze is removed from the infected fruit. Flies did not exhibit a preference between infected saplings and uninfected saplings. The volatiles of infected fruit did not attract D. platura, indicating that diseased fruit odor is not responsible for the observed preference for infected fruit. Flies did not differentiate between sapling odors until infected trees had died, at which point they preferred uninfected tree odors. This study supports previous hypotheses suggesting that E. amylovora takes advantage of existing plant–insect interactions, though it is not fully understood how significantly behavioral changes affect transmission. Additional pathosystems with occasional, nonspecific vectors should be studied to further understanding of the VMH. Diptera, hemiptera, hymenoptera, Erwinia amylovora, fire blight The Vector Manipulation Hypothesis (VMH) posits that phytopathogens develop strategies that enhance dissemination through direct and indirect effects on insect vectors (Ingwell et al. 2012). One such effect acting on insect vectors is behavioral manipulation, which occurs directly between the pathogen and the insect or indirectly when mediated by the host plant (Nadarasah and Stavrinides 2011, Eigenbrode et al. 2018). For instance, Erwinia tracheiphila (Smith) (Enterobacterales: Erwiniaceae), a bacterial pathogen of cucurbits vectored by the beetle, Acalymma vittatum (Luperini) (Coleoptera: Chrysomelidae) (Yao et al. 1996), indirectly alters the behavior of vectors by reducing volatile organic compound (VOC) production in flowers of diseased plants and enhancing VOC production in wilting leaves (Shapiro et al. 2012). The beetles, which consume leaf tissue but mate in flowers, are subsequently more likely to feed on diseased leaves and to mate in uninfected flowers, where the pathogen is transmitted through frass (Shapiro et al. 2012). This example of the VMH is one of only a few plant–bacteria pathosystems in which plant-mediated behavioral manipulation is demonstrated, and evaluation of additional pathosystems is necessary to expand our understanding of the VMH (Eigenbrode et al. 2018). Erwinia amylovora (Burrill) (Enterobacterales: Erwiniaceae) is a devastating bacterial pathogen of pomaceous crops that causes fire blight in apple and pear and was recently shown to have plant mediated behavioral effects on honeybees (Apis mellifera (Linnaeus) (Hymenoptera: Apidae)) (van der Zwet and Keil 1979, Cellini et al. 2019). Erwinia amylovora can infect any tree tissue including flowers, fruit, and succulent shoots through natural openings or damage associated with severe storms, management practices, and insect feeding (Norelli et al. 2003). This bacterium can be transmitted by rain and wind as well as insects and is relatively unique in that it exhibits a low degree of vector specificity, meaning that many insects can act as occasional vectors of the pathogen (Purcell 1982, van der Zwet et al. 2012). The role of different insects in the spread of E. amylovora varies across the season, as the disease has distinct phases in the spring (blossom blight) and summer (shoot blight) before overwintering at the margins of cankers formed on lignified tree tissue (van der Zwet et al. 2012). In the spring, a characteristic bacterial ooze exudes from the margins of overwintering cankers and acts as primary inoculum for new infections (Norelli et al. 2003). Ooze consists of E. amylovora encased in a bacterially produced exopolysaccharide and exudes from canker margins after bacterial population growth puts pressure on the plant epidermis and ruptures it (Koczan et al. 2009, Slack et al. 2017). This bacterial ooze is washed into flowers by wind and rain or by insects that visit the ooze and subsequently visit flowers (Norelli et al. 2003). Once deposited on flowers, E. amylovora populations can grow rapidly, exceeding 1 × 107 CFU/flower, and pollinators such as honeybees facilitate disease dissemination by transmitting bacteria from flower to flower (Johnson and Stockwell 1998, van der Zwet et al. 2012). Infected flowers release an alternative suite of VOCs relative to uninfected flowers and this alternative odor profile is somewhat deterrent to honeybees (Cellini et al. 2019). Incomplete deterrence means that honeybees will occasionally visit infected flowers, but subsequently avoid them and prefer to forage on uninfected flowers, allowing for occasional acquisition and regular transmission thereafter (Cellini et al. 2019). In this example of the VMH, it is hypothesized that the bacterium takes advantage of existing interactions between honeybees and apple flowers to enhance dissemination (Cellini et al. 2019). As bloom ends, E. amylovora migrates into young, succulent shoots, which develop bacterial ooze that serves as inoculum for secondary spread of shoot blight (van der Zwet et al. 2012). Secondary shoot blight infections occur when E. amylovora colonizes a wound or natural opening and can be spread from ooze droplets by rain, wind, and insects (Bogs et al. 1998). The role of insects in the spread of shoot blight is poorly understood and over 50 species across several insect orders have been implicated throughout history (van der Zwet et al. 2012). Of these insect orders, anecdotal observations regularly implicated Diptera in the movement of E. amylovora from ooze droplets (Stewart and Leonard 1916, Gossard and Walton 1922, Parker 1936, Emmett and Baker 1971), and flies have received renewed consideration in recent years (Ordax et al. 2015, Boucher et al. 2019). In this scenario, flies feed on ooze and acquire the bacteria, then deliver it to plant surfaces via shedding or defecation, after which E. amylovora theoretically persists until the opportunity to invade a wound or natural opening occurs (Stewart and Leonard 1916, Bogs et al. 1998). Indeed, the high bacterial populations in ooze are hypothesized to support this transmission mechanism, which may have been a significant dissemination method during the evolution of the pathogen (Slack et al. 2017). The nonspecific nature of this mechanism lends it a cosmopolitan suite of potential, occasional vectors (van der Zwet et al. 2012). Whether plants infected with E. amylovora mediate behavioral effects on occasional vectors during shoot blight and whether these effects can be understood through the VMH lens is currently unknown. This topic was broached as early as 1934, when researchers showed that flies strongly preferred oozing bark chips to molasses and hypothesized that the fermenting odors of the oozing cankers were attractive to flies (Thomas and Ark 1934). Indeed, E. amylovora infection alters the VOCs released from infected plants and includes emission of compounds that are known to affect insect behavior, though biological effects of these compounds have not been established in the this system (Cellini et al. 2018). The goal of this study was to evaluate whether E. amylovora infection in apple shoots and fruit alters the behavior of a potential occasional vector, Delia platura. Adult D. platura feed on water droplets, honeydew, and flower nectar and seek out wooded habitats such as apple orchards when temperatures exceed 29°C (Miller and McClanahan 1960, Higley and Pedigo 1984). Recently, this anthomyiid was anecdotally observed to regularly feed on E. amylovora ooze (Boucher et al. 2019) and was shown to transmit the pathogen to uninfected apple saplings from this source (Boucher et al. 2020). Broadly, we sought to