Oral uptake of dsRNA targeting specific Asian citrus psyllid genes can induce psyllid mortality and reduce Liberibacter titer in infected psyllids. Significant progress has been made with research on the use of RNAi as a means of Asian citrus psyllid control. We have made use of a Citrus tristeza virus (CTV) dsRNA expression system which when inoculated into C.Marcrophylla results in leaf phloem containing dsRNAs which target essential ACP genes. When ACP feed on leaves from these “paratrangenic” (CTV transfected) citrus plants, mortality was double that observed when the ACP fed on artificial diets containing this synthetic dsRNA, greater than 80% mortality. Currently, new versions of paratransgenic citrus are being produced to determine the best sequence to use targeting this essential gene. The use of the Ion Torrent (by Life Technologies) next generation sequencing has provided rich insight into the transcription profile of “paratransgenic” fed “sick” ACP using comparative RNA Seq analysis. The data support the specific down-regulation of the dsRNA target gene as the cause of mortality as seen by the significant perturbation of genes in the molecular pathway in which this gene functions. RNA Seq data continues to be collected from ACP fed on a variety of gene specific dsRNA containing diets. Experimentation has also begun using artificial diets containing dsRNAs synthesized by a novel “mass-production” technique that would make it practical for the use of RNAi technology as a field application. The use of topical application in conjunction with “paratransgenic” plant varieties presents a strategy for effective delivery and a multiple gene targeting employment of the RNAi pest control technology.
We hypothesized that groves with high bicarbonate stress are suffering from HLB because they support lower fibrous root density compared to groves with lower bicarbonates (less than 100 ppm) in irrigation water and/or soil pH (less than 6.5). To confirm this relationship, we surveyed 37 grove locations in Highlands and Desoto counties with varying liming history and deep vs. shallow wells mostly on Swingle and Carrizo. Lower root density is significantly related to well water pH greater than 6.5 and to soil pH greater than 6.2. Yield records from these blocks reveal that groves under high bicarbonate stress production have declined 20 percent over the last three seasons (2009-2012) in contrast to Ridge groves with low bicarbonate stress which have increased 6 percent in production even though HLB incidence has accelerated. The yield losses are correlated with less fibrous root density, which reduces root system capacity for water and nutrient uptake. Evidence from research on other crops indicates that bicarbonate impairs the root’s ability to take up important nutritional cations including Ca, Mg and K, as well as micronutrients, especially Mn and Fe. In Florida, Bryan Belcher of Davis Citrus Management has acidified irrigation water with sulfuric or N-furic acid (a mixture of urea and sulfuric acid) by injection at the well in the same way as fertigation. N-furic has the advantages of being safer to handle and providing some additional N due to the urea component, but the disadvantage is higher cost of treatment compared to sulfuric acid. Since irrigation is not necessary when it rains, acidification treatment only occurs during the dry season when the bicarbonates are loading into the wetted area under the tree. When the rain begins, these bicarbonates are flushed from the rhizosphere. Our labs and grower cooperators are also evaluating acidification of soil by amendment with elemental sulfur applied in prilled or finely ground form. Sulfur (S) releases acid when it interacts with Thiobacillus bacteria in soil to form acid (H+) ions. This process of acidification with S is slower than treatment of the water but provides for longer lasting reduction in soil pH. Sulfur can be applied in prilled form with a fertilizer spreader or with a herbicide boom as a slurry. Sulfur can also be added to dry and fertigation formulations to lower the pH by as much as one unit after repeated ground applications.
