Between Oct. 5th and 30th, five experiments were conducted that included Fallglo (1 time), sunburst (2 times), red grapefruit (4 times), and navel oranges (1 time). Fruit were treated on a commercial packingline (3 experiments) or on the Indian River Research and Education Center research line (2 experiments). Treatments included 1) full wash (brush bed + high-pressure wash) + waxing (carnauba), 2) full wash, 3) brush bed only, 4) brush bed with brushes rotating half normal speed, 5) high-pressure wash (HPW) only, 6) HPW for 10 seconds, 7) HPW for 5 seconds, 8) running fruit only over PVC rollers, and 9) a control (not washed or waxed). On the commercial line, fruit remained on the brush bed for ~ 1 min. 10 seconds, and on the full HPW for ~35 seconds. Normal brush rotation speed was ~100 rpm. Fruit were also evaluated for how surface dirt obscured the ability to grade the fruit for canker and other grade defects. Fruit from all treatments were degreened under simulated commercial conditions (5 ppm ethylene, 85F, 95% RH) and color development and weight loss measured almost daily. Fruit were subsequently stored and evaluated for the development of decay and disorders during storage under ambient conditions (~70-75F). In general, all very early season Fallglo fruit were relatively clean and did not need washing for adequate grading. This changed somewhat by the end of October when grapefruit that received more extensive washing (i.e., full washing) was significantly easier to grade compared to unwashed fruit. HPW producing fruit with intermediate gradeability. However, even minimally washed fruit were sufficient for adequate grading. These experiments need to be repeated next season to determine variability in initial fruit cleanliness from year to year. Washing and waxing the fruit gave the greatest inhibition of degreening, almost stopping color development completely. Compared to preliminary results in 2008, results again showed that full washing of fruit on both the brush washer and HPW, or washing on the brush bed along inhibited degreening significantly more than did washing fruit only as they passed over the HPW. Fruit that were not brushed at all, but only passed over rollers experience a slight, but significant delay in color development compared to the control, but the delay was relatively minor compared to the other washing treatments. As the season progresses, fruit exterior surfaces become more soiled with dirt and sooty mold that makes grading more difficult without washing. Experiments are currently underway using late-season red grapefruit, harvested from a block with citrus canker. Unwashed fruit were run past commercial graders and the number of fruit with any canker or surface defects that would justify elimination based on export grade standards were counted. These fruit were then taken to a commercial packinghouse and either left un-washed (control), or washed over 1) the entire line (brush + HPW), 2) the brush line only, or 3) HPW only. Fruit were then evaluated by commercial graders again. Data is still being collected and analyzed from these experiments.
Research results from the first (2007-08) season were successfully completed, the results reported at the Florida State Horticultural Society (FSHS) meetings and at Packinghouse Day and the Indian River Postharvest Workshop in Sept. 2008, a manuscript published in the FSHS proceedings (http://www.fshs.org/; Proc. Fla. State Hort. Soc. 2008. 121:322’325), and subsequent results reported in FCPRAC progress reports. To summarize all experiments and results, preharvest treatments included 1) foliar K (8lb K2O), 2) foliar Mg (6% solution), 3) foliar K + Mg, 4) foliar Vapor Gard’ (1% or 2%), and 5) induced tree water stress by discontinuing irrigation and covering the ground with large plastic sheets. Treatments were conducted in commercial groves on White grapefruit three times, Flame red grapefruit twice, and Valencia oranges once. Fruit were harvested at various times after foliar applications (sometimes weekly), held for 3 to 4 days at ~70F with ~60% RH, washed & sometimes waxed (carnauba), and stored under ambient conditions. Fruit were evaluated at regular intervals for peel breakdown and decay. While foliar treatments did not always significantly reduce peel breakdown, even when the differences were not significant, such applications tended to reduce peel breakdown. In general, foliar K or Mg treatments significantly reduced peel breakdown once each. In one experiment, foliar K actually increased peel breakdown, but the reason for this is unclear. However, a combination of foliar K + Mg or use of an antitranspirant (Vapor Gard’) significantly reduced postharvest peel breakdown more than the control or K or Mg alone. Blocking irrigation and rainfall for 49 days before harvest increased peel breakdown in the first year. In the second year, while differences in peel breakdown were not significantly different, trees under water stress tended to develop twice as much peel breakdown compared to the control. In postharvest experiments, fruit were harvested, held for 3 to 4 days, washed and waxed (carnauba) on a research packingline at the IRREC, and then stored at ambient conditions. Postharvest treatments including 1) holding the fruit for 3 or 4 days after harvest at 30%, 60%, or 100% RH at ~70F, 2) including a fungicide (TBZ or Imazalil) in the wax, 3) not waxing the fruit, or 4) washing, but not rinsing the detergent from the fruit before storage. Exposing fruit to low (30%) RH after harvest dramatically increased peel breakdown and often significantly increased postharvest decay as well. Exposure to intermediate RH occasionally increased peel breakdown compared to the control (100% RH), but results were not consistent and sometimes were not significantly different from the control. Not waxing fruit or rinsing detergent from the fruit usually had no effect on peel breakdown. Interestingly, postharvest treatments with Imazalil occasionally reduced peel breakdown. The reason for this is unclear but worth further investigation.
