Work proceeded on several fronts. First, updated budgets that reflect the increase in grove maintenance costs were completed. These budgets indicate that a full program to minimize the effects of greening can increase grove care-taking costs by $450 per acre. Under typical Florida yields, this increase adds more than one dollar per box to the breakeven price. Second, plans are being formulated to hold an “economic summit” to held in April at the CREC. The participants will include economists from both Florida and Sao Paulo. During the last half-day of the meeting, representatives from the citrus industry are invited to listen and respond to the discussion. Third, a model of disease spread through a particular grove that was first developed in Sao Paulo by FUNDECITRUS is being adapted to Florida.
We want to obtain an in vitro culture of Candidatus Liberibacter by co-culturing the bacteria with insect cells to study Candidatus Liberibacter physiology, metabolism, virulence and its interactions with the insect vector. We tested various stable insect cell lines to in vitro culture Candidatus Liberibacter asiaticus (LAS), the Asian form of HLB disease, also found in Florida. We tested different inoculums (from various citrus and periwinkle plants) on different insect cell lines. We checked for the presence of LAS in inoculated cell cultures by direct PCR. Current results: – After Mamestra, Spodoptera, Drosophila cell lines, we tested an Aedes albopictus insect cell line. We didn’t detect LAS presence after inoculation. Actually, citrus inoculums have a deleterious effect on Aedes cells and we are now focusing on periwinkles inoculums. – We detected LAS in two lines of drosophila cell cultures by direct PCR. One line lost the detection after 6 transfers. Signal of detection was confirmed to be LAS. For one line, we are currently reaching the 15th culture transfer with a detection by PCR to be confirmed by sequencing. The detection remains very weak and we need to improve the conditions to get a higher titer. – The detection of the bacteria by direct PCR peaks at day 7 after transfer and then declines. The bacteria are multiplying and seem to follow the drosophila cells growth. We began to set up multiplex qPCR assays. Primer sets suitable for qPCR and specific of LAS and drosophila detection were selected and tested with success. – In order to reach higher bacterial concentrations, we tested complements of the culture medium: various sugars, vitamins or trace elements. We analyzed metabolic pathways potentially encoded by the released Liberibacter genome sequences to improve growth conditions and to define limiting factors and/or growth inhibitors. Of the complements tested some had a positive effect on the detection of LAS but negative effects on insect cells survival. With the decline of the insect cell cultures the bacteria was no longer detected. Currently, sodium pyruvate and sterile citrus juice had a positive effect on LAS detection but a lethal effect on insect cells. Proline and fructose had a positive effect on the bacteria detection and are now added to our co-cultures. – We analyzed insect cell culture medium sugar and amino-acids variations over culture time to identify potential LAS growth limiting factors. Rapid depletion of glucose and of some amino-acids by drosophila cells could be limiting factors and are candidates for our medium improvement. Current work: – We are now setting up multiplex qPCR assays to monitor the ratio of LAS and drosophila DNA over culture time. – We are combining complements to the insect cell culture media and looking for new ones to improve LAS concentration – We are analyzing insect cell culture medium minerals and trace elements variations over culture time to identify potential LAS growth limiting or improving factors. – We are testing new insect cell lines to get more consistent presence of the bacteria and higher concentrations. We are reaching our milestones for the first year of this project and we will look into the axenization of our primo-cultures.
The citrus extension agents conducted the “Low Volume Application Technology for Citrus Pests” program in six locations around central and south Florida in October. These programs discussed: low volume labeled materials; low volume application; application considerations; and worker safety. Program attendance exceeded 300 participants. Full program presentations can be viewed by going to the Citrus Agents web site at http://citrusagents.ifas.ufl.edu/events/PDF/LowVolumeApplication/LVA_program.htm . As part of greening control, selected agents are actively involved in assisting with the development of area wide spray events. These targeted sprays are to suppress psyllids over a wider geographic area. In addition to the above program, agents are actively involved in conducting grower programs (6), field visits (47), educating Master Gardeners (2) and the general public in greening identification and control issues.
