A replicated field study was conducted in a 5.4-ha commercial block of young ‘Valencia’ orange trees through four harvests employing a factorial design to evaluate individual and compound effects of a popular foliar nutrient program and threshold-based vector management. ACP populations were maintained at contrasting levels in insecticide-treated and untreated plots despite proximity. Nevertheless, incidence of HLB, estimated by PCR at nearly 30% at the beginning of the study, rose to almost 95% early in the third year without measurable reduction from vector control. However, insecticide treatments did result in higher threshold cycle (Ct) values, indicating reduced disease intensity. Vector control significantly improved yields all but year one, while the nutrition only treatment made a significant contribution to yield only in year four. Combined foliar nutrition and vector control was the best treatment all 4years, and provided yields close to the pre-HLB regional average in the 4th year. Although the combined treatment was not profitable at current juice prices, this shortcoming could be remedied by reducing costs of the nutrient package and/or the insecticide regime. In another location where HLB incidence and tree health was monitored on grapefruit and Hamlins receiving various treatments, including initially, tree removal, S. Halbert has been conducting trapping of psyllids. Psyllids from the traps are being analyzed for HLB titer by K. Hendricks, SWFREC. Four suction traps were operated at the SW Florida Research & Extension Center from July 2009 to present. These included an 8 meter tall trap and three 2 meter traps. Of the latter, one was in managed citrus, one was in unsprayed citrus, and the other was in an open field. Samples were collected approximately weekly. The psyllids were removed and identified in Gainesville. Beginning in 2011, all Diaphorina citri Kuwayama were tested singly for presence or absence of the HLB pathogen. All three short traps collected D. citri. Both traps located in citrus collected at least occasional D. citri throughout the year, but the trap in the unsprayed citrus collected the most. The trap in the open field showed peak activity in March, coinciding with the spring flush. These collections could indicate that longer distance flights away from the crop occur at that time of the year. Overall, there were few psyllids positive for Las. There were positive samples collected from all three short traps. There was no difference in the numbers of positives by trap. This can be attributed to the fact that citrus greening disease is widespread and common in the Immokalee area. Data indicated that neither nutritional nor insecticidal sprays impacted the disease progress of HLB, because either the treatments were initiated during the long lag time between inoculation/symptom expression or another reason. Recent yield data indicates that trees in plot receiving nutritionals and insecticide are benefiting by increased yields.
Data collection and analysis of field trials with ‘Ray Ruby’ grapefruit treated with foliar sprays with copper bactericides or soil drench or trunk application with the non-insecticidal SAR-inducer acibenzolar-S-methyl (Actigard) and the neonicotinoids imidacloprid (Admire Pro) and thiamethoxam (Platinum) to evaluate their efficacy for canker control on young bearing trees has been completed. Data analysis for the Hamlin orange trial is still pending. For evaluation of contact activity, nine film forming coppers treatments sprayed every 21 days with a handgun were compared with two rates of the soluble copper formulation, Magna-Bon (copper sulfate pentahydrate) or was applied as a combination of one film forming copper with Magna-Bon at lower rates. In spring 2012, early season rains were sporadic, hence incidence of old canker lesions on fruit was 13% on the non-treated control (NTC) or half of the incidence of newer late season lesions on the NTC (27%). All the copper treatments were effective for reducing lesion incidence by more than 50% (6% old sand 10% new lesions). Magna-Bon was as effective as the film-forming copper formulations alone, mixed with film-forming copper, or applied later in the season after the filming forming copper early in the season. Comparable cankercontrol achieved with the combinations of Magna-Bon substantially reduced total seasonal application of metallic copper. Four 60-day interval soil drench and trunk applications of Actigard, Admire and Platinum reduced canker incidence on fruit from 47% to 20% compared to 16% with 21-day interval foliar sprays season-long with copper hydroxide (Kocide 3000).