understand whether vector manipulation occurs at the shoot blight stage of the E. amylovora disease cycle and whether these interactions resemble those that occur between honeybees and E. amylovora infested flowers during blossom blight (Cellini et al. 2019). We evaluated the effects of both plant odor and physical disease characteristics on the behavior of D. platura, and our specific aims were to 1) evaluate D. platura preference for infected or uninfected fruit; 2) determine whether bacterial ooze induces preference for diseased fruit; 3) establish whether D. platura exhibit preference for diseased apple saplings at various time points post inoculation; and 4) observe whether D. platura exhibit preference for the odor produced by diseased fruit and saplings. Materials and Methods Delia Platura Colony Maintenance This colony was started in summer 2018 with wild D. platura captured in low tunnel strawberry plantings at a single Cornell AgriTech research farm in Geneva, NY. Flies were collected on days when D. platura were likely to be active (between 15.6 and 29.4°C) between May and September (Miller and McClanahan 1960). Adult flies were confined to 24 × 24 × 24” cubic BugDorms (catalog no. 1452, BioQuip Products, Inc., Rancho Dominguez, CA) in a walk-in environmental growth chamber with a 16:8 D:N cycle at 24°C and 50% RH and maintained on diet as previously described (Webb and Eckenrode 1978). The diet included two separate sources, a petri dish filled with brewer’s yeast powder for protein, and a dry mixture of brewer’s yeast (1 part by weight), soy peptone (1 part by weight), skim milk powder (10 parts by weight), and table sugar (10 parts by weight; Webb and Eckenrode 1978). A water source was provided by threading a parafilm sealed 125-ml Erlenmeyer flask with dental wick. Water and diet sources were replaced weekly or as needed. Colony cages were switched out once a month and washed thoroughly to avoid disease outbreaks and poor growing conditions. Dead flies were removed from cages daily. Adults were provided oviposition boxes in a 1.9-liter plastic container filled with 1.2 l of greenhouse sand modified from Webb and Eckenrode (1978). In total, 50 lima beans were pushed into the sand and a 4-ml layer of meat and bone meal fertilizer (Keystone Mills, Romulus, NY) was dispersed across the top of the sand. This container was threaded with dental wick through a hole in the center of its bottom and nested into an identically sized container containing roughly 300-ml tap water to moisten the sand and to encourage the germination of the beans, which stimulates oviposition in D. platura (Weston and Miller 1989). Oviposition boxes were replaced twice weekly, and old boxes were moved to a development cage, where 5–10 more beans were added along with another layer of meat and bone meal. Most larvae pupated after 14 d (Webb and Eckenrode 1978). Two weeks after oviposition boxes had been removed from colony cages, pupae were separated from sand and plant debris by running warm water (~49°C) over the surface of the bin and pouring through a two tiered soil sieve, with a 12.5-mm mesh sieve to catch plant debris placed above a 0.250-mm, 60-mesh sieve to catch pupae. This was done continuously until boxes were empty. Pupae were floated in warm water (~49°C) and skimmed from the surface, and butterfly forceps were used to remove extra debris stuck to the pupae. Some pupae sank to the bottom, and those pupae were collected using butterfly forceps. Pupae were washed an additional time in warm water (~49°C) and set to dry on a paper towel, after which the dry weight was measured for tracking colony production. Pupae were then spread across the surface of moist sand in a 47-ml deli cup to avoid desiccation and placed in a separate 24 × 24 × 24” cubic BugDorm with food and water sources as described earlier. Newly emerged flies were transitioned into colony cages daily or weekly depending on which age cohorts were being used for experiments. Fruit Preference Delia platura fruit preferences were tested in a two-choice bioassay using an arena design based on Drosophila suzukii (Matsumura) (Diptera: Drosophilidae) fruit preference assays (Cha et al. 2019, 2020). The arena was constructed from 2 l cylindrical, lidded plastic buckets roughly 15 cm in diameter with opaque walls (Supp Fig. 1 [online only]). Two 3 × 5 cm rectangular windows were cut into the sides directly across from each other and covered with 1-mm mesh fabric (Outdoor Wilderness Fabrics, Caldwell, ID) to allow for airflow in the bucket. An 8 mm in diameter section was cut out of the center of the lid, and the resulting hole was covered with 1-mm mesh fabric. During experiments, a folded half piece of paper towel was placed over the lid to avoid stimulating flies with light or objects outside of the arena. Two 5-mm-diameter holes were made using a hole punch about 1.5 cm from the top lip of the bucket. Holes were placed directly across from each other and a 5-mm wooden dowel was secured through the holes, running across the middle of the bucket. This dowel was used to suspend two fruit ‘swings’ that each contained a single fruit for the flies to choose between. Fruit ‘swings’ were constructed by bending two standard size paper clips into an ‘S’ shape, and puncturing them through the lid of a 44 ml plastic shot glass opposite each other so that the bottom loop of the ‘S’ of each paper clip held the lid. Paper clips were secured to the lids with glue, and the swing was hung on the wooden dowel from the top loop of the paper clip. Fruit swings were kept 5 cm apart for all replicates. Fruit were suspended within the arena near each other because infected and uninfected fruit can be found within fruit clusters in orchards. With this design, flies had the opportunity to walk or fly to the fruit after being released into the arena. Flies were released into arenas through a 5-mm hole drilled into the base of the bucket at its center. The cap of a 15-ml conical tube with a 5-mm hole drilled through its center was glued to the external surface of the bucket so the hole drilled through the middle aligned with the hole at the center of the bucket base. The cap’s threading was facing outward so a 15-ml conical tube containing flies could be screwed onto the cap. Flies would then crawl or fly out of the conical tube into the arena, ensuring introduction into the arena was not biased towards one side and so flies were not significantly jostled. All experiments were conducted in an environmental chamber with conditions matching those used in colony maintenance. Immature apples (cv. Gala) used in this assay were prepared as previously described (Boucher et al. 2019). Fruit were collected from fire blight free blocks at a single Cornell AgriTech research farm in Geneva, NY in the spring of 2019 when they reached roughly 2–3 cm in diameter and held in cold storage (4°C) until inoculation with E. amylovora. Single colonies of Erwinia amylovora strain Ea273 (originally isolated from Malus × domestica cv. Rhode Island Greening in New York) maintained on Crosse–Goodman (CG) medium were grown overnight in Luria-Burtani (LB) broth (catalog no. 12795-027; Invitrogen) at 28°C in a shake incubator (Crosse and Goodman 1973, Norelli et al. 1984). Fruit were surface sterilized by wiping fruit surface thoroughly with 95% ethanol, and then soaking in a 10% bleach solution for 10 min. After soaking, fruit were removed, rinsed in sterile water, and allowed