With the program being recently funded, we are in the process of receiving order items to conduct the study. All materials were ordered in December of 2009 and are expected to be received within the next 30 days. Once delivered, studies will be designed to conduct proposed studies.
Six months after shearing or clipping citrus trees off above the soil, the study has shown that Remedy is effectively controlling stump sprouts at all treatment times of 0, 24, 48 or 72 hours after shearing the tree off above the soil. However, in a few cases lateral roots did exhibit limited sprout formation even with the stump being treatment with Remedy. Complete and thorough coverage of the cut surface is essential for adequate control. Sprouts were produced from the untreated stump as well as on limited lateral roots. The size of stump also impacts sprout formation with larger trunks producing more sprouts and quicker after shearing as compared to smaller stumps. Rootstock also has an impact on sprout formation. Rough lemon rootstock tends to produce more sprouts from the trunk and lateral roots than other rootstocks in this study. However, with limited data the difference between rootstock is not able to be determined. Sprouts from stumps will be collected and tested for HLB (greening) during the next quarter.
The project has two specific aims. We outline below the progress made for each of them. Specific Aim 1: Identify sweet orange responses to Huanglongbing disease (HLB) through deep transcriptome profiling using next generation DNA sequencing technologies. Transcriptome analysis was conducted using Illumina Genome Analyzer II. We obtained read lengths up to 125 bp pair-ended for the entire set of mature fruits analyzing the four types of samples: symptomatic, asymptomatic and apparently healthy from infected locations and healthy plants from disease-free location. 25 and 31 million reads/sample were obtained and 12,8-18.99 million reads were mapped to the NCBI citrus Unigene set (15808). Using velvet assembly we obtained 158,656 contigs that were BLASTed to the following databases : JGI 1.2x citrus genomic sequences (78.6% of contigs had hits), NCBI C. sinensis Unigene set (45.7% of contigs had hits), TAIR9 protein sequences (53.6% of contigs had hits) poplar protein sequences (58.7% of contigs had hits). About 1156-1734 of the citrus unigenes were differentially regulated in pair wise comparisons of symptomatic, asymptomatic, and apparently healthy fruits with control fruits. We have also completed construction of three of the four cDNA libraries from immature leaves. We are planning to obtain the last sample (control) to soon run this set of samples before the analysis of the mature leaves and immature fruits. Additional RNA was isolated from both mature fruits and leaves for the validation of the results using qRT-PCR. Specific Aim 2: Define and validate gene networks and identify host (sweet orange) response biomarkers regulated by HLB at different stages of infection. We have analyzed transcriptome data using BLAST2go, mapman, pathexpress software. A gene set enrichment analysis has revealed the following HLB-regulated pathways: starch metabolism, lignin, stilbene biosynthesis, phenylpropanoid, sucrose metabolism, coumarin, salicylic acid and hormone metabolism. A pair wise comparison between the four types of samples has identified several genes that can be used to distinguish infections before the symptoms appear. After qRT-PCR validation these genes can be confirmed as biomarkers for asymptomatic diagnosis. A list has been made including genes belonging to sucrose metabolism, photosynthesis, cell wall metabolism, lipid metabolism and certain volatile metabolite pathways. Terpene metabolism was highly regulated in apparently healthy compared to HLB-infected fruits. Jasmonate, salicylic acid, ethylene and auxin were the key hormones involved in host response to HLB-infection in the fruit peel tissues. Among transcription factors, WRKY, NAC, MYB, EREBP, bZIP were mostly highly regulated after the infection proceeding from the asymptomatic stage. With respect to cell biology processes the most significant up regulation was observed in pathways related to, protein modification, protein degradation, transport, metal handling pathways, while cell division and cell cycle and organization were down regulated. Up regulation was observed in HLB symptomatic fruits for genes involved in Photosystem II and redox chain reactions as well as ubiquitin dependent protein degradation. The following receptor like proteins were found to be involved in the host responses in the fruit: leucine-rich-repeat protein receptors, DUF26, WAK, thaumatin. We are conducting a network analysis of citrus responses based on a Arabidopsis knowledge base of protein-protein interactions to determine which genes play key roles in the cascade of HLB-regulated pathways (hubs or connecting hubs). Such genes will be excellent candidates as biomarkers for diagnosis and as targets for therapeutics. Currently, qRT-PCR has been performed to validate 6 genes belonging to starch, sucrose, hormone metabolism and additional analysis will include up to 20-30 genes to identify biomarkers for asymptomatic host responses in fruit to HLB.