Objective 1: Preliminary trials conducted with potted nursery and young field trees indicate that soil drench is an effective and consistent application method for increasing copper status of young trees over the baseline concentration in leaves. Hence, in a trial in Picos Farm at USDA-Ft. Pierce, Hamlin on Swingle citrumelo trees were root drenched in late spring with the copper bactericides EXP, Cop-R-Quik, (CQ) Magna-Bon (MB), and copper phosphite (CP) and a zinc phosphite (ZP). Spring flush leaves were collected in July to measure copper and zinc concentration in leaves to test for systemic movement of the metals. Overspray of micronutrients obscured detection of treatment differences. Leaf samples for PCR testing were also collected to track Las infection progress. So far incidence of infected trees is less than 1%. To further evaluate soil drenches of the two commercial bactericides containing copper chelate (CQ and MB), three locations were selected. Two trials are in blocks with 1 year old trees that are apparently healthy and a third trial in a block of 4-yr old Valencia trees with symptomatic and pre-symptomatic HLB infection as determined by PCR status. Hence, the preventative and curative activity of the copper bactericides are being followed in pre- and early stage HLB infected trees. In the two trials with 1 yr-old young trees, a wide range of copper concentrations in flush leaves has been established. In the 4 yr old block, no effect of MB or CQ on HLB decline rating was detected in December 2012. Samples for PCR status of the trees in all trials will be collected and analyzed this spring.
Florida growers have reported that enhanced nutritional programs (ENPs) maintain productivity of HLB-infected trees. However, efficacy and sustainability of the nutritional approach for HLB disease management remains uncertain. Complementary studies of multiple ENPs and their individual components compared to the standard nutritional program (SNP) on nursery and field trees were initiated in 2010. Two independent nursery trials were initiated with final data collection finished and data analysis currently underway. From monitoring temporal bacterial movement and replication it was found that Candidatus Liberibacter asiaticus (Las) populations are similar for ENPs and the SNP. Minor differences in Las movement have been observed. Las invaded new flush tissue faster in ENP treated trees than SNP trees. Phosphite treatments have caused Las to favor early invasion of root tissue compared to other treatments. Preliminary analysis of the final takedown results from nursery trials shows a reduction in visual symptom expression with ENPs. However, there were no differences in the HLB-associated phloem plugging or reduction in leaf number, leaf size, and fibrous root mass. The field site was chosen for its mix of healthy, presymptomatic, and HLB symptomatic trees to determine if observed differences resulted from effects on healthy or infected trees. After 3 years of nutritional treatments, fruit yield and HLB symptoms in field trees treated with ENPs have not differed significantly from the standard nutritional program. The only significant predictor of yield differences was the percent of declining trees, but there was no interaction with nutritional treatment, suggesting that nutritional programs have not affected yields of healthy or HLB affected trees in this study.
Entomopathogenic nematodes: Amending soils to increase biological control of insect pests Conducted duplicate experiments to investigate effects of soils from ACPS and conventional plots on nematophagy by fungi and subsequent entomopathogenic nematode (EPN) efficacy against D. abbreviatus larvae. Soils were oven desiccated prior to use. Trapping nematophagous fungi (NF) or zoosporous NF were added to soils with either Steinernema diaprepesi (Sd) or Heterorhabditis indica (Hi). After 10 days EPNs were extracted from half experimental units and weevil larvae were exposed to soil from the other units. ACPS soil reduced survival of both EPN species regardless of NF presence. Effect of soil very pronounced on Sd, relatively mild on Hi. Effect of NF pronounced on Sd, with additive effects of soil and trappers and significant synergism between soil and zoosporous NF. Effects of NF on Hi relatively minor with no indication of synergism. Previous study of soils under both management regimes noted increased incidence of a bacterial parasite of Sd in ACPS soils. The bacterium was not present in current experiments, thereby extending potential mechanisms for ACPS effects on EPNs to encompass biocontrol by bacteria and fungi as well as direct inhibition of ACPS to EPNs, especially Sd. Plant parasitic Nematodes: Characterizing a new nematode pest and the prevalence of resistance breaking populations of the citrus nematode. PCR primers and taqman probes designed (from published sequences) to distinguish Xiphinema citricolum from X. laevistriatum (two dagger nematodes that infest citrus orchards) conflicted with our morphological characterization of several populations of dagger nematodes. The results indicate the need to re-evaluate the molecular and morphological characterization of dagger nematodes from type localities. This work has been initiated using specimens isolated from localities given in the original report (2 localities for each species). Continued survey of sites identified by DuPont soil sampling program as positive for dagger and/or citrus nematode coincident with tree decline.
Progress on development of an acoustic trap to capture male Asian citrus psyllids (ACP) that are searching for females on citrus tree branches continued on two fronts in the last quarter of 2012. In the third quarter, we had recorded male and female communicatory duets and found that both sexes responded readily to playbacks of recorded signals and to synthetic mimics that had the same temporal and spectral patterns as the recorded calls. The playbacks were produced by a computer that output signals to a commercially available (Bruel and Kjaer) vibration exciter attached to the plant. Both the recorded calls and synthetic mimics attracted males to the signal source, which is a necessary first step in the trapping process. This work was prepared and submitted for publication. Based on the successful results of these studies, we began efforts to develop a less expensive system for detecting and producing the ACP communicatory signals. We purchased an inexpensive Arduino Uno microcomputer system and programmed it to output synthetic mimics of ACP communicatory signals to a piezoelectric buzzer. In addition, we have begun the process of programming the microcomputer to recognize and reply to calls made by males searching for females on a tree. We found that it is possible to use one single buzzer that both produces a female signal and detects the male signal on the plant, which reduces the total cost of the Arduino-buzzer system to less than $50. In comparison, the cost of the signal analysis/vibration exciter system that was used to conduct the initial studies is about $3000. We expect to complete the programming tasks and begin testing the new instrument in the laboratory during the next quarter.
The overall goal of this 3-year research project is to efficiently deliver antimicrobial molecules against HLB bacteria into citrus phloem. This quarterly (Oct.2012 to Jan. 2013) research continued to optimize the W/O (water in oil) and O/W (oil in water) nanoemulsion formulations by evaluating the ideal adjuvant mixtures of solvents (ethanol, acetone, methyl, acetate, and ethyl acetate), oils (soy oil, cremorohor, carvacrol, p-cymemen, orchex796), lipophilic (span80, span85) and hydrophilic surfactants (tween20, tween80) using pseudo-ternary phase diagram and thermodynamic stability methods. Several ideal W/O and O/W nanoemulsion formulations have been optimized for loading the effective antimicrobial compounds. Orchex796, one of the horticultural mineral oils, has been used to control insects in citrus production. The nanoemulsion formulations of Orchex796 will be applied to control both citrus insects and the Las bacteria. Eight compounds from Citrus Research and Education Center, IFAS, UF were tested for their effectiveness against HLB bacteria by the grafted-based chemotherapy method. The results will be showed in next quarterly report. In the next quarter, the research will focus on screening the penetration compounds finalize the applicable formulations.
Ultra High Performance Liquid Chromatography – Pesticide Residue Analysis (Dec 2012 Update) This quarter, use of the equipment purchased focused on validating methods used for analysis of pesticides that we will be working with. We will continue this method development and will begin analyzing research samples in Jan/Feb.
Using highly purified salivary sheaths we were able to study composition of different parts of the sheaths to show that their composition is not homogeneous throughout. This finding has allowed targeted research focused on blocking the initial stages of sheath synthesis as a new strategy for blocking the psyllids ability to transmit the bacterium associated with citrus greening. Research has now been initiated on developing transgenic plants to express genes encoding peptides that could function to block the ability of the psyllid to reach the phloem sieve tube elements with their piercing mouth parts. Because the sieve tube element is the location where the citrus greening associated bacterium resides, it is hoped that by blocking the psyllids form reaching these cells we will block the ability of the psyllid to acquire or transmit the bacterium, thereby blocking the spread of citrus greening disease.