to air dry under a laminar flow hood. Once dry, fruit were cut in half, and any that appeared to contain rot at the center were discarded. The exposed flesh of the fruit was scored 6 to 8 times using a sterile dental pick, and 100 µl of E. amylovora grown in LB broth was pipetted onto the flesh surface. The sterile dental pick was then used to collect E. amylovora colonies maintained on CG media, which were rubbed across the fruit flesh surface, dispersing the LB broth in the process. The fruit was allowed to absorb the liquid flesh side up in the laminar flow hood for 5 min, before being placed flesh side down on a piece of filter paper that had been soaked in sterile, distilled water in a 180-ml deli cup. The cup was capped with a plastic lid, sealed with parafilm, and incubated at 28°C in a growth chamber with high humidity (>90%) for 7–14 d, until ooze droplets were abundant on the skin of the fruit. Mock inoculations were done according to the same process, but sterile LB broth was used and extra colonies from CG media were not added. For fruit used in ooze removed experiments, ooze was washed from the fruit by rinsing with sterile, distilled water, wiping the surface dry, and then repeating four times to ensure all ooze was removed from the surface rather than dispersed across the surface. We could not control the severity of infection on the immature fruit halves, meaning that some infections at 8 d could have reached the desired severity that other fruit would need 10 d to reach. However, all fruit used on a given day were inoculated on the same date (e.g., on a given testing day, all fruit used were 10 d old), and we selected fruit to roughly match in severity, discarding fruit that was over or under our standard. In general, we chose fruit that had roughly 20–50 opaque ooze droplets on the skin, with two to three patches of yellow brown to brown skin. Any fruit that was dark black, shriveling, or growing mold was discarded. Any fruit that did not produce significant amounts of ooze was discarded. Prior to use in preference assays, the flesh side of the fruit was covered with parafilm, as flies in pretrials that located the flesh of the fruit often roosted permanently on the flesh, which could have confounded the results. On the morning of each experiment day, six (three males, three females), 5- to 7-d-old D. platura per replicate were removed from the colony and transitioned into a cubic BugDorm containing only a water source in the environmental chamber that experiments would be conducted in so flies could acclimate to the chamber for 4–6 h. At 2 pm, fruitlet halves were placed on the fruit swings and arenas were sealed. Six flies (three males, three females) were captured into a sterile 15-ml conical tube, and the tube was screwed onto the cap that had been glued to the bottom of the arena and left in place until the end of the experiment. Over the next 2 h, every 2 min, the number of flies on each fruit was recorded. The time period and observation intervals were chosen based on previous experiments finding high rates of activity during this time (Hough-Goldstein and Bassler 1988). At the end of 2 h, flies and fruit were discarded, and all parts of the arena were rinsed with 95% ethanol and washed with warm soapy water. Total counts of flies on each fruit in the arena were tabulated. We conducted three independent preference assays with 20 replicates in each assay: 1) mock inoculated fruit versus oozing fruit; 2) mock inoculated fruit versus inoculated fruit with ooze removed; and 3) oozing fruit versus inoculated fruit with ooze removed. We also conducted controls for each fruit type (fruit with same conditions on both sides) and another with no fruit to test for any ‘side’ biases in the arenas. We ran 10 control replicates for each, totaling 40 control replicates, and there were no side biases in any of the controls. The fixed effect of treatment (the two fruit types used in each assay) on count of flies landing on each fruit in the arena was analyzed in R with a generalized linear mixed effects model and a negative binomial distribution. Replicate nested within block was included as a random effect and the analysis was conducted using the glmmTMB package (Brooks et al. 2017, R Core Team 2019). Comparisons between fruit treatments were made using estimated marginal means in the emmeans package (Lenth 2020). Data were visualized in R using the ggplot2 package (Wickham 2016). Shoot Preference Delia platura preference for infected or uninfected apple saplings were tested in groups of 100, 5- to 7-d-old flies in 93 cm × 47.5 cm × 47.5 cm BugDorm cages (catalog no. 4S4590, MegaView Science Co., Ltd., Taiwan) using the 47.5-cm square as the base. The arena design was based on similar preference bioassays conducted on various Diptera (Hough-Goldstein and Bassler 1988, Weston and Miller 1989, Sánchez-Alcañiz et al. 2017, Cha et al. 2020). All experiments were conducted in an environmental chamber matching the conditions used for colony maintenance, and flies were transitioned from the colony chamber to the experiment chamber in a cage containing diet and a water source 12–15 h prior to testing so flies could acclimate to the chamber. In preparation for experiments, apple saplings (cv. Brookfield Gala [Malus domestica] bench grafted onto EMLA 26 rootstock) 6.35 mm in diameter (Schlabach Nursery, Medina, NY) were potted into 15.24-cm square pots with a 3:1 mixture of LM-3 All Purpose Mix (Lambert, Quebec, Canada) and greenhouse sand and maintained in a greenhouse for 3 wk with a 14:10 D:N cycle and 22:18°C D:N temperature cycle. Plants were watered every other day or as needed, and all rootstock growth and bloom was removed by hand. Hypoaspis mites (Biobest Group, Westerlo, Belgium) were used as a biocontrol for thrips in the greenhouse. After 3 wk of growth, saplings were inoculated with E. amylovora or mock treated. Erwinia amylovora inoculum was prepared in LB broth as described earlier. Two of the three shoots growing on the sapling were removed, and the youngest leaf on the remaining shoot was bisected with the scissors what was dipped into the inoculum. Mock inoculated saplings were treated the same way, but sterile LB broth was used for the scissor dip. Saplings were then transferred to a mist chamber and arranged on the floor of the chamber about 1 m apart to avoid cross-contamination. The chamber was set to 24°C, and plants were kept in the chamber for 2, 5, or 10 d depending on the assay being conducted. The mist in the chamber was turned on for 3–4 h/d or as needed to maintain a relative humidity between 85 and 100%. Two potted apple saplings were placed 13 cm apart, 2 cm from the sides of cage, and 5 cm from the back of the cage. Two-liter paint strainer bags were used to sheath the pots (W. W. Grainger, Inc., Lake Forest, IL) to keep flies away from the soil such that only ~10 cm of scion and the entire shoot were available to the flies. Two webcams (model no. C615, Logitech, Lausanne, Switzerland) were attached to ring stands with rubber bands so that each camera was facing a single sapling, and adjusted for height on the ring stand so that the entire shoot was visible. The front edge of the ring stand base was 10 cm in front of the sapling, such that the back edge of the ring stand was flush with the front edge of the base of the cage. The webcam cables were run through the upper arm sleeve and connected to a laptop computer running ImageJ image processing software (NIH, Rockville, MD) with the Webcam Capture plugin installed. Using this plugin, webcams were programmed to take a photo every minute for 8 h from 8:30 am to 4:30 pm for a total of 480 frames per sapling (960 total frames per rep, Supp Fig. 2 [online only]). Flies were collected into a single vial and released into the cage by placing the vial at the exact center of the cage base and removing the cap immediately prior to starting the camera feed. A large piece of black cloth was placed over the top of the arena such that it hung 30 cm on all sides to avoid stimulating flies with light and chambers were closed off for the duration of the experiment to avoid disturbing or interrupting fly behavior. The location of infected and uninfected plants was switched every other testing day to ensure no side biases were present. The number of flies on each sapling shoot was counted for every frame captured. Only flies that were visible were counted, and if we could not tell if a fly was on a shoot or not, it was not counted. If part of the fly was visible (e.g., if the fly was on the back-facing side of the shoot), that fly was counted as present on the shoot. We conducted three independent assays to evaluate if preferences change as disease symptoms become more severe: 2 d post-mock inoculation (DPMI) versus 2 d post-inoculation (DPI); 2) 5 DPMI versus 5 DPI; and 3) 10 DPMI versus 10 DPI. Plants were never reused (e.g., plants used in 2 DPI trials were discarded, not used for 5 DPI trials), and 10 replicates of each assay were conducted. We spot checked for side biases in the assay by running five control replicates using uninfected, 3-wk-old saplings and no side biases were found. Five and 10 DPI shoots were exhibiting symptoms at the time of testing, but 2 DPI shoots were not. The point in testing 2 DPI shoots was show whether flies responded to infection before physical symptoms developed. After use, 2 DPI plants were allowed to develop symptoms to confirm E. amylovora infection. The number of flies in each frame for a given sapling was summed and data were analyzed and visualized as described for fruit preference assays. Odor Preference Individual D. platura were tested for preference between infected and uninfected shoots or infected and uninfected fruit using a glass Y-tube olfactometer (Cha et al. 2011, Wong et al. 2018). The olfactometer was 3 cm in diameter, with a 30-cm-long base and 5-cm-long arms. 3.175-mm inert plastic tubing (W. W. Grainger, Inc., Lake Forest, IL) was secured to an airline tap with a hose clamp and connected to an activated charcoal filter to purify the air. The filter output was connected via tubing to a round bottom flask containing dH2O to humidify the air, which was then passed through a Y-shaped splitter to divide the flow in two. Each arm of the splitter was connected via tubing to a flow regulator (catalog no. 97004–640, VWR International, Radnor, PA), and air was pushed into each odor source chamber at 1 liter/min. Odor sources were connected via tubing to specialized glass inserts that slid into the arms of the Y-tube and sealed the flow, so all air was pushed down to the base of the Y-tube (Supp Fig. 3 [online only]). This apparatus was set up in an environmental chamber matching the conditions using for colony maintenance. A black canvas was erected around the Y-tube to avoid stimulating flies with human movements. Two independent assays were conducted to compare fruit odor preferences: 1) mock inoculated fruit versus inoculated fruit and 2) fresh fruit versus inoculated fruit. Mock and infected fruit were prepared as described earlier, and fresh fruit was prepared by halving room temperature, immature fruit that had been removed from cold storage on the morning of experimentation. Fruit odor sources were contained in 75-ml glass jars, with a plastic lid fitted with input and output sources. Tubing was run from the flow regulator to the jar, then separately from the jar to the Y-tube. Air was run through the system for at least 1 h prior to experimentation to ensure no odors associated with handling the fruit were present when flies were tested. Three independent assays matching those conducted for shoot preference were conducted to evaluate shoot odor preference. Shoots were prepared as described earlier and contained in large (406 mm × 444 mm oven bags (Reynolds, Inc., Richmond, VA) that had been punctured on both sides with a #5 cork borer (Raguso and Willis 2005). Each hole in the bag was fitted with a plastic fitting, so tubing could be attached, with one acting as the input, receiving air after it passed through the flow regulator, and the other acting as the output, pushing air into the Y-tube. Only shoots and a wax covered section of the scion were contained in the bag, which was sealed tightly at its open end with twist ties. Bags were held upright by attaching binder clips to the top of the bag and the running wire that was secured to a wall through the clips, which prevented bending and breaking of shoots that could have led to the emission of damage associated volatiles. Once the bags were in place, the bag inflated naturally and plants were left with air running for 1 h to flush the chamber of volatiles associated with handling (Cha et al. 2008). On each testing day, 25, 5- to 7-d-old D. platura were isolated at 8 am in individual plastic vials that had the bottom replaced with mesh fabric so air could flow out of the Y-tube when vials were attached. Vials were then plugged with cotton and laid on their side in the testing chamber. Replicates were tested from 2 to 4 pm to match previous experiments (Hough-Goldstein and Bassler 1988). Flies were sexed, the cotton plug was removed, and the vial was inserted into the base of the Y-tube. Flies were timed for 5 min, after which they were removed. Flies that had not made a choice were recorded. Replicates were terminated before 5 min if it was determined that the fly had made a choice. We counted choices when flies exhibited settling behavior. In this assay, that behavior was defined as having entered an arm and remained in that arm for greater than 30 s. If a fly remained for 30 s, it generally stopped walking and roosted for extended time. The leads pushing air into the Y-tube were switched to the opposite arm after every replicate was tested to avoid side biases, and all glassware that a fly could contact were cleaned after every fifth replicate. Glassware was cleaned by first soaking for 10 min in warm water (49°C) mixed with powdered enzymatic detergent (catalog no. 04-358-23, Fisher Scientific, Waltham, MA), then rinsed with 95% ethanol using a squirt bottle and rinsed again with acetone before being set to dry completely. The same odor sources were used for every replicate tested on a given day and then discarded. We conducted 50 replicates for each of the five assays, with a replicate defined as a fly that successfully made a choice. This usually required running 60–70 total flies to reach 50 successes. We also conducted a 50-replicate control in which only humidified air was pushed through the apparatus to demonstrate that flies would be active in the given experimental conditions. Overall preferences for odor sources tested were analyzed in R using a χ 2 goodness of fit test with the null hypothesis that flies would choose each treatment in a 1:1 ratio (R Core Team 2019). Sex-based preferences for odor sources were analyzed the same way. Data