We continue to make progress in our goal to identify proteins that are specific to Liberibacter-infected citrus. These proteins, and the antibodies made against them, will be used in the development of a rapid diagnostic tool for the disease in the field. We have previously reported that all IgYs that will be necessary to perform pCMAT are in hand. Briefly, IgYs were raised in chickens against diseased plants at 3 months, 1 year and 2 years post-infection. The material from each time point was injected to two hens for IgY production. One of the two hens used for the 3 months time point unfortunately entered molting and had to be discarded. The last batch of antibodies was shipped to Oragenics on October 7th. It was demonstrated by Western analysis that the IgYs that were raised against diseased plants are broadly immunogenic. All pools of hen serum had a strong reactivity against plant extracts. Several hundred bands were observed as well as a smear, which indicates a broad and measurable reactivity against a large number of plant (and presumably pathogen) proteins. The next phase of this study aimed at optimizing the steps of IgY adsorption using healthy plant material. This was to deplete the IgYs from those antibodies that are reactive with proteins constitutively expressed in healthy and in diseased plant tissues. This was performed using a proprietary method called PCMAT. Briefly, IgYs were adsorbed against insoluble fractions of homogenized and sonicated healthy plant tissues. To monitor the adsorption process, sequential absorption of pooled IgYs was performed using PCMATTM against extracts from healthy plant tissues. The IgYs that did not bind to healthy plant tissue were collected and purified, and used to probe cellular extracts in an ELISA format. Clearly, the adsorption process is not yet complete after 5 rounds of adsorptions with insoluble material, yet resulted in a significant decrease of reactivity against healthy plant extracts. We are currently optimizing the conditions necessary to bind soluble healthy plant extracts to solid matrices, which will be necessary for further absorb the IgYs; until the reactivity against healthy plant extracts reaches background levels. Thoroughly adsorbed IgYs which specifically retain their reactivity against extracts from diseased specimens should be available by the end of the month. In summary, we are on schedule to perform the experiments described in the original application. The next few months will be spent on refining the adsorptions process with soluble plant extracts. The resulting IgYs will then be used as a probe to isolate proteins of citrus that are specifically induced during infection as well as gene products of the bacterial pathogen produced during that process. This will be performed by immunoaffinity and LC-MS/MS, as originally proposed.
Fall tests of herbicide applications to trunks continued into the winter. Efficacy of alternatives to 1 1/2 inch hatchet blade cuts are being evaluated. Spring and Summer tests, when trees are more difficult to kill, are critical for this testing. An electro-hydraulic mechanized tree trunk incision clamp system was developed and tested first on a 4 X 4 ATV, and later transferred onto a utility tractor to assist with weight distribution and hydraulic power source issues. Also, to deal with operating stresses, the clamp system was revised to incorporate a larger bore cylinder and a lesser resistant serrated clamping mechanism. The increased force from the larger cylinder made it necessary to build components from heavier materials. Also, due to fluctuating tree trunk shapes and sizes, the clamping mechanism wasn’t able to consistently contact the tree without shearing pivot point components. Further testing identified the need to decrease the system weight and to create a lesser resistant means for making tree trunk incisions, in order to become an efficient system. A less-resistant methodology consisted of changing the system function for the cutting mechanism. These changes compensate for the trunk size and shape variances and will allow reductions in system weight. Recent orchard tests have indicated the cutting mechanism functions with less resistance and it is more consistent in performance and efficiency. The spray system was added to provide a complete prototype and several trees were treated with either undiluted or 1 to 1 diluted Arsenal solution. The efficacy of the complete system in these treatments will be evaluated over the next 4 weeks. Full evaluation of the system needs to occur during the spring-summer when trees are more difficult to kill. We are now asking CRDF for guidance on issues about patents and a manufacturing agent for the machine. The results of the 600 acre aerial spray test applied by South Florida were reported last quarter. The one area not completely killed will be evaluated this quarter. Aerial sprays treatments to the additional 40 acres of abandoned grove at this site has been delayed until some issues on chemicals to be tested are resolved. It is still planned to have 4 treatments with 2 replications (5 acre plots) in order to evaluate selected chemicals. Work on use of the soil fumigant, Midas, has also been delayed until the parent company can make their representatives available to continue the tests. Apparently, they are currently involved in testing their product on vegetable sites. They expect to be available by early spring.