Funding for this project has enabled us to develop a comprehensive program for ACP management. We are building programs targeting ACP adults using dormant sprays of broad-spectrum insecticides and treatments during the growing season based on scouting. Dormant sprays have evolved into a successful annual area-wide cooperative effort in SW Florida. For the growing season, we have developed a wide range of choices such as (1) recurrent low-volume applications of horticultural oil, (2) wide selection of tested selective and broad spectrum insecticides, and (3) strategies to conserve and augment biological control. A survey of local managers covering 106,148 acres of commercial citrus in SW Florida showed widespread adoption of these management strategies. Our first objective was to assist growers to plan and implement effective strategies for ACP control. All growers surveyed used dormant sprays during the 2008-2009 dormant season in cooperation with a voluntary area-wide program initiated and promoted in cooperation with Gulf Citrus Growers Association and Cooperative Extension. 71,916 acres were sprayed by air in 08/09 and 73,180 in the 09/10 seasons with an additional estimated 30,000 sprayed by ground, virtually all the present commercial acreage in SW Florida. Field counts obtained from DPI-CHRP and growers’ personnel trained by us indicated that populations by air were only 1.5X winter numbers three months into the growing season compared to a 28X increase in untreated blocks. Results of the post application survey will be evaluated this May. Our 2nd objective was to assist in development of efficient monitoring methods. The “tap” sampling method we developed has proved to be an effective and efficient tool for monitoring ACP adults, as well as other pests and beneficial insects. We integrated this and a technique for assessing immature stages into a rapid and easy to use system promoted in Citrus Industry Magazine, fully explained on our website http://swfrec.ifas.ufl.edu/ and used by 95% of surveyed growers. Our 3rd objective was to accelerate testing of new chemical tools for psyllid suppression. We completed four field evaluations of insecticides for ACP control in 2009 and submitted reports for publication in the ESA journal Arthropod Management Tests. These join 17 previously published reports on insecticidal control of ACP available online. We found that foliar applications of spirotetramat, fenproximate + 435 Oil, spirotetramat, or 435 Oil alone during bloom suppressed ACP populations for two months. Our 4th objective was to evaluate the economic feasibility of psyllid management strategies. Two aerial dormant sprays could cost less than $25/ac. The cost of monitoring psyllids using the tap method is about $25 per block up to 100 acres so could be done 20 times per year for as little as $5/ac, and would avoid unnecessary sprays during the growing season that could cost up to $50/ac. Unnecessary applications also cause secondary pest outbreaks of mites, scales and leafminers, reported by 70% of surveyed growers. Our 5th objective was to provide an information bridge between growers and researchers. Experimental results and recommended management programs have been presented at 18 extension meetings across Florida including two production managers meetings and two CCA trainings, as well as the Entomological Society of America, the Florida State Horticultural Society, the Florida Entomological Society, Citrus HLB & Potato Zebra Chip (ZC) in Mc Allen, TX, and the XIII Simposio Internacional de Citricultura. CD Victoria, Mexico. 91% percent of surveyed growers were satisfied at some level by their present management programs and 85% of respondents were monitoring ACP more than 3 times per year. We are working on emphasizing the use of this information to make informed decisions to use insecticides only when needed. (1)Stansly, PA, JA Qureshi, and HA Arevalo. 2009. Why, when and how to monitor and manage Asian citrus psyllid. Citrus Industry 90(3):24-26. (2) Stansly, PA, HA Arevalo, M Zekri, and R Hamel. 2009. Cooperative dormant spray program against Asian citrus psyllid in SW Florida. Citrus Industry 90(10):14-15.