were visualized in R as described earlier. Results Fruit Preference Delia platura exhibited a preference for diseased fruit when given a choice between infected fruit and mock inoculated, uninfected fruit (Fig. 1A: t-ratio = 2.077, P = 0.045). Flies given a choice between infected fruit and infected fruit with no ooze exhibited no preference (Fig. 1B: t-ratio = 0.875, P = 0.388). No preference between infected fruit with ooze removed and mock inoculated fruit was found (Fig. 1C: t-ratio = 0.069, P = 0.946). Fig. 1. Open in new tabDownload slide Mean ± 95% confidence interval of total D. platura counts on apple fruit when given a choice between (A) a fruit infected with Erwinia amylovora with ooze and a mock inoculated fruit; (B) a fruit infected with E. amylovora with ooze removed and a mock inoculated fruit; and (C) a fruit infected with E. amylovora and a fruit infected with E. amylovora with ooze removed. Black points show total counts of flies on fruit for individual replicates. There were significantly more flies counted on infected fruit when given a choice between infected fruit and mock inoculated fruit (A: P = 0.045), but not when given a choice between infected fruit with ooze removed and mock inoculated fruit (B: P = 0.388), or when given a choice between infected fruit and infected fruit with ooze removed (C: P = 0.946). *P < 0.05. Fig. 1. Open in new tabDownload slide Mean ± 95% confidence interval of total D. platura counts on apple fruit when given a choice between (A) a fruit infected with Erwinia amylovora with ooze and a mock inoculated fruit; (B) a fruit infected with E. amylovora with ooze removed and a mock inoculated fruit; and (C) a fruit infected with E. amylovora and a fruit infected with E. amylovora with ooze removed. Black points show total counts of flies on fruit for individual replicates. There were significantly more flies counted on infected fruit when given a choice between infected fruit and mock inoculated fruit (A: P = 0.045), but not when given a choice between infected fruit with ooze removed and mock inoculated fruit (B: P = 0.388), or when given a choice between infected fruit and infected fruit with ooze removed (C: P = 0.946). *P < 0.05. Shoot Preference Delia platura did not exhibit preference between infected and mock infected saplings that had been allowed 2 d to develop symptoms (Fig. 2A: t-ratio = −0.032, P = 0.975), or after 5 d of symptom development (Fig. 2B: t-ratio = −0.129, P = 0.899), or at 10 d of symptom development (Fig. 2C: t-ratio = −0.848, P = 0.410). Fig. 2. Open in new tabDownload slide Mean ± 95% confidence interval of total Delia platura counts on apple saplings when given a choice between a sapling infected with Erwinia amylovora and a sapling that was mock inoculated at three time points post-inoculation (DPI). Black points show total count of flies on saplings for individual replicates. There was no statistical difference in fly choice between infected and mock infected saplings after 2 d of symptom development (A: P = 0.975). There was no statistical difference between infected and mock infected saplings after 5 d of symptom development (B: P = 0.899). There was no statistical difference between infected and mock infected saplings after 10 d of symptom development (C: P = 0.410). Two DPI saplings were not exhibiting disease symptoms; 5 and 10 DPI saplings were exhibiting disease symptoms. Fig. 2. Open in new tabDownload slide Mean ± 95% confidence interval of total Delia platura counts on apple saplings when given a choice between a sapling infected with Erwinia amylovora and a sapling that was mock inoculated at three time points post-inoculation (DPI). Black points show total count of flies on saplings for individual replicates. There was no statistical difference in fly choice between infected and mock infected saplings after 2 d of symptom development (A: P = 0.975). There was no statistical difference between infected and mock infected saplings after 5 d of symptom development (B: P = 0.899). There was no statistical difference between infected and mock infected saplings after 10 d of symptom development (C: P = 0.410). Two DPI saplings were not exhibiting disease symptoms; 5 and 10 DPI saplings were exhibiting disease symptoms. Odor Preference Seventy-five percent of D. platura responded to odor stimuli when presented with a choice between an infected sapling and a mock infected sapling after 2 d of symptom development (DPI). There was no difference between choice of sapling (Fig. 3A: χ 2 = 1.28, P = 0.337) and no effect of sex on sapling choice (χ 2 = 0.335, P = 0.774). Seventy-one percent of D. platura responded to odor stimuli when presented with a choice between an infected sapling and a mock infected sapling after 5 d of symptom development. There was no difference between choice of sapling (Fig. 3B: χ 2 = 0.08, P = 0.883) and no effect of sex on sapling choice (χ 2 = 2.826, P = 0.156). Eighty-six percent of D. platura responded to odor stimuli when presented with a choice between an infected sapling and a mock infected sapling after 10 d of symptom development (DPI). Significantly more flies chose the mock inoculated sapling than the infected one (Fig. 3C: χ 2 = 9.680, P = 0.003) and there was no effect of sex on sapling choice (χ 2 = 0.397, P = 0.764). Fig. 3. Open in new tabDownload slide Frequency of Delia platura that chose each odor source when presented with a sapling infected with Erwinia amylovora and a mock inoculated sapling of the same day post-inoculation (DPI). There was no preference for either sapling at 2 DPI (A: P = 0.337) and 5 DPI (B: P = 0.883), but D. platura preferred mock inoculated plants at 10 DPI (C: P = 0.003). **P < 0.01. Fig. 3. Open in new tabDownload slide Frequency of Delia platura that chose each odor source when presented with a sapling infected with Erwinia amylovora and a mock inoculated sapling of the same day post-inoculation (DPI). There was no preference for either sapling at 2 DPI (A: P = 0.337) and 5 DPI (B: P = 0.883), but D. platura preferred mock inoculated plants at 10 DPI (C: P = 0.003). **P < 0.01. When presented with a choice between an infected immature apple fruit and a mock inoculated immature apple fruit, 83% of D. platura responded to odor stimuli. Significantly more flies chose the mock inoculated fruit (Fig. 4A: χ 2 = 13.52, P = 0.002), and there was no effect of sex on fruitlet choice (χ 2 = 0.967, P = 0.500). When presented with a choice between an infected immature apple fruit and a fresh immature apple fruit, 85% of D. platura responded to odor stimuli. There was no difference between fruit choice (Fig. 4B: χ 2 = 0, P = 1), and there was no effect of sex on fruit choice (χ 2 = 0.739, P = 0.551). Fig. 4. Open in new tabDownload slide Frequency of Delia platura that chose each odor source when presented with an immature fruit infected with Erwinia amylovora and (A) a mock inoculated immature fruit of the same day post-inoculation or (B) a fresh immature fruit. Delia platura preferred mock inoculated fruit over infected fruit (A: P = 0.002) but exhibited no preference when offered a fresh fruit and an infected fruit (B: P = 1.000). **P < 0.01. Fig. 4. Open in new tabDownload slide Frequency of Delia platura that chose each odor source when presented with an immature fruit infected with Erwinia amylovora and (A) a mock inoculated immature fruit of the same day post-inoculation or (B) a fresh immature fruit. Delia platura preferred