The new variable rate controller continues to perform well in the approximately 50 units currently deployed in Florida citrus groves. Testing is continuing on the addition of a liquid crystal display, with the current focus being to implement all the user input and operation status functions via a single color touchscreen and rugged weatherproof enclosure. The elimination of all physical switches and pilot lights from the enclosure will increase the durability of the controller and make it easier to upgrade all of its functions by using virtual switches and status lights on the display. We have completed tests on the latest range of ultrasonic and optical sensors with the variable rate controller in order to offer alternative sensor packages to the users. The different sensor beam widths and speeds now available could have advantages in different grove situations (e.g. bedded flatwoods versus ridge) and the controller should be able to accept an array of different sensors, including optical, ultrasonic and microwave (radar). Effective systemic insecticides like imidacloprid have a labeled annual application rate limit of active ingredient (0.56 kg/ha) which constrains their application frequency to about 4.5 per year when broadcast with air-blast sprayers. Variable rate technology and band application can extend the efficacy of such rate-limited agrochemicals per planted hectare by precision placement of active ingredient only to the tree canopies. A sprayer used in young groves must have a very fast response speed because the tree canopies and adjacent spaces are small. A precision row sprayer with a hoop-shaped boom was developed for young citrus, using the prototype variable rate control system. By using a fast solenoid valve (<0.2 s), the sprayer is capable of real-time pesticide flow control according to the presence of citrus tree canopies. The sprayer system consists of a WAAS GPS receiver, near-infrared diffuse reflectance optical sensors, electric solenoid valve and two embedded microcontrollers with synchronized look-ahead and adjustable buffering. Spray savings were estimated to exceed 90% in the first year, and average 50 to 75% by the time the trees become productive, thus allowing more efficient and flexible use of rate-limited pesticides and reducing production costs for greater profitability. A video of this sprayer is available at http://128.227.177.113/pa/Video.html We are conducting research on finding ways to change the pesticide spray flow rate in real time instead of the simple on/off actions currently employed. In conjunction with sensor information about canopy density, the proportional flow control would allow customization of spray amounts as needed per unit of canopy and its condition. So far the stepper motor valves tested have shown potential for effective proportional flow control, with response times better than 0.5 s.