Progress continues on the primary objectives of the project:1) Develop a generic epidemiological model that can be used to compare control scenarios and to optimize the probability of controlling and managing high-risk pathogens of agricultural significance. 2) Development and testing of efficient methods of statistical inference to estimate epidemiological parameters from maps of emerging epidemics. 3) Develop user friendly model ‘front ends’ that can be used by researchers and regulatory agencies. During this cycle we have focused primarily on making the models more robust and flexible models to most efficiently analyze and predict the spread of emerging pathogens across a range of scales extending from within-plantation to the landscape and regional scales for both HLB and canker. We are continuing to test the models extensively to assess how host planting age affects the transmission of HLB . We consider the models to be and are developed as part of a flexible tool-box so that they can be readily adapted to new disease threats as well. Work has continued on parameter estimation and writing up the methodology. Removal or treatment of infected trees in a region in which the dispersal and transmission are being estimated for an emerging epidemic can have profound effects in under-estimating some parameter values if not allowed for. Accordingly we have also developed methods to allow for this during parameter estimation. This requires an understanding of how the control is likely to affect epidemiological parameters and where this is unknown how to compare alternative models. Once parameterized, the models allow us to predict where disease is likely to spread most rapidly. We have tested a risk-based strategy that uses epidemiological knowledge on where disease is likely to spread most rapidly to optimize control that involves removal of more susceptible hosts around key infected sites. Using a parameterized model of citrus canker on realistic landscapes, we have shown that the risk-based strategy would outperform a conventional fixed-radius approach when the dispersal and transmission parameters are accurately known. We are testing to what extent performance degrades under varying degrees of uncertainty. Additional work has also been completed in this quarter in collaboration with a graduate student on a model of host-vector systems, considering the spread of HLB within an individual tree in particular. The relative importance of pathogen transmission within the vascular system of the tree and between leaves via psyllid vectors has been studied. Currently, the models are being used to investigate the efficacy of potential methods for the control of HLB, including the use of insecticide, roguing, use of nutritional products and thermotherapy, amongst others. Publications: Cunniffe, N.J., R.O.J.H. Stutt, R.E .DeSimone, T. R. Gottwald, and C.A. Gilligan. 2012. Webidemics: Webbased Interactive Demonstration of Epidemiological Modelling Informing Control Strategies. To be submitted to PLoS One. Gilligan, C. A., F. M. Neri, A.R.. Cook, G. J. Gibson, and T. R. Gottwald. 2012. Bayesian analysis of an emerging epidemic: citrus canker in urban Miami. To be submitted to Journal of the Royal Society Interface. Parry, M.F., G. J. Gibson, T. R. Gottwald, S. Parnell, and C. A. Gilligan. 2012. Spatio-temporal modelling of the spread of HLB/Citrus Greening on orchard scales. To be submitted to Proceedings National Academy Science USA / Journal of the Royal Society Interface.
Xambr’ Brazil Plots: Designed to examine the effect of windbreaks, copper sprays to reduce infection, and leafminer treatments to determine their individual and combined effects on control of citrus canker. Windbreaks were completed and plants were be established in Mid April 2010, but severe winds damaged the windbreaks during two storm events. These windbreaks have been reinforced and rebuilt. From October through December 2012, we completed plot staking and planting of a short line of Casuarina in front of the gates in each block, 7 m outward from the main casuriana perimeter, to allow equipment entry but still adequately break the wind from the interior of the plots. Windbreak and citrus plants will be of sufficient size to begin the experiment in March 2013. New windbreak trials established in Saint Lucie County. Further disease data has been recorded and is awaiting analysis. An additional 10 weather stations were purchased for deployment at a second site (cv. Rubyred grapefruit, Scott Farms, St. Lucie Co.). Weather stations were deployed equidistantly in an East-West and North-South orientation in the citrus block. Data collection has been started. Programmable leaf wetness controller (PLWC): During this quarter we developed a very complex sensor that probably closest emulates a leaf. This sensor, currently named the pin sensor, is a flat surface, made from wax. Contact wires are then connected together in a pattern, so that a drop of water will bridge a ‘+’ to a ‘-‘, creating a connection. The more drops of water, the more connections; the more connections, the more current flow. Currently the sensor is too sensitive and thus we are currently reducing the sensitivity of the electronics, and retesting. Project publications: Bock, C. H., Graham, J. H., Gottwald, T. R., Cook, A. Z., and Parker, P. E. 2010. Wind speed effects on the quantity of Xanthomonas citri subsp. citri dispersed downwind from canopies of grapefruit tree infected with citrus canker. Plant Di Bock C.H., Graham, J.H., Gottwald, T.R., Cook, A.Z., and Parker, P.E. 2010. Wind speed and wind-associated leaf injury affect severity of citrus canker on Swingle citrumelo. Eur J. Plant Path 128:21-38 Bock, CH, Parker, PE, Cook, AZ, Graham, JH and Gottwald, TR. 2001. Infection and decontamination of citrus canker and inoculated the surfaces. Crop Protection 30:259-264. Hall, D.G., Gottwald, T.R. and C.H. Bock. 2010. Exacerbation of citrus canker by citrus leafminer Phyllocnistis citrella in Florida. Florida Entomologist. Florida Entomologist 93:558-566. Bock, C.H., Gottwald, T.R. and Parker, P.E. 2011. Distribution of canker lesions on the surface of diseased grapefruit. Plant Pathology (Accepted). Bock, C.H., Cook, A.Z., Parker, P.E., Gottwald, T.R., and Graham, J.H. 2011. Some characteristics of the dispersal plume of bacteria of Xanthomonas citri subsp citri in wind-driven splash downwind of canker-infected grapefruit tree canopies. (Plant Disease). Bock C.H., Graham J.H., Cook A. Z., Parker P.E., and Gottwald T.R. 2012. Predisposition of citrus foliage to infection with Xanthomonas citri subsp. citri. (Submitted to Plant Disease).
During summer/fall 2012, we completed a gap study trial in St. Lucie Co. We started 2 replicated trials to study efficacy of mating disruption in plots treated with pheromone in rubber septa and 2 validation trials with pheromone loaded in rubber septa in collaboration with ISCA, grove managers, owners, and crop consultants. These trials address pheromone carrier, timing, coverage, formulation, longevity, and efficacy of mating disruption under various crop phenologies. 1. Winter and spring application of SPLAT CLM. We submitted the manuscript ‘Seasonal abundance of Phyllocnistis citrella (Lepidoptera: Gracillariidae) and dormant season application of pheromone in citrus to control overwintering and spring populations’ (Fla Entomol). Winter application of SPLAT CLM did not increase trap catch disruption in spring compared with spring application alone. Moth numbers declined in winter, suggesting that preventive control might help manage this pest. Leaf flushes on south sides of groves may provide favorable habitat during winter. Moth flight in spring preceded mining, suggesting early colonizers arrived from outside the grove. Area-wide management of this pest may be appropriate. 2. Effect of formulation on longevity and efficacy across treatment gaps (St. Lucie Co.). Cooperator: Packers of Indian River. Established 27 April to compare current SPLAT CLM triene pheromone formulation with a formulation that incorporates the ‘natural’ 3:1 triene:diene pheromone blend. Treatments were applied 29 May and 3 June. Mines were evaluated 18 June and 26 July when widespread early-stage leaf mines were present. Data are being analyzed. There appears to be no difference in efficacy between the formulations. 3. SPLAT CLM versus rubber release devices (St. Lucie Co.). Cooperator: Scott Lambeth (Golden River Fruit Co.) An experiment was started 10-13 July to compare trap catch disruption in small plots (0.14 ha) treated with SPLAT versus rubber septa loaded with an equivalent amount of pheromone. Treatments included ‘natural’ 3:1 blend and triene only. In a companion trial, five blocks (71 ha) were treated with SPLAT CLM (10-12 July). In small plots, septa provided superior and long-lasting trap catch disruption. Disruption in SPLAT-treated large blocks was weak and short-lived. We identified a formulation problem in SPLAT. Three of the large blocks (42 ha) were treated (24 Aug) with rubber septa distributed at 318 septa/ha. Septa containing either formulation have provided >95% disruption for 11 weeks in small plots and 5 weeks (so far) in large blocks. Evaluation of leafminer damage is ongoing. 4. Large plot septa trial (St. Lucie Co.). Cooperator: Packers of Indian River. On 7 Sept, an experiment was started to compare trap catch disruption in plots (0.87 ha) treated with rubber septa containing ‘natural’ 3:1 blend (330 septa/ha) versus untreated plots (0.87 ha). 5. Large plot septa trial (Charlotte Co.). Cooperator: TRB Groves. On 20-23 Aug, previous SPLAT CLM grower validation trial was continued with deployment of rubber septa loaded with natural blend. 6. SPLAT grower validation trials. Grower validation trials with SPLAT are complete at 4 sites (Charlotte, DeSoto, Okeechobee and St. Lucie). SPLAT CLM underperformed compared with previous studies due to an apparent interaction between pheromone and a new wax component.