Obj. 1.DNA bar coding in progress using mtCOI primers to amplify a 780 bp fragment from psyllid field and colony collections from various U.S. states and off-shore locations. The purpose of this objective is to examine diversity within and between varied populations in relation to the associated Liberibacter spp. based on 16S rDNA. Over 50 collections are in hand. Obj. 2. qPCR has been applied to detect Ca. Liberibacter presence (or absence) in the psyllid colony cohorts used for light and TEM studies, and by FISH (over different AAPs), for immatures and adults (Roberts, SWFREC, HLB diagnostics lab). The FL psyllid rearing system is fully functional and Ca. Liberibacter-infected and bacterium-free colonies are being maintained routinely. About 200 field-collected adult psyllids testing 10% PCR positive were placed into 2 more cages each with 6 ungrafted plants and 100 psyllids. Grafted plants were held in an air conditioned greenhouse at (28 ‘C 64% RH) and psyllid inoculated plants were held in an air conditioned insectary under similar conditions and 14 h light. All plants tested negative for HLB before they were infected with either method. After about one to one and a half month 85-90% grafted plants tested HLB positive based on PCR conducted on dark green mature leaves, whereas, plants exposed to infected psyllids still tested negative. When grafted plants were pruned to induce new shoots, 31-35% tested HLB positive by PCR. Two weeks later when PCR was conducted on leaves from young shoots 90-100% were positive. Four weeks later 97% tested positive by (dark green mature leaves). Two months after psyllid exposure to infected (PCR +) plants, 83% of the psyllid infested plants tested HLB positive by PCR of the dark green mature leaves. Adult psyllids collected on 8/12/2009 and tested at the SWFREC HLB lab were 61% and 67% HLB positive from budded and psyllid infected plants, respectively. Adult psyllids collected on 8/25/2009 and PCR tested at the USDA-ARS Riverside lab were 36% and 13% HLB positive from budded and psyllid infected plants, respectively. Obj. 3. Because it can be reared year-round and because of quarantine constraints we have adopted Bactericera cockerelli as a surrogate species for Diaphorina citri. All three sections, V1,V2 & V3, of the B. cockerelli ventriculus support bacterial growth and replication that advances into huge colonies in the epithelia and on the outside surfaces as well. The esophagus and external hindgut harbor small numbers on their outside surfaces. Thus focus on transmission is now shifted to the oral region. The salivary glands apparently do not support massive, organ-destroying cultures, however, high titers of bacteria occur in the hemolymph surrounding them. Their role in the terminal stages of transmission is still under study. FISH technique, involving coagulant fixation, adopted from prior works by various authors, have proven unsatisfactory for this species. New techniques are under development, including the rendering of individual organs on slides, and the de-shelling of aldehyde fixed whole bodies for acceptable in-situ localization. Obj. 4. We are investigating whether bacteria gain lateral access to the salivary canal without invading the glands. The most likely site would be near the precibarium, where the maxillary stylets converge (Ullman and McLean 1989) and glandular secretions are present. Studies are underway to determine the configuration of the oral area to aid in interpretations of liberibacter-anatomical TEM visualization.
Our initial efforts in using canines to detect canker started in 2000 but were thwarted by the September 11 attack on the US, which permanently diverted the canine we had been training (it was included in the much needed efforts to detect explosives). However, prior to this dog’s reassignment, our initial research with canines for canker detection demonstrated their ability to detect citrus and citrus canker volatiles/aromas/essences, i.e., minute concentrations of volatile/aromatic compounds, deposited on cloth and other materials as well as directly from the air. A second attempt with USDA, APHIS agriculture detector dogs was cut short by decreasing budgets combined by increasing needs of detection of contraband agricultural products at points of entry into the US. However, our third and ongoing attempt has been much more successful. This most recent canine, ‘Juice’, has demonstrated an ability to detect canker infected citrus fruit with high reliability. The initial success of this project with regards to canker detection, suggests that the same technique may be useful for detection of HLB, perhaps presymptomatic while the disease is still latent. Such a tool would be very useful and is much needed, as it would allow both detection and early removal of presymptomatic (latently infected) trees, which also may be below the titer required for psyllids transmission. Two dogs have been trained to recognize the scent of citrus canker. Dogs were trained on infected versus non-infected trees and fruit. Training of the canines for the detection of canker infected citrus trees was accomplished via a commercial canine training facility in North Florida. As indicated above, ‘Juice’, demonstrated an ability to detect canker infected citrus fruit with high reliability (> 99.3%). Juice was capable of discriminating canker infected from non-infected fruit in 30 runs of 5 blind stations each, with only one false positive and no false negatives (Peruyero and Gottwald, unpublished). Results indicate that canines can detect and differentiate citrus fruit infected with canker from non-infected tissues. The canine was then trained on infected versus non-infected Duncan grapefruit seedlings. Preliminary tests at the canine training facility are very promising. Field trials were conduced in early April at the USHRL farm. Trees (75) were placed in the field in a 3 row by 25 tree/row design. Ten replications were conducted in which the proportion of disease trees ranged from 2-10%. The conditions were sunny but with a 15-20 MPH breeze. Even so, detection highly reliable. Data are presently under analyses. Short movie footage was also taken to document the dogs search and detection behavior. More field trials are scheduled over the next few months to validate the canine’s performance. At the conclusion of the canker field testing (proof of concept) the dog will the be cross-trained for HLB detection and field tests will be conducted to assess performance and detection ability to HLB. This will go beyond the time frame of this grant, but we will pursue it to conclusion utilizing USDA, ARS funding internal.