mock inoculated fruit over infected fruit (A: P = 0.002) but exhibited no preference when offered a fresh fruit and an infected fruit (B: P = 1.000). **P < 0.01. Discussion Collectively, our results indicate that physical disease characteristics and changes to odor in plants infected with E. amylovora cause minor behavioral changes to D. platura in accordance with the Vector Manipulation Hypothesis (Ingwell et al. 2012). We show that D. platura preferred infected fruit when given a choice between an infected fruit and an uninfected fruit, but this preference subsided when ooze droplets were removed from the fruit surface. When given a choice between infected fruit and infected fruit with ooze droplets removed, D. platura again exhibited no preference for either choice. When challenged with the odors of infected and uninfected fruit, D. platura strongly preferred odors of uninfected fruit. However, this preference subsided when flies were given a choice between infected fruit odor and fresh fruit odor. In sapling assays, D. platura exhibited no preference between infected and uninfected saplings at 2, 5, and 10 DPI based on physical characteristics, and no preference based on odors at 2 and 5 DPI. Interestingly, flies did exhibit a preference for uninfected sapling odor at 10 DPI. The behaviors observed in this study align with the extension of the VMH presented by Cellini et al. (2019), which posited that the nonspecific nature of insect transmission of E. amylovora creates opportunities for the pathogen to exploit existing plant–insect interactions to enhance transmission (Cellini et al. 2019). In this scenario, the bacterium benefits from D. platura foraging behavior as the insect searches for water droplets or hemipteran honeydew on infected fruit (Miller and McClanahan 1960), and the preference and odor dynamics mediated by the plant increase transmission likelihood. We did not expect to find that D. platura preferred uninfected fruit odor relative to infected fruit odor. However, this finding combined with the flies’ propensity for physical, oozing fruit could benefit transmission. In our assay, uninfected fruit were mock inoculated in part by wounding the fruit surface, mimicking the wound requirement for E. amylovora colonization. Ecologically, occasional visitation and preferential arrestment on oozing fruit followed by preference for the odor of damaged, uninfected fruit could allow for the significant spread of E. amylovora to new hosts. In terms of the VMH, the odor produced by infected fruit could be deterrent, or the odor produced by wounded, uninfected fruit could be attractive. Given that flies preferred infected fruit when given access to the fruit itself, we do not believe that infected fruit odors were deterrent to D. platura. The observed preference for wounded, uninfected fruit odor could be caused by production of attractive volatiles by the microbial community present on the fruit (Farré-Armengol et al. 2016). Microbes and their byproducts are capable of producing volatile blends that elicit behavioral responses from insects (Becher et al. 2012), and D. platura are responsive to microbe-associated cues (Hough-Goldstein and Bassler 1988, Weston and Miller 1989). In our study, mock-treated fruit were maintained in identical conditions to infected fruit, whereas fresh fruit were kept in cold storage until the day of testing. The former provided ample opportunity for microbial colonization and subsequent volatile production, whereas the latter did not (Mercier and Wilson 1994, Teixidó et al. 1999, Shen et al. 2018), potentially explaining why preference for uninfected fruit differed between mock treated and fresh fruit and suggesting that flies are overall less likely to visit undamaged fruit. This case thus provides another example of how E. amylovora may take advantage of existing plant–insect(–microbe) interactions (Cellini et al. 2019). We hypothesized that the physical preference for infected fruit relative to uninfected fruit occurred because bacterial ooze acted as an arrestant that kept flies on fruit if they discovered an ooze droplet. We attempted to demonstrate the arresting qualities of ooze by comparing D. platura preference for infected fruit without ooze against 1) uninfected fruit or 2) oozing, infected fruit. We predicted that 1) flies would prefer oozing, infected fruit over infected fruit without ooze and that 2) there would be no preference between infected fruit with ooze removed and uninfected fruit. We found no preference in both assays, potentially confounding our understanding of how ooze affects D. platura choice. However, it is possible that ooze and the sugars that constitute it were not completely removed from the fruit despite our best efforts to do so. The remnants of bacterial ooze on the fruit could have diluted preference for an oozing fruit relative to a diseased fruit with ooze removed, as flies could have spent time feeding on the remnants. Biologically, this scenario may approximate what occurs in the field after a rainstorm and suggests that the overall behavioral effect of E. amylovora ooze on D. platura is relatively weak and is easily interrupted. The physical characteristics of saplings infected with E. amylovora appear to have no behavioral effect on D. platura under the conditions we tested. Infected shoots on saplings turn brownish black over time and wilt, developing a characteristic shepherd’s crook appearance (Koczan et al. 2009). Other Delia spp. use plant shape and color to identify suitable hosts, but D. platura is hypothesized to rely more strongly on odor cues than vision and is able to identify a wide range of potential hosts (Gouinguené and Städler 2006). Our data align with this hypothesis of D. platura host finding, as the changes in sapling shoot appearance caused by E. amylovora infection did not alter the behavior of D. platura. Importantly, apple saplings are not considered ovipositional hosts for D. platura adults, serving only as a potential refuge from high temperatures and as a food source location (Miller and McClanahan 1960, Higley and Pedigo 1984). Thus, the cues generally employed by D. platura during host finding may not be active in this system, precluding behavioral manipulation based on them. Then again, a lack of overt behavioral manipulation might allow the pathogen to take advantage of existing interactions between insects and plants. This principle explains D. platura’s response to shoot odor in this assay as well. At 2 and 5 DPI, flies exhibit no preference between infected and uninfected shoot odors, suggesting regular, nonodor-based interchange between infected and uninfected shoots. If a fly acquires E. amylovora from ooze and randomly lands on different shoots, then transmission is more likely to occur relative to a scenario where a fly has an odor-based preference for infected or uninfected shoot odors. The strong preference for uninfected plant odors at 10 DPI is likely a byproduct of infected shoot death at this stage, which still benefits E. amylovora because interactions between flies and uninfected shoots would increase, likely leading to higher rates of transmission to new hosts if flies are infectious. Overall, the VMH provides a useful framework for understanding interactions between D. platura and E. amylovora (Eigenbrode et al. 2018). In this system, low vector specificity might preclude the development of specific manipulative cues in favor of maximizing