We proposed to identify and assess gene sequences for their negative effects on sap-sucking Hemipteran insects via RNAi using both in vitro and in planta dsRNA feeding assays. To date, we have cloned sequences from nine homologous D. citri and M. persicae transcripts. In addition, we have carried out artificial feeding assays on M. persicae using dsRNA derived from the salivary gland-specific Coo2, midgut-specific glutathione-S-transferase S1 (GSTS1) and constitutively expressed S4e ribosomal protein from M. persicae, as well as a control derived from green fluorescent protein (GFP) sequence. Since recent evidence suggests that RNAi in sap-sucking insects may operate more effectively in planta than in vitro, we evaluated the RNAi strategy in planta for its effects against our model insect, M. persicae (objective 2). In this objective, Gateway-based vectors were used to express the selected insect dsRNA (Coo2, GSTS1 and S4e) either constitutively (35S promoter) or in a phloem-specific manner. Our results suggest that the M. persicae-specific dsRNA expressed in planta has a negative effect on both the lifespan of the insects and the number of offspring generated. In the fall of 2010, we began working on objective 3: to transform citrus with RNAi-inducing transgenes against D. citri. Previously, we conducted 3′ rapid amplification of cDNA from vacuolar ATP synthase subunit G, S4e, and .-tubulin transcripts from D. citri. We have now inserted sequences of the aforementioned transcripts into Gateway-based vectors downstream of both the constitutive 35S and our novel phloem-specific citrus CsSUS1 promoters. To date, we are in the process of transforming and regenerating citrus with the D. citri-specific gateway vectors for evaluation and use by the Florida citrus industry. Initial attempts to transform and regenerate Citrus sinensis ‘Valencia’ and ‘Hamlin’ containing reporter gene constructs were successful. Currently we have completed transformation of citrus callous tissue using gateway vectors with the vacuolar ATP synthase subunit G or S4e transcripts inserted downstream of a phloem-specific citrus CsSUS1 promoter. We have regenerated at least one line for each of the constructs of interest, and we are in the process of in vitro micro propagation to produce additional copies of each line. We are still regenerating other transformed lines, and are preparing to generate additional lines with the other transcript/promoter combinations.