Our initial efforts in using canines to detect canker started in 2000 but were thwarted by the September 11 attack on the US, which permanently diverted the canine we had been training (it was included in the much needed efforts to detect explosives). However, prior to this dog’s reassignment, our initial research with canines for canker detection demonstrated their ability to detect citrus and citrus canker volatiles/aromas/essences, i.e., minute concentrations of volatile/aromatic compounds, deposited on cloth and other materials as well as directly from the air. A second attempt with USDA, APHIS agriculture detector dogs was cut short by decreasing budgets combined by increasing needs of detection of contraband agricultural products at points of entry into the US. However, our third and ongoing attempt has been much more successful. This most recent canine, ‘Juice’, has demonstrated an ability to detect canker infected citrus fruit with high reliability. The initial success of this project with regards to canker detection, suggests that the same technique may be useful for detection of HLB, perhaps presymptomatic while the disease is still latent. Such a tool would be very useful and is much needed, as it would allow both detection and early removal of presymptomatic (latently infected) trees, which also may be below the titer required for psyllids transmission. Two dogs have been trained to recognize the scent of citrus canker. Dogs were trained on infected versus non-infected trees and fruit. Training of the canines for the detection of canker infected citrus trees was accomplished via a commercial canine training facility in North Florida. As indicated above, ‘Juice’, demonstrated an ability to detect canker infected citrus fruit with high reliability (> 99.3%). Juice was capable of discriminating canker infected from non-infected fruit in 30 runs of 5 blind stations each, with only one false positive and no false negatives (Peruyero and Gottwald, unpublished). Results indicate that canines can detect and differentiate citrus fruit infected with canker from non-infected tissues. The canine was then trained on infected versus non-infected Duncan grapefruit seedlings. Preliminary tests at the canine training facility are very promising. Field trials were conduced in early April at the USHRL farm. Trees (75) were placed in the field in a 3 row by 25 tree/row design. Ten replications were conducted in which the proportion of disease trees ranged from 2-10%. The conditions were sunny but with a 15-20 MPH breeze. Even so, detection highly reliable. Data are presently under analyses. Short movie footage was also taken to document the dogs search and detection behavior. More field trials are scheduled over the next few months to validate the canine’s performance. At the conclusion of the canker field testing (proof of concept) the dog will the be cross-trained for HLB detection and field tests will be conducted to assess performance and detection ability to HLB. This will go beyond the time frame of this grant, but we will pursue it to conclusion utilizing USDA, ARS funding internal.
A series of flexible stochastic models to predict the temporal increase and spatial spread of diseases. The models were initially characterized for citrus canker spreading in plantation and urban (backyard) environments. They have subsequently been extended to HLB in this project. The models can be used in a number of ways: to predict spread and to analyze the effectiveness of control strategies. Most attention has been given to spread within plantations, including allowance for proliferation of infection along boundaries in response to vector behavior. The models can readily be extended to consider spread at larger scales including spread through heterogeneous environments up to State-wide scales. We have also considered the effects of uncertainties in the distribution of host crops for example the effects of small areas of crop that may not be recorded but which can act as ‘bridges’ in transmitting disease. The effects of uncertainty in parameter estimates for dispersal parameters and transmission rates have also been included. Additional computer-friendly formulations of the models have also been developed to aid in education of stake-holders to illustrate the effects of uncertainty in predicting future disease spread and the effectiveness of alternative methods of control. Estimation of parameters for dispersal of HLB poses considerable statistical challenges, especially where trees may become infectious before they are symptomatic/detected. Here we use an SEIDR model (Susceptible, Exposed (latently infected but not yet infectious), Infectious but not yet symptomatic/detected, Detected and infectious and Removed trees). Using MCMC methods, and extensive data from Southern Gardens for successive snapshots of the occurrence of symptomatic/detected trees in known populations of susceptible trees, we are able to estimate model parameters for the transmission rates and dispersal kernel for the disease. Current work is focused on the differential effects of host age on epidemiological parameters as well as variability across the plantation. From these it is possible to allow for uncertainty in the parameters as well as variability over time and through space. The uncertainty is then incorporated in models to predict spread and to allow for uncertainty in the efficiency and comparison of control methods. A front-end, a web based version of the model that users can easily access and use is under final development and will be made available later this year. In this version both residential and commercial citrus scenarios can be tested. A wide variety of epidemiological and climate/weather variables have been included and are user selectable and changeable via sliding controls. The simulator then generates epidemics that progress through time and can be visualized on the screen. In addition various disease control/mitigation parameters can be selected and simulations can been visualized to see the effect of these various control strategies. This user selectable “front end” is highly instructive to growers and regulators. The Front end runs simulations one at a time for users to visualize. It is based on a more formal analytical model that can run thousands of simulations based on the same parameters and make more statistically valid predictions for regulatory intervention strategy building and regulatory/industry decision making.