The two primary objectives of the current proposal are: (1) determine and compare the metagenomes of healthy and infected citrus phloem, and (2) continue to develop and apply AthenaBio’s cultivation technologies towards producing a pure culture of Ca. Liberibacter. Objective 1 Discussion. Laser cutting microdissection (LCM) will be developed to purify phloem-containing sieve cells from both healthy and infected citrus samples. DNA extracted from the purified cells will be analyzed by 16S rRNA sequencing to evaluate the contained microbial diversity. The most diverse DNA samples will be selected for large-scale metagenomic analysis. All DNA extraction and sequencing work will be performed by Dr. Ravel’s group (U. of Maryland) Objective 2 Discussion. AthenaBio will continue to develop and apply its cultivation technologies to isolate a pure cultures of Ca. L. species. In addition to Ca. Liberibacter species, we aim to cultivate other phloem endophytes to provide reference genomes for metagenomic analysis. Two approaches will be employed to disrupt quorum sensing (QS) and thereby enhance growth: (1) physical methods (i.e. a ‘flow cell’) will be used to wash QS signals away from cells, and (2) enzymes will be used to inactivate QS signals through chemical modification. The advantage of the flow cell approach is that knowledge of the chemical structure of the QS signal is not required, and this approach is therefore expected to work on a broad variety of QS systems. Both approaches (flow cell and enzyme) will be combined with various media formulations and incubation conditions. Work has continued in method development for LCM of citrus phloem cells. Healthy leaf tissue was dissected into small pieces (~5mm), immediately placed in tissue freezing medium and placed on dry ice. Various thickness levels and staining procedures were tested. We believe we have a working method but the instrument was out of order for a period of time. The instrument is now fully functional and we are anticipating DNA analysis to begin by mid-February.
For the first year of research on healthy and HLB samples from one Hamlin and two Valencia harvests (15+ trees/sample), reports were given on sugar, acid, brix, titratable acidity, ratio and oil content, as well as sensory perception of flavor. The results showed that there were minimal differences between juice from healthy and HLB trees for asymptomatic fruit, but that there were differences for symptomatic fruit, both chemically and in terms of sensory perception. Now we present the data for these samples for aroma, which includes both gas-chromatography-olfactometry (GCO) and GC-mass spectroscopy (GC-MS) analyses. For GCO analysis, each sample evaluated was analyzed twice using two trained olfactory assessors. Assessors rated aroma intensity continuously throughout the chromatographic separation process using a linear potentiometer whose output was recorded and quantified using chromatography software. Retention times and verbal descriptors were recorded to permit aroma descriptors to be coupled with computerized aroma time-intensity plots. Only those compounds which were described similarly at a similar retention time and observed by half the panel considered as aroma active components. Both pasteurized and unpasteurized samples for healthy and symptomatic HLB juice were analyzed Forty-six aroma components were observed in the set of juices. Total aroma intensity from the juices were summed and averaged. Early season HLB symptomatic fruit had less overall aroma intensity than control (healthy). However, in more mature fruit there is little difference in overall intensity. On average, pasteurization increased total GCO aroma intensity. The samples were analyzed by GC-MS, primarily to identify and confirm the GCO data. Forty-two MS analyses on three sets of juice from the 2008-2009 season were analyzed. One set was Hamlin and two sets were Valencia. Within each set there were two reps of control (healthy), symptomatic and asymptomatic juices. Each set also contained paired pasteurized and unpasteurized juices of each juice type. In examining the GCO data set, it was observed that the vast majority of aroma active volatiles were very similar between HLB symptomatic and control juices. However, a few volatiles were found either exclusively or at higher aroma intensity in certain sample types. In the case of Valencia juices, 2-methylbutanol (minty, green) and decanal (citrus, fatty) were found primarily in HLB pasteurized juices. ‘-myrcene (geranium, earthy, grass) and octen-3-one, 1 (mushroom, metallic) were identified in both HLB and control unpasteurized juices, but with a higher intensities in HLB juice. An unknown minty, ocimene (tropical); 1-decanol (spicy, cilantro), farnesyl acetate (perfume); p-cymene (citrus, green) were observed only in unpasteurized control juice. Volatiles such as methional (cooked grain); dodecanal (citrus, soapy, green) and cadinene (minty, burned) were found almost exclusively in pasteurized control juice. Early season HLB symptomatic fruit had less overall aroma intensity than control (healthy). However, in more mature fruit there is little difference in overall intensity. On average, pasteurization increased total GCO aroma intensity. Since the late season Hamlin harvest was missed in 2009, samples for Hamlin were harvested last week. Analyses on these samples will be added to the final quarter first year report. The second year of research will include the establishment of thresholds for off-flavored symptomatic HLB juice in normal juice and of identified off flavored compounds in normal juice. This work has already been initiated. Also studies are underway to compare HLB juice from trees under normal cultivation and spray regimes to those undergoing nutritional and salicylic acid sprays as well as to healthy trees for chemical and sensory analyses.