The questions to be addressed by this research project are 1) how does Imidicloprid move in the sandy soils of south Florida, 2) how long does imidicloprid persist in sandy soils, and 3) how much Imidicloprid leaches below the root zone of citrus trees. Soil sample analysis and field shylid populations counts have been completed for all studies except these ending in the past quarter. Two field studies were conducted the spring and summer of 2012. Initial analysis of soil samples from the spring study indicate that Imidicloprid concentrations decreased rapidly in the soil and was either taken up by the plants or leached within 5 to 10 days at recommended microsprinkler irrigation rates, however reduced irrigation following applications allowed for greater soil residual time and uptake. During these initial application studies, adult psyllid populations per tree decreased from approximately 1.7 for both treated and non-treated trees to 0.1 psyllids per tree for the treated trees and remained suppressed greater than 50 days after application. These data would indicate great persistence of the insecticide in citrus tissue and is supported by elevated leaf tissue concentrations. Isotherms studies to determine the sorption rate and persistent concentration in soil with time was conducted. The sorption studies, relating Imidicloprid concentrations in soil solutions with Imidicloprid concentrations in the solid phase of sandy soil were conducted using Immokalee fine sand. Imidicloprid decreased in soil concentrations due to biodegradation at a rate of 0.013 .g g-1 d-1. Imidicloprid was found to penetrate to a depth of 45 cm and have greatest affinity in the 0-15 cm soil depth due to higher soil organic matter at that depth. These data were supported by soil partition coefficients (Kd) of 1.68 mg of Imidicloprid. in one liter of soil solution at the 0-15 cm depth and decreases to 0.33 and 0.25 mg l-1 at 15-30 and 30-45 cm, respectively. These data suggest that Imidicloprid is leached rapidly from the soil if not taken up by the target plant.
Progress continues on the primary objectives of the project:1) Develop a generic epidemiological model that can be used to compare control scenarios and to optimize the probability of controlling and managing high-risk pathogens of agricultural significance. 2) Development and testing of efficient methods of statistical inference to estimate epidemiological parameters from maps of emerging epidemics. 3) Develop user friendly model ‘front ends’ that can be used by researchers and regulatory agencies. During this cycle we have focused primarily on making the models more robust and flexible models to most effeciently analyze and predict the spread of emerging pathogens across a range of scales extending from within-plantation to the landscape and regional scales for both HLB and canker. We are continuing to test the models extensively to assess how host planting age affects the transmission of HLB . We consider the models to be and are developed as part of a flexible tool-box so that they can be readily adapted to new disease threats as well. The principal modeling results for this quarter relate to further testing and refinement of a tool box to estimate parameters for dispersal and transmission using citrus canker and HLB. We now have a set of rigorous Markov-chain Monte Carlo (MCMC) methods to estimate parameters for dispersal and transmission rates of emerging epidemics from successive mapped snapshots of disease. Specifically, we have developed data augmentation to infer the sequence of infections between successive snapshots of disease and used reversible jump MCMC methods to compare alternative models, for example when it is unclear how important a particular biological process is likely to be. MCMC methods were combined with techniques from time series analysis (such as sliding windows estimates) in order to track the value of time-varying parameters such as the transmission rate. We also developed and applied statistical methods to test the goodness-of-fit of the model using inferred parameters against the observed temporal and spatial patterns. We believe this to be the most rigorous analysis of how to fit epidemiological models to emerging epidemics. We have also made significant progress in the implementation of Approximate Bayesian Computation (ABC) algorithms for stochastic, spatially explicit epidemics as an alternative method suited to some epidemics that if successful is likely to be easier to use by non-specialists. We have successfully tested our algorithms on synthetic datasets and small real datasets (including citrus canker data). Work is currently under way to test the robustness of the method, and to apply it to large epidemic datasets. Publications: Cunniffe, N.J., R.O.J.H. Stutt, R.E .DeSimone, T. R. Gottwald, and C.A. Gilligan. 2012. Webidemics: Webbased Interactive Demonstration of Epidemiological Modelling Informing Control Strategies. To be submitted to PLoS One. Gilligan, C. A., F. M. Neri, A.R.. Cook, G. J. Gibson, and T. R. Gottwald. 2012. Bayesian analysis of an emerging epidemic: citrus canker in urban Miami. To be submitted to Journal of the Royal Society Interface. Parry, M.F., G. J. Gibson, T. R. Gottwald, S. Parnell, and C. A. Gilligan. 2012. Spatio-temporal modelling of the spread of HLB/Citrus Greening on orchard scales. To be submitted to Proceedings National Academy Science USA / Journal of the Royal Society Interface.
(863) 956-8817
(863) 956-5894
700 Experiment Station Road
Lake Alfred, FL 33850
Email Us