Disease control of citrus huanglongbing (HLB) by interplanting with guava. HLB is the most devastating disease of citrus worldwide and presently threatens the existence of the citrus industry in Florida. In Vietnam guava has been shown to be an effective deterrent to HLB. For all plots and experiments, Guava trees, (Vietnamese white cultivar) were propagated and grown to appropriate size requiring about one year. Guava vs no guava nurseries: Two nursery sites, a guava protected citrus nursery versus an unprotected nursery, have been established. Disease free, PCR-negative citrus trees (2 sweet orange and 1 grapefruit cultivars) were located in the protected and unprotected plots in June 2009. The guava were established over a year ago and grown to appropriate size as indicated in Vietnam. Trees are assayed for HLB every 60 days, and are in their second assay. Psyllid populations are also being monitored continuously every two weeks within plots to document any repulsion of the vector due to guava. To date no HLB+ plants have been identified in the nursery plantings after multiple assays. Citrus/guava interplantings: 2 commercial plantings with multiple replications each have been established. This has taken considerable time. Guava trees were propagated and grown to transplant size. These were then out planted and grown for a year per Vietnam protocols. One trial was established in a commercial orchard with collaborators in Southern Gardens Citrus. A second trial planting was established at the USHRL Picos Farm in Fort Pierce. Both nursery and field citrus trees are assayed for HLB every 60 days, and have been assayed multiple times. Psyllid populations are also being monitored continuously every two weeks within interplanted plots to document any repulsion of the vector due to guava. Data collection continues and is currently ongoing. A severe frost last winter affected both the USHRL and the Southern Gardens plots causing a delay in the experiment. Damage was extensive in both plots. In the Southern garden plots, damage was more severe and the guava have now been renovated sufficiently that the Southern Gardens plot and plots have now been interplanted with citrus as of November 2009. The damage to the guava was overcome by pruning and replanting of damaged guava trees. Renovation of the USHRL plot was less extensive, and the guava were interplanted with the citrus as of August 2009 in the USHRL Picos Farm plot. A second series of freezes during the 2009/2010 winter season hit both Southern Gardens and Picos plots and both sets of plots have suffered, but both are recovering with pruning and some replanting of guava. To date, 2 HLB+ plants have been identified in the USHRL plots after multiple assays. However this data is insufficient to draw any conclusions as yet. To date no HLB+ plants have been identified in the SG plantings after multiple assays.
The control of citrus sprouts from stumps or remaining roots after tree removal has been a major concern for years within the Florida citrus industry and now more important since the arrival of huanglongbing (HLB) or citrus greening. Sprouts from stumps of clipped trees that are HLB positive can harbor the greening bacterium. This study is conducted to determine effectiveness of Remedy Ultra on the control citrus stump sprouting. This study is examining the impact of delaying application of Remedy on the control of sprout formation. Observations were made on 30-day intervals over an 8-month period after clipping or shearing the tree as the removal method. Studies are being conducted at Lake Placid and Arcadia. Treatments of Remedy Ultra were applied at 25 or 50% solution mixed with diesel fuel at time of clipping, 24, 48 and 72 hours later. At the Lake Placid site, 2, 2, 1, 1, and 1 of the 7 (100%) stumps sprouting at 30, 60, 90, 120 and 150 days after tree removal, respectively. At this site, 4 stumps within the 56 treated stumps had root sprouts but not on the treated stump. All of the root sprouts occurred on trees originally grown on rough lemon rootstock, whereas none of the other rootstocks sprouted at this site. Remedy Ultra was slightly less effective in controlling sprouts when applied at the 25% Remedy/75% diesel than at the 50% rate especially on rough lemon stumps. Delaying application using the lower rate also had a slight negative impact on sprout control. At the Arcadia site, 5 of the 7 untreated stumps sprouted at 60 days after removal with one additional stump sprouting by 90 days. At this site, 86% of the untreated stumps sprouted within 90 days of clipping. Only one stump exhibited root sprouts which occurred within 60 days after treatment regardless of the time of application of the herbicide. The number of sprouts per stump ranged from 1 to 8. Rootstock and trunk diameter have an impact on root or stump sprouting. Larger trunk diameter supports greater sprouting frequency as compared to smaller trunks. Additionally, rootstocks such as rough lemon and Cleopatra mandarin are more prone to sprouting than other rootstocks like sour orange. Historically, some herbicides used for stump treatment have caused injury to adjacent healthy trees by translocation of materials via root grafting to the healthy trees. Resets planted nearby treated citrus stumps were damaged when excessive materials were applied to the soil surface or moved off treated stumps by rain or irrigation. In these studies, no problems were associated with Remedy Ultra damaging adjacent trees.
All field equipment has been ordered and is being tested to ensure that they work as intended. The equipment to measure droplet size being developed by low volume sprayers is a continuing issue. The system developed by Drop Vision has proven to be unacceptable and has been returned to the company. The method used by Drop Vision measures droplet size on glass slides coated with telfon or MgO. Both coating methods will not work where spray mixture contains a higher volume of water and either evaporates or completely covers the slide. Other testing methods are being considered.