For the first year of research on healthy and HLB samples from one Hamlin and two Valencia harvests (15+ trees/sample), reports were given on sugar, acid, brix, titratable acidity, ratio and oil content, as well as sensory perception of flavor. The results showed that there were minimal differences between juice from healthy and HLB trees for asymptomatic fruit, but that there were differences for symptomatic fruit, both chemically and in terms of sensory perception. One exception, chemically, that was not presented in previous reports, is that limonin and nomilin were higher in asymptmatic and especially symptomatic fruit compared to controls for January Hamlin and April Valencia, but not June Valencia, reflecting sensory results. Levels were below published taste thresholds for these compounds, however. Now we present Russ Rouseff’s data for aroma, which includes both gas-chromatography-olfactometry (GCO) and GC-mass spectroscopy (GC-MS) analyses. For GCO analysis, each sample evaluated was analyzed twice using two trained olfactory assessors. Only those compounds which were described similarly at a similar retention time and observed by half the panel considered as aroma active components. Both pasteurized and unpasteurized samples for healthy and symptomatic HLB juice were analyzed. Forty-six aroma components were observed in the set of juices. Total aroma intensity from the juices were summed and averaged. Early season HLB symptomatic fruit had less overall aroma intensity than control (healthy). However, in more mature fruit there is little difference in overall intensity. On average, pasteurization increased total GCO aroma intensity. The samples were analyzed by GC-MS, primarily to identify and confirm the GCO data. Forty-two MS analyses on three sets of juice from the 2008-2009 season were analyzed. One set was Hamlin and two sets were Valencia. Within each set there were two reps of control (healthy), symptomatic and asymptomatic juices. Each set also contained paired pasteurized and unpasteurized juices of each juice type. In examining the GCO data set, it was observed that the vast majority of aroma active volatiles were very similar between HLB symptomatic and control juices. However, a few volatiles were found either exclusively or at higher aroma intensity in certain sample types. In the case of Valencia juices, 2-methylbutanol (minty, green) and decanal (citrus, fatty) were found primarily in HLB pasteurized juices. ‘-myrcene (geranium, earthy, grass) and octen-3-one, 1 (mushroom, metallic) were identified in both HLB and control unpasteurized juices, but with a higher intensities in HLB juice. An unknown minty, ocimene (tropical); 1-decanol (spicy, cilantro), farnesyl acetate (perfume); p-cymene (citrus, green) were observed only in unpasteurized control juice. Volatiles such as methional (cooked grain); dodecanal (citrus, soapy, green) and cadinene (minty, burned) were found almost exclusively in pasteurized control juice. Early season HLB symptomatic fruit had less overall aroma intensity than control (healthy). However, in more mature fruit there is little difference in overall intensity. On average, pasteurization increased total GCO aroma intensity. Since the late season Hamlin harvest was missed in 2009, samples for Hamlin were harvested last week. Analyses on these samples will be added to the final quarter first year report, but the limonin and nomilin analyses are done and show higher levels in HLB juice, similar to the 2009 Hamlins. The second year of research (already underway) will include the establishment of thresholds for off-flavored symptomatic HLB juice in normal juice and of identified off flavored compounds in normal juice. Also studies are underway to compare HLB juice from trees under normal cultivation and spray regimes to those undergoing nutritional and salicylic acid sprays as well as healthy trees for chemical and sensory analyses.