interactions with as many insect species as possible. The resultant low intensity of manipulation may then be a hallmark of low-specificity pathosystems, and some occasional vectors may be more actively manipulated than others depending on how the various cues they use to forage, oviposit, and mate overlap with the physical and volatile characteristics of infected plants. This principle is exemplified by stronger manipulation of honeybees during blossom blight relative to the case explored in this study (Cellini et al. 2019). Moreover, the existence of noninsect transmission may reduce selection for traits that promote insect behavior change, further incentivizing pathogen capitalization on interactions already at play in the system (Cellini et al. 2019). For example, D. platura are known to forage on hemipteran honeydew (Miller and McClanahan 1960), and could do so on a shoot that was simultaneously infested with E. amylovora ooze and hemipterans (Stewart and Leonard 1916). The indirect interaction between D. platura and another insect in the community could thus bring the fly into close contact with the pathogen and increase transmission potential. Likewise, this type of indirect interaction requires further attention in other pathosystems with nonspecific vectors. For instance, Lonsdalea quercina (Hildebrand & Schroth) (Enterobacterales: Pectobacteriaceae), an ooze-producing, emerging bacterial pathogen of red oak, retains a suite of insect visitors to ooze (Sitz et al. 2019), but it is unknown whether some form of vector manipulation occurs. Soft rot phytopathogens represent another low-specificity system of interest as do various exudate producing phytopathogenic fungi (Russin et al. 1984, Eigenbrode et al. 2018, Rossmann et al. 2018). Evaluating these pathogens through a VMH lens will provide support for this hypothesis across a spectrum of pathogen-vector relationships and provide context for how complex behavioral and community interactions contribute to the success of economically important phytopathogens. Acknowledgments We acknowledge the efforts of Françoise Vermeylen at the Cornell Statistical Consulting Unit for helpful comments and guidance regarding statistics. This research was funded by National Science Foundation Graduate Research Fellowship (no. DGE-1650441); Northeast Sustainable Agriculture Research and Education Graduate Student Grant (no. GNE16-115–29994); New York State Department of Agriculture and Markets Apple Research Development Program (no. C200849); Federal Capacity Funds (no. 2016-17-199), Grace Griswold Endowment, Arthur Boller Research Fund. References Cited Becher , P. G. , G. Flick, E. Rozpedowska, A. Schmidt, A. Hagman, S. Lebreton, M. C. Larsson, B. S. Hansson, J. Piškur, P. Witzgall, and M. Bengtsson. 2012 . Yeast, not fruit volatiles mediate Drosophila melanogaster attraction, oviposition and development . Funct. Ecol . 26 : 822 – 828 . Google Scholar Crossref Search ADS WorldCat Bogs , J. , I. Bruchmüller, C. Erbar, and K. Geider. 1998 . Colonization of host plants by the fire blight pathogen Erwinia amylovora marked with genes for bioluminescence and fluorescence . Phytopathology 88 : 416 – 421 . Google Scholar Crossref Search ADS PubMed WorldCat Boucher , M. , R. Collins, K. Cox, and G. Loeb. 2019 . Effects of exposure time and biological state on acquisition and accumulation of Erwinia amylovora by Drosophila melanogaster . Appl. Environ. Microbiol . 85 : e00726 – 19 . Google Scholar Crossref Search ADS PubMed WorldCat Boucher , M. , R. Collins, K. Harling, G. Brind’Amour, K. Cox, and G. Loeb. 2020 . Interactions between Delia platura and Erwinia amylovora associated with insect mediated transmission of shoot blight . PhytoFrontiers (in press). doi:10.1094/PHYTOFR-08-20-0013. Google Scholar OpenURL Placeholder Text WorldCat Brooks , M. E. , K. Kristensen, K. J. van Benthem, A. Magnusson, C. W. Berg, A. Nielsen, H. J. Skaug, M. Maechler, and B. M. Bolker. 2017 . glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling . R J . 9 : 378 – 400 . Google Scholar Crossref Search ADS WorldCat Cellini , A. , G. Buriani, L. Rocchi, E. Rondelli, S. Savioli, M. T. Rodriguez Estrada, S. M. Cristescu, G. Costa, and F. Spinelli. 2018 . Biological relevance of volatile organic compounds emitted during the pathogenic interactions between apple plants and Erwinia amylovora . Mol. Plant Pathol . 19 : 158 – 168 . Google Scholar Crossref Search ADS PubMed WorldCat Cellini , A. , V. Giacomuzzi, I. Donati, B. Farneti, M. T. Rodriguez-Estrada, S. Savioli, S. Angeli, and F. Spinelli. 2019 . Pathogen-induced changes in floral scent may increase honeybee-mediated dispersal of Erwinia amylovora . ISME J . 13 : 847 – 859 . Google Scholar Crossref Search ADS PubMed WorldCat Cha , D. H. , S. Nojima, S. P. Hesler, A. Zhang, C. E. Linn , Jr., W. L. Roelofs, and G. M. Loeb. 2008 . Identification and field evaluation of grape shoot volatiles attractive to female grape berry moth (Paralobesia viteana) . J. Chem. Ecol . 34 : 1180 – 1189 . Google Scholar Crossref Search ADS PubMed WorldCat Cha , D. H. , G. M. Loeb, C. E. Linn, and W. L. Roelofs. 2011 . Electrophysiological and behavioral identification of a volatile blend involved in host location of female strawberry sap beetle, Stelidota geminata . Entomol. Exp. Appl . 140 : 153 – 162 . Google Scholar Crossref Search ADS WorldCat Cha , D. H. , S. P. Hesler, G. Brind’Amour, K. S. Wentworth, S. Villani, K. D. Cox, M. T. Boucher, A. Wallingford, S. K. Park, J. Nyrop, et al. 2020 . Behavioral evidence for contextual olfactory-mediated avoidance of the ubiquitous phytopathogen Botrytis cinerea by Drosophila suzukii . Insect Sci . 27 : 771 – 779 . Google Scholar Crossref Search ADS PubMed WorldCat Crosse , J. E. , and R. N. Goodman. 1973 . A selective medium for and a definitive colony characteristic of Erwinia amylovora . Phytopathology 63 : 1425 . Google Scholar Crossref Search ADS WorldCat Eigenbrode , S. D. , N. A. Bosque-Pérez, and T. S. Davis. 2018 . Insect-borne plant pathogens and their vectors: ecology, evolution, and complex interactions . Annu. Rev. Entomol . 63 : 169 – 191 . Google Scholar Crossref Search ADS PubMed WorldCat Emmett , B. J. , and L. A. E. Baker. 1971 . Insect transmission of fire blight . Plant Pathol . 20 : 41 – 45 . Google Scholar Crossref Search ADS WorldCat Farré-Armengol , G. , I. Filella, J. Llusia, and J. Peñuelas. 2016 . Bidirectional interaction between phyllospheric microbiotas and plant volatile emissions . Trends Plant Sci . 21 : 854 – 860 . Google Scholar Crossref Search ADS PubMed WorldCat Gossard , H. A. , and R. C. Walton. 1922 . Dissemination of fire blight . Bull. Ohio Agric. Exp. Stn . 537 : 83 – 126 . Google Scholar OpenURL Placeholder Text WorldCat Gouinguené , S. P. , and E. Städler. 2006 . Oviposition in Delia platura (Diptera, Anthomyiidae): the role of volatile and contact cues of bean . J. Chem. Ecol . 32 : 1399 – 1413 . Google Scholar Crossref Search ADS PubMed WorldCat Higley , L. G. , and L. P. Pedigo. 1984 . Seedcorn maggot (Diptera: Anthomyiidae) population biology and aestivation in Central Iowa . Environ. Entomol . 13 : 1436 – 1442 . Google Scholar Crossref Search ADS WorldCat Hough-Goldstein , J. A. , and M. A. Bassler. 1988 . Effects of bacteria on oviposition by seedcorn maggots (Diptera: Anthomyiidae) . Environ. Entomol . 17 : 7 – 12 . Google Scholar Crossref Search ADS WorldCat Ingwell , L. L. , S. D. Eigenbrode, and N. A. Bosque-Pérez. 