During the last three months our lab temporary stopped all PCR analysis because the building was being fixed. It delayed all the results from this period. Then, no progress was obtained in the three experiments. Experiment 1 – Since the last report, no new symptomatic plant were observed in any compartment. Leaf samples of all the plants were collected with the objective to detect the presence and to identify the bacterial species in each plant in November/09 (total of 1268 samples collected since the beginning of the experiment in April/08). Psyllids were collected in November and December/09 to evaluate their infectivity. The samples were submitted to conventional PCR and real-time PCR. These analyses are in progress. In December/09 we stopped to release new ACP and killed all psyllids present in the compartments. All plants were moved to other insect-proof screenhouse and will be there for symptoms observation and leaf sampling for PCR analysis. New inoculum source Citrus and Murraya plants are being prepared to repeat this experiment after March/10. At this time we want to include sources of M. paniculata infected with CLas. Experiment 2 – During January/10, ACP were reared on symptomatic inoculum sources for CLam and CLas. The emerging adults from nymphs reared on such inoculum sources were used for inoculation in March/10. After the inoculation, adults of ACP free of liberibacters will be periodically feed on these inoculated plants to detect the moment when they can acquire the bacteria from inoculated plants. Experiment 3 – Nine insect-proof screenhouses were built in a commercial citrus farm to protect Hamlin, Pera and Valencia sweet oranges with three different age. Plants from Hamlin and Valencia were already inoculated at the beginning of July/09 with infected adults of ACP and plants from Pera were not inoculated yet, because some of screenhouse suffered some wind damages. However, until now, all encaged Pera trees did not show any HLB-symptoms, proving they are really healthy. Monthly assessments for symptom severity have been done, but no HLB-symptoms were observed yet. Also, leaf samples have been collected to detect the presence of Liberibacter species on inoculated shoot. PCR analyses are in progress, no result yet.
During last three months data still have been collected from this field experiment. In average, after the symptomatic tree have been observed it was exposed in the field before be eliminated 1, 7, 29 and 67 days, respectively for frequencies of local inoculum reduction of 14, 28, 84 and 182 days. However, the shortest frequency of local inoculum reduction treatment (every 14 days) did not result in significant differences on HLB progress rate and HLB incidence 46 months after planting compared with longer frequency (every 182 days), probably because primary infections were more important and intensive than secondary ones. The tested program of ACP has been efficient to reduce the number of adult psyllids captured on yellow stick traps in 79% and to reduce the number of eggs and nymphs observed on new shoots in 94%. However this vector control reduced HLB-incidence in only 48%. The disease progress rate (estimated by Gompertz model) in plots with ACP control program was significantly reduced in 27% compared to the disease progress rate in plots without ACP control (Average Gompertz rate of 1.35 for plots without vector control and 0.99 for plots with vector control). From December/09 to March/10 the disease incidence increased from 28.9% to 35.2% in plots with ACP control program and from 61.8% to 67.1% in plots without vector control. All psyllids captured on yellow stick traps from the winter/07 to fall/09 were tested for the presence of Candidatus Liberibacter spp. by conventional PCR. PCR-positive ACP have been observed only during spring/07 (1.8% of tested samples), summer/08 (2.1%), spring/08 (4.5%), summer/09 (11.1%), and fall/09 (11.1%). Infective ACP were found both in plots with and without vector control program. Samples from spring/09 and summer/10 are being processed. Annual maps of HLB-symptomatic trees were prepared for spatial analysis using stochastic models (MCMC) to verify the effects of each treatment on primary and secondary spread of HLB. Because of the nature of the data (based on visual symptoms and not on infection) and because the long incubation period, new data set are being prepared to be analysed using only year based data from months were the new symptoms development is null. Also, to avoid the edge effect, only central area of each plot will be analysed. This analysis will be done at USDA lab in Fort Pierce by Gottwald’s team. The results of this experiment are being compared with the results of other similar experiment, but carried out in a region with lower external inoculum pressure. So far, it could be comments that HLB managent is much easier and economically sustainanble under low pressure of external inoculum and that only local disease management could be not enough to control HLB in a region with high external inoculum pressure. These and others results were presented in four semminars given in CREC, Florida Citrus Show, Syngentha Experimental Station, and Florida Natural’s in the last week of January/10. The assessments on this experiment will continue at least for one more year to allow more detailed temporal and spatial analysis and get final conclusions.