Five-band Sensor Development: At Citrus Research and Education Center (CREC), a passive five-band sensor was developed and integrated as a portable, relatively inexpensive, robust, mobile sensor module that can be incorporated with the sensor platform (developed in CREC) in a gator for stress-detection applications. The sensor can measure the reflectance in five spectral bands (three visible and two infrared), and has the capability of storing the GPS coordinates and for incident light intensity correction for acquiring stable spectral reflectance. The sensor is being field tested for its ability to accurately identify stressed citrus trees in different citrus varieties. This sensor will be further tested for HLB detection. Greenhouse Experiments: The preliminary results from greenhouse experiments (CREC) for detecting HLB in early stages indicated that the spectroradiometer reflectance data could be used to identify HLB-infected trees with a classification accuracy of > 90% for Hamlin and Valencia varieties. The results are being further validated with a larger number of dataset. Aerial Hyperspectral Imaging: A new set of aerial hyperspectral imagery has been acquired in a citrus grove with the help from Dr. Yang and his colleagues (USDA ARS, Texas). The imaging site was the 20 acres Summerland grove (Consolidated Citrus Ltd.). The grove contained some dead trees (scaffold), probably from tristeza, citrus blight and HLB. Ground truthing was conducted for diseased trees. The aerial imageries along with ground data are currently being analyzed to further develop and evaluate the HLB detection algorithms. The results will be reported in next progress report. Differential mobility spectrometer sensor: In this quarter, we designed, manufactured and tested a volatile collection device holder that improved the reproducibility and eliminated false positives that occur when the collection device makes direct contact with the leaf surfaces. In addition, we expanded on our previous study of measuring volatile organic compounds (VOCs) of symptomatic and healthy sweet navel oranges. The sample collected in this quarter consisted of three groups: control plants in grove A (commercial grove, Lake Alfred, FL), healthy & asymptomatic plants in grove B (Experimental Station, Oakland, FL), and symptomatic plants in grove B. The VOCs were collected using polyacrylate solid phase microextraction (SPME) fibers and shipped to Dr. Davis’ lab for gas chromatograph/mass spectrometer analysis to classify the disease status by the VOC profile and possibly identify chemical biomarkers. As there was a large overlap between the chemical peaks among the groups, the distinction between the groups was not very clear using linear classification techniques. However, using higher order machine learning algorithms, we were able to create a model that has > 85% classification accuracy. Specifically, a couple of data analysis methods were applied to this data set. First, through a peak-by-peak comparison, we detected 11 peaks (potential biomarkers) where we could see a visible difference between the peak amplitude distributions of two groups: 5 healthy vs. 15 infected samples (p<0.15). Working on the entire chromatogram data of the samples, principal component analysis and neural network model was used for a further quantitative examination of the seperability of these two groups. Based on the leave-one-out strategy, we obtained a promising result with the specificity being 80% and the sensitivity being 87%. This diagnosis result is a promising direction for our biomarker detection and disease diagnosis research of the next step. The results of this quarter were significant because we learnt that experimental on-field conditions are very challenging because the wind mixes the VOCs and makes securing the SPME fibers difficult, and the presence of other diseases such as melanosis and leaf minor on some of the plants impact our capability of getting a clean VOC profile for HLB. We have taken these into account for our next set of studies that are currently underway.
Research results from the first 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 results reported in previous FCPRAC progress reports. Several experiments were conducted since the 2007-08 season. Preharvest treatments consisted of 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. Treatments were conducted on White grapefruit twice, Valencia oranges once, and Flame red grapefruit once. So far, we have found less promising results with foliar K alone than we did in the previous season. In one experiment, foliar K even increased peel breakdown. The reason for this is unclear. However, applying a combination of foliar K and Mg or use of an antitranspirant resulted in a significant reduction in peel breakdown during postharvest storage. Application of foliar Mg (Epsom salt) alone showed a trend in reducing peel breakdown that, while not significant, is worth repeating. We are still conducting harvests from the Flame grapefruit block that was treated in October. In general, the fruit has been very healthy with fruit harvested in early Nov. showing very little decay or peel breakdown even after over 2 months storage at ambient temperatures. We have also conducted four sets of postharvest experiments that gave similar results as the previous year. Interestingly, postharvest treatments with Imazalil sporadically reduced peel breakdown as was noted the previous year. The reason for this is unclear but worth investigating further. We are currently setting up another experiment in different Flame grapefruit block that will include a cocktail of materials, in addition to the previous treatments, to see about reducing peel breakdown. Soil water content probes are installed in a water-stress and control plot and we have arranged to have leaf and peel samples for nutrient content.