2012 . Plant viruses alter insect behavior to enhance their spread . Sci. Rep . 2 : 578 . Google Scholar Crossref Search ADS PubMed WorldCat Johnson , K. B. , and V. O. Stockwell. 1998 . Management of fire blight: a case study in microbial ecology . Annu. Rev. Phytopathol . 36 : 227 – 248 . Google Scholar Crossref Search ADS PubMed WorldCat Koczan , J. M. , M. J. McGrath, Y. Zhao, and G. W. Sundin. 2009 . Contribution of Erwinia amylovora exopolysaccharides amylovoran and levan to biofilm formation: implications in pathogenicity . Phytopathology 99 : 1237 – 1244 . Google Scholar Crossref Search ADS PubMed WorldCat Lenth , R . 2020 . emmeans: estimated marginal means, aka least-squares means. (https://cran.r-project.org/package=emmeans). Mercier , J. , and C. I. Wilson. 1994 . Colonization of apple wounds by naturally occurring microflora and introduced Candida oleophila and their effect on infection by Botrytis cinerea during storage . Biol. Control 4 : 138 – 144 . Google Scholar Crossref Search ADS WorldCat Miller , L. A. , and R. J. McClanahan. 1960 . Life history of the seedcorn maggot, Hylemya cilicrura (Rond.) and of H. liturata (Mg.) (Diptera: Anthomyiidae) in Southwestern Ontario . Can. Entomol . 92 : 210 – 221 . Google Scholar Crossref Search ADS WorldCat Nadarasah , G. , and J. Stavrinides. 2011 . Insects as alternative hosts for phytopathogenic bacteria . FEMS Microbiol. Rev . 35 : 555 – 575 . Google Scholar Crossref Search ADS PubMed WorldCat Norelli , J. L. , H. S. Aldwinckle, and S. V. Beer. 1984 . Differential host x pathogen interactions among cultivars of apple and strains of Erwinia amylovora . Phytopathology 74 : 136 – 139 . Google Scholar Crossref Search ADS WorldCat Norelli , J. L. , A. L. Jones, and H. S. Aldwinckle. 2003 . Fire blight management in the twenty-first century: using new technologies that enhance host resistance in apple . Plant Dis . 87 : 756 – 765 . Google Scholar Crossref Search ADS PubMed WorldCat Ordax , M. , J. E. Piquer-Salcedo, R. D. Santander, B. Sabater-Muñoz, E. G. Biosca, M. M. López, and E. Marco-Noales. 2015 . Medfly Ceratitis capitata as potential vector for fire blight pathogen Erwinia amylovora: survival and transmission . PLoS One 10 : e0127560 . Google Scholar Crossref Search ADS PubMed WorldCat Parker , K. G . 1936 . Fire blight: overwintering, dissemination, and control of the pathogen . New York Agric. Exp. Stn. Mem . 193 : 1 – 42 . Google Scholar OpenURL Placeholder Text WorldCat Purcell , A. H . 1982 . Insect vector relationships with procaryotic plant pathogens . Annu. Rev. Phytopathol . 20 : 397 – 417 . Google Scholar Crossref Search ADS WorldCat R Core Team. 2019 . R: a language and environment for statistical computing . (https://www.r-project.org/). Raguso , R. A. , and M. A. Willis. 2005 . Synergy between visual and olfactory cues in nectar feeding by wild hawkmoths, Manduca sexta . Anim. Behav . 69 : 407 – 418 . Google Scholar Crossref Search ADS WorldCat Rossmann , S. , M. W. Dees, J. Perminow, R. Meadow, and M. B. Brurberg. 2018 . Soft rot Enterobacteriaceae are carried by a large range of insect species in potato fields . Appl. Environ. Microbiol . 84 : e00281 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat Russin , J. S. , L. Shain, and G. L. Nordin. 1984 . Insects as carriers of virulent and cytoplasmic hypovirulent isolates of the chestnut blight fungus . J. Econ. Entomol . 77 : 838 – 846 . Google Scholar Crossref Search ADS WorldCat Sánchez-Alcañiz , J. A. , G. Zappia, F. Marion-Poll, and R. Benton. 2017 . A mechanosensory receptor required for food texture detection in Drosophila . Nat. Commun . 8 : 1 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Shapiro , L. , C. M. De Moraes, A. G. Stephenson, M. C. Mescher, and W. van der Putten. 2012 . Pathogen effects on vegetative and floral odours mediate vector attraction and host exposure in a complex pathosystem . Ecol. Lett . 15 : 1430 – 1438 . Google Scholar Crossref Search ADS PubMed WorldCat Shen , Y. , J. Nie, Y. Dong, L. Kuang, Y. Li, and J. Zhang. 2018 . Compositional shifts in the surface fungal communities of apple fruits during cold storage . Postharvest Biol. Technol . 144 : 55 – 62 . Google Scholar Crossref Search ADS WorldCat Sitz , R. A. , V. M. Aquino, N. A. Tisserat, W. S. Cranshaw, and J. E. Stewart. 2019 . Insects visiting drippy blight diseased red oak trees are contaminated with the pathogenic bacterium Lonsdalea quercina . Plant Dis . 103 : 1940 – 1946 . Google Scholar Crossref Search ADS PubMed WorldCat Slack , S. M. , Q. Zeng, C. A. Outwater, and G. W. Sundin. 2017 . Microbiological examination of Erwinia amylovora exopolysaccharide ooze . Phytopathology 107 : 403 – 411 . Google Scholar Crossref Search ADS PubMed WorldCat Stewart , V. B. , and M. D. Leonard. 1916 . Further studies in the role of insects in the dissemination of fire blight bacteria . Phytopathology 6 : 152 – 158 . Google Scholar OpenURL Placeholder Text WorldCat Teixidó , N. , J. Usall, N. Magan, and I. Viñas. 1999 . Microbial population dynamics on Golden Delicious apples from bud to harvest and effect of fungicide applications . Ann. Appl. Biol . 134 : 109 – 116 . Google Scholar Crossref Search ADS WorldCat Thomas , H. E. , and P. A. Ark. 1934 . Fire blight of pears and related plants . Calif. Agric. Exp. Stn. Bull . 586 1-43. Google Scholar OpenURL Placeholder Text WorldCat Webb , D. , and C. Eckenrode. 1978 . Simplified rearing and bioassay for the seedcorn maggot, Hylemya platura (Meigen) . NY State Agric. Exp. Stn . 72 : 1 – 4 . Google Scholar OpenURL Placeholder Text WorldCat Weston , P. A. , and J. R. Miller. 1989 . Ovipositional responses of seedcorn maggot, Delia platura (Diptera: Anthomyiidae), to developmental stages of lima bean . Ann. Entomol. Soc. Am . 82 : 387 – 392 . Google Scholar Crossref Search ADS WorldCat Wickham , H . 2016 . ggplot2: elegant graphics for data analysis . Springer-Verlag , New York . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Wong , J. S. , A. K. Wallingford, G. M. Loeb, and J. C. Lee. 2018 . Physiological status of Drosophila suzukii (Diptera: Drosophilidae) affects their response to attractive odours . J. Appl. Entomol . 142 : 473 – 482 . Google Scholar Crossref Search ADS WorldCat Yao , C. , G. Zehnder, E. Bauske, and J. Kloepper. 1996 . Relationship between cucumber beetle (Coleoptera: Chrysomelidae) density and incidence of bacterial wilt of cucurbits . J. Econ. Entomol . 89 : 510 – 514 . Google Scholar Crossref Search ADS WorldCat van der Zwet , T. , and H. L. Keil. 1979 . Fire blight. A bacterial disease of rosaceous plants . US Dep. Agric. Agric. Handb . 510 : 1 – 200 . Google Scholar OpenURL Placeholder Text WorldCat van der Zwet , T. , N. Orolaza-Halbrendt, and W. Zeller. 2012 . The disease cycle of fire blight, pp. 155 – 182 . In Fire blight: history, biology, and management . American Phytopathological Society , St. Paul, MN . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC © The Author(s) 2020. Published by Oxford University Press on behalf of Entomological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - The Effect of Erwinia amylovora Infection in Apple Saplings and Fruit on the Behavior of Delia platura (Diptera: Anthomyiidae)
JO - Environmental Entomology
DO - 10.1093/ee/nvaa153
DA - 2021-02-17
UR - https://www.deepdyve.com/lp/oxford-university-press/the-effect-of-erwinia-amylovora-infection-in-apple-saplings-and-fruit-8SPwgn9xVB
SP - 117
EP - 125
VL - 50
IS - 1
DP - DeepDyve
ER -