Several sensors and sensor aiding tools such as four-band sensor, five-band sensor (with GPS and self-incident light correction), volatile collection device holders (for volatile organic compound- VOC sampling), and moving sensor platform with retractable mast (for holding multiple sensors) were developed during the first year of this project. During the last quarter of this project period, significant progress was made in the application of mid-infrared spectroscopy (5.15-10.72 ‘m) for HLB detection. The healthy, nutrient-deficient, and HLB-infected leaf samples were analyzed using a portable mid-infrared spectrometer. The k-nearest neighbor (kNN)-based statistical algorithm yielded a healthy and diseased class average classification accuracy of 91% and 96%, respectively after data preprocessing (baseline correction, narrow waveband, principal component analysis-PCA). In addition, greenhouse experiments indicated that the visible-near infrared reflectance data (350-2500 nm) from a spectroradiometer could be used to identify HLB-infected plants with a classification accuracy of > 60% and > 90% for Hamlin and Valencia varieties, respectively (using quadratic discriminant analysis based algorithm). Aerial images were collected across 2,000 acres (Southern Gardens) during previous years to develop a spectral library. Forty-four critical wavebands (of 128 bands) were selected using image processing algorithm, which enhanced the average overall classification accuracy from 45% to 65% in 150 acres blocks. These bands also decreased the false positives by 60%. On December 9, 2009, a new set of georeferenced aerial hyperspectral imagery was acquired from a new citrus grove. Healthy and infected spectral data for ground measurement showed a clear separation in the two wavelength ranges: 500-800 nm and 1500-2000 nm. Studies on VOC profiling for HLB detection were performed in the greenhouse. These experiments were conducted to identify volatile biomarkers. Volatile samples were collected from two varieties (Valencia and Hamlin) of healthy and HLB-infected samples. After data pre-processing (baseline correction, computation of auto-regression coefficients), PCA and PC regression with leave-one-out validation were applied for HLB diagnosis. The corresponding systematical accuracy, false positive rate and false negative rate are: 80.7%, 10.0%, and 36.4% for Valencia; 76.0%, 16.7%, and 30.1% for Hamlin; and 70.0%, 18.8%, 45.8% for the varietal independent case. With these results, the next step will be to eliminate background variations between sampling batches and varietals. In regard to the application of fluorescence-based HLB detection, a subcontract was signed between University of Florida and Embrapa Agricultural Instrumentation to allow partnership. However, the process could not be completed due to some complications resulting from the change in rules by the Brazilian Government. Embrapa is making efforts to resolve this issue to receive funding for this project. For HLB detection, a sensor prototype PHOTONCITRUS was developed. In the laboratory, the accuracy for detecting HLB-infected citrus leaves and healthy leaves using the calibrated system increased to 95% and 80%, respectively. The sensor will be further evaluated to detect HLB-infected trees under field conditions. The work done in Brazil will be joint and partially moved to Florida to avoid funding transfer problems. In summary, progress has been made in the developing sampling protocols, establishing experimental protocols, evaluating multiple sensors under laboratory and field conditions, as well as developing data analysis algorithms and improving their performances for accurately detecting HLB-infected citrus trees. In the second year, we will further expand our research to validate the current findings on a larger dataset under field experiments and further refine the sensing techniques and classification accuracies.
Two years funding for this project was recommended by NAS panel. Significant progress was made during the first year for all 3 objectives. 1) Evaluation and refinement of the tap sample method developed by Qureshi and Stansly (2007) to monitor psyllid adults. We compared tap sample, the Yellow Corn Rootworm Trap (YCRW), the ACP Trap, and a sweep net, at low and high density psyllid populations in insecticide treated and untreated blocks of mature citrus at two locations. At low density, adult psyllids on YCRW traps and ACP traps hung within the canopies of citrus trees in a commercial grove for two weeks averaged 1.8 ‘ 0.3 and 1.04 ‘ 0.2, respectively, compared to 0.08 ‘ 0.04 per tap sample made once per tree at the end of the two week period. Averages of 9.3 ‘ 1.2, 4.8 ‘ 0.6, and 0.5 ‘ 0.1 were obtained for YCRW trap, ACP trap, and tap sample, respectively, at the high density. One sticky trap cost approximately $1, only provides data after at least a 1-week delay, and requires an average 7 minutes to deploy and collect data compared to an average of 3 seconds for a tap sample that provides instant data with no more supplies than a laminated sheet of letter size paper or a smooth white surface such as a clipboard, and a one-foot piece of PVC tube. Therefore, many tap samples can be taken in the time one sticky trap can be processed and cost much less. Comparison of the two sampling methods in mature citrus at Southwest Florida Research and Extension Center showed that significantly more adults were detected using the tap sample compared to the sweep net at low density, with means of 0.1 ‘ 0.02 and 0.06 ‘ 0.01 per tap or sweep sample per tree, respectively. Means of 0.3 ‘ 0.1 per tap and 0.4 ‘ 0.1 per sweep per tree at high density were not significantly different. Time to conduct one sweep (3.1 ‘ 0.04 s) was significantly greater than one tap sample (2.9 ‘ 0.04 s) at low density but did not differ at high density and averaged 3.4 ‘ 0.1 s. However, sweep samples are more tiring, require more equipment, knock off fruit and can spread canker, especially when foliage is wet. 2) Test the influence of adult density and shoot infestation rates on precision of estimated means and distribution of population within blocks. Regression analysis based on bootstrapping of the data from the first trial indicated that the precision targets of 0.25 and 0.01 SEM:mean were reached at high psyllid density with 8 and 66 tap-samples, respectively, compared to 12 and 75 YCRW traps and 13 and 74 ACP traps. At low psyllid density, 267 and 1929 tap-samples, 16 and 130 YCRW traps, and 43 and 219 ACP traps were needed to attain these precision ratios, respectively. In the second trial at low psyllid density, 98 and 554 tap-samples and 158 and 1281 sweep net samples were needed at ratios of 0.25 and 0.1, respectively, compared to 35 and 196 tap samples and 37 and 198 sweep net samples at high density. Therefore, the sweep net would take more time to provide the same data and more work and difficulty in counting at high density. We are conducting more sampling and integrating shoot density and infestation data into overall sampling for psyllid. 3) Evaluate and integrate methods for assessing psyllid density, shoot density, and infestation rates into a user friendly system accessible to consultants and managers. A rapid system that includes tap samples, flush inspections and natural enemy evaluations along with forms and spreadsheets is available on our website swfrec.ifas.ufl.edu/entlab/. We made 5 presentations specifically on monitoring psyllids and conducted 5 workshops at SWFREC with a total of over 250 attendants. The tap sampling method has gained wide acceptance and is adopted by growers and consultants in SW Florida and our cooperators in California and Mexico. A recent survey of grove managers by SWFREC and IFAS extension covered 106,148 acres of commercial citrus and reported that all 95% conducting sampling use the tap method, along with shoot examination (76%), sticky traps (14%), and sweep nets (9%). Findings from the above studies have been submitted for publication in citrus industry and Florida Entomologist.