Objective 1: We conducted tests with nutrient analyses of leaf samples collected from HLB-infected and healthy trees to establish relationships that could be used for (early) diagnosis of HLB. Starch accumulation in symptomatic leaves can significantly alter the interpretation of nutrient status when diagnosed on a dry weight concentration basis. When corrected for starch accumulation by using specific leaf area, Ca, Mg and B are consistently lower and K is higher in HLB-infected trees compared to healthy trees under the same fertilization program. These changes exist in both symptomatic and asymptomatic leaves from infected trees. Similarly, separation between asymptomatic healthy leaves and asymptomatic infected leaves could be achieved using the Diagnosis and Recommendation Integrated System (DRIS), particularly when assessing the Dry Mass Index (DMI) generated by the method. DMI integrates the imbalances caused by nutrient losses and starch accumulation in HLB-infected leaves, thus could indicate infection early on despite the absence of visible symptoms. We are currently sampling leaves from trees infected with other diseases (CTV, blight, phytophthora) to determine if these changes are HLB-specific and can be used as a diagnostic method. Due to stress-related abscission of HLB-symptomatic leaves from the exterior of infected canopies for much of the year, we are often forced to sample symptomatic leaves from within the canopy and asymptomatic leaves from the outside. We tested the effect of this unavoidable sampling position difference on the measured leaf nutrients analyzed in healthy trees and found no significant differences in 11 elements measured. Objective 2: Two field studies have been established to evaluate the effects of various foliar nutrient applications on the expression of HLB in infected trees by evaluating tree nutrient status, growth, yield and visual tree appearance through photographic documentation. The first of these trials is in a heavily infected mature Hamlin grove in south Florida. Since the initiation of the project the trees in this study have received four foliar applications of one of nine different treatments. Untreated trees serve as controls. The trees were harvested in December 2009. The second study is in a young (3-5 years old) commercial Valencia grove in Haines City. Treatments in this study have been underway for approximately 4 months and include fertigation in addition to foliar nutrient sprays. Our initial efforts at this site have been to demonstrate the ability to raise the levels of specific nutrients involved in plant defenses within trees. Since treatments began, B levels have been successfully raised to near toxic levels within infected and healthy trees using both foliar and fertigation applied B. This has demonstrated that good nutrient uptake can be achieved through the treatment methods in a relatively short period of time. Analysis of the large quantity of these studies have generated is still being analyzed as of the writing of this report. As soon as the analyses are complete information will be passed along to the Florida citrus community. Objective 3: A hydroponics system has been constructed in an HLB approved greenhouse at the CREC. Hamlin trees grown from rooted cuttings were purchased from a commercial nursery and are currently being established in the hydroponics system. Once the plants have made the transition from soil to hydroponics culture half of them will be inoculated with HLB by grafting material from greenhouse-grown trees with known HLB infections. All of the plants will be maintained with a complete nutrient solution until HLB symptoms develop on the inoculated trees. Previous experience with greenhouse graft inoculation indicates that this will take about 90 days. At that time, healthy and infected trees will be subjected to various changes in the nutrient solution composition. Based on the data from objective 1, the levels of Ca, Mg, and B and combinations of these three nutrients will be changed to determine plant response with and without HLB. Trees remaining in complete nutrient solution will serve as controls. This experiment will allow us to begin to separate nutrient and HLB effects on plant growth and understand the interaction of nutrition and HLB infection. Data from this project has been the basis for two Citrus Industry magazine articles in 2009. In addition, the data has been presented at Citrus Expo, the Florida State Horticultural Society conference and at six different regional grower meetings. An EDIS document about the general effects of plant nutrition on disease tolerance is being written.
(863) 956-8817
(863) 956-5894
700 Experiment Station Road
Lake Alfred, FL 33850
Email Us