All objectives for the first year of this project have been completed or exceeded. Objective 1: Compare aerial and conventional ground application of insecticides for psyllid control. Two trials were conducted for this objective. Low-volume (LV) aerial applications at 10 GPA were compared with ground airblast applications at125 GPA during the summers of 2008 and 2009. Results in 2008 indicated that broad spectrum insecticides work well by air and ground but that the selective insecticides, spinetoram and imidacloprid, provided much better control when applied by ground. During 2009, we followed up with two selective insecticides (spinetoram and spirotetramat) compared with the broad spectrum phosmet, all applied at the same volumes as above by air and ground in replicated (N=4) 48 or 24-acre plots respectively of mature orange trees. In this second trial, all materials were more effective by ground, however differences between broad spectrum and selective insecticides were less clear, phosmet providing best control with both methods. Nevertheless, the ‘Cooperative Dormant Spray’ program we initiated with Gulf Citrus Growers Association resulting in over 70,000 total acres sprayed by air the last two winters demonstrates that aerial applications are efficient and effective during the “dormant” season (see report on FCPRAC #000). A second year of ‘Cooperative Dormant Sprays’ is presently being evaluated . We also compared LV applications using a modified London Fogger 18-20 by Chemical Containers @ 2GPA applied to bed tops only with conventional airblast sprayer that treated both tops and swales @ 116GPA on a highly infested plot of ‘Pineapple’ oranges. Applying only tops weighted the odds against the LV, but the objective was to test this typical LV application. Spinetoram (4oz/ac) and dimethoate (24oz/ac.) were both applied with 2GPA of horticultural mineral oil (HMO). Conventional ground applications resulted in fewer ACP for the duration of the one-month trial compared with the control, whereas LV applications resulted in statistically fewer ACP for only the last two weeks. So again, conventional application proved superior in terms of control with the products and under the conditions tested. Objective 2: Assess the effects of frequent nocturnal LV applications of horticultural mineral oil (HMO) on psyllid populations. In a preliminary replicated trial testing LV @1GPA of only HMO every two weeks for 4 months in summer of 2008, we found 14.1’3.8% of the flush was infested in untreated plots compared with 2.1’1.2% for the HMO-treated plots. In 2009, we compared the modified London Fogger and the Proptec P400D @ 2GPA, applied every 2 to 4 weeks depending on ACP populations, and monitored ACP every two weeks. High frequency LV applications of HMO with the Proptec maintained lower populations (1.4 ‘ 0.75 ACP adults x days) compared with the London Fogger (4.4 ‘ 1.9), and the untreated control (6.3 ‘ 3.9). In 2010, we are conducting experiments to evaluate frequent LV applications of HMO alone or tank mixed with micronutrients as a complete ACP/HLB management package. We are also collaborating with CREC to assess the deposition of oil on citrus foliage by LV application with protocols we are developing using gas chromatography to quantify residues eluted from leaves of treated trees. Objective 3: Extend results to the citrus industry. Results from these experiments have been presented at 18 extension meetings across the state in 2009 including two production managers meetings and two CCA trainings, the Entomological Society of America (ESA), and the Florida State Horticultural Society (FSHS). – Stansly, PA, HA Arevalo, M Zekri, and R Hamel. 2009. Cooperative dormant spray program against Asian citrus psyllid in SW Florida. Citrus Industry 90(10):14-15 – Arevalo, H. A. and P. A. Stansly. 2009. Comparison of Ground and Aerial Applications for Control of Adult Asian Citrus Psyllid, Diaphorina citri Kuwayama. Proc. Fla. State Hort. Soc. 122: (186-188)
(863) 956-8817
(863) 956-5894
700 Experiment Station Road
Lake Alfred, FL 33850
Email Us