We hypothesized that groves with high bicarbonate stress are suffering from HLB because they support lower fibrous root density compared to groves with lower bicarbonates (less than 100 ppm) in irrigation water and/or soil pH (less than 6.5). To confirm this relationship, we surveyed 37 grove locations in Highlands and Desoto counties with varying liming history and deep vs. shallow wells mostly on Swingle and Carrizo. Lower root density is significantly related to well water pH greater than 6.5 and to soil pH greater than 6.2. Yield records from these blocks reveal that groves under high bicarbonate stress production have declined 20 percent over the last three seasons (2009-2012) in contrast to Ridge groves with low bicarbonate stress which have increased 6 percent in production even though HLB incidence has accelerated. The yield losses are correlated with less fibrous root density, which reduces root system capacity for water and nutrient uptake. Evidence from research on other crops indicates that bicarbonate impairs the root’s ability to take up important nutritional cations including Ca, Mg and K, as well as micronutrients, especially Mn and Fe. In Florida, Bryan Belcher of Davis Citrus Management has acidified irrigation water with sulfuric or N-furic acid (a mixture of urea and sulfuric acid) by injection at the well in the same way as fertigation. N-furic has the advantages of being safer to handle and providing some additional N due to the urea component, but the disadvantage is higher cost of treatment compared to sulfuric acid. Since irrigation is not necessary when it rains, acidification treatment only occurs during the dry season when the bicarbonates are loading into the wetted area under the tree. When the rain begins, these bicarbonates are flushed from the rhizosphere. Our labs and grower cooperators are also evaluating acidification of soil by amendment with elemental sulfur applied in prilled or finely ground form. Sulfur (S) releases acid when it interacts with Thiobacillus bacteria in soil to form acid (H+) ions. This process of acidification with S is slower than treatment of the water but provides for longer lasting reduction in soil pH. Sulfur can be applied in prilled form with a fertilizer spreader or with a herbicide boom as a slurry. Sulfur can also be added to dry and fertigation formulations to lower the pH by as much as one unit after repeated ground applications.
Stress intolerance of HLB trees is a direct consequence of a more than 30 percent loss of fibrous root density compared to non-diseased trees. This root loss may be compounded by interactions with root pathogens and pests such as Phytophthora spp. Progagules of Phytophthora spp. are elevated in the rhizosphere soil of HLB trees. When Las interacts with Phytophthora, fibrous root loss can be greater than that caused by HLB alone, depending on the grove location and time of year. Although not directly studied, HLB associated root loss has been attributed to carbohydrate starvation. Initially, canopy symptoms are not apparent in trees with significant root loss, so starch concentration in roots was determined. Surprisingly, early root loss occurred without a drop in starch. This disappearance of roots could be due to the lack of new growth to replace old roots during the normal cycle of death and regeneration of fibrous roots, or may be associated with premature root death unrelated to phloem plugging, or a combination of the two processes. Also, it is highly likely that Las damage alters the concentration of soluble sugars in roots by increasing leakage from roots that attracts and accelerates infection by root pathogens such as P. nicotianae.
Seasonal root sampling continues in two field sites with a third site identified. Sampling has already revealed seasonal variation in root infections and apparent shifts in the root flush cycle caused by Liberibacter. We have swapped out root cages twice and collected data on root flushes in response to HLB during a normal root flush and nonflushing time. Data analysis on initial root flush data in comparison to overall root density is underway and data collection will continue through the year. Some difficulty is being encountered with finding sufficient presumed healthy trees in the field, so more emphasis will be placed on greenhouse experiments as the project continues. We will continue to look for new sites with moderate infection rates and trees old enough for routine root sampling. Sampling at a rootstock trial site is underway with a full year of data on the effects of HLB on these new experimental rootstocks. This has already begun to demonstrate how these new rootstock lines respond to Liberibacter infection. We are in the final preparation stage of testing the most promising rootstock lines in more controlled greenhouse studies and are awaiting improved weather conditions for high efficiency inoculation Liberibacter.
FireWall (22.3% streptomycin sulfate; Agrosource, Inc.) was granted by an EPA section 18 registration for control of citrus canker in Florida grapefruit. The label for FireWall restricts use to no more than two applications per season. As a condition for FireWall registration, EPA requires monitoring of Xanthomonas citri subsp. citri (Xcc) for streptomycin resistance in treated groves. The objective of this survey was to apply our published protocol for sampling canker-infected grapefruit leaves for isolation and detection of streptomycin resistant Xcc. Two sites in two commercial grapefruit groves and a trial site treated the last 4 out of 5 years with FireWall (including 4 sprays in 2013) were sampled for streptomycin resistance on September 24 and 25, 2013. From the last mature flush that earlier in the season was directly contacted by the FireWall in spray treatments, leaf samples with canker lesions and one sample without lesions (check) were collected at five locations (four corners and the center) in each of the grove blocks. From each sampling location, 4 fully expanded leaves with canker lesions were placed in a flask. As a streptomycin-positive Xcc check, a single leaf with canker lesions produced by injection-infiltrated with a streptomycin resistant Xcc strain was placed in the flask with four asymptomatic leaves. Total colonies recovered from KCH selective medium ranged from 6 to 107 . 104 cfu/ml of washate . The number of bacteria that grew on KCH-S from the samples ranged from 0 to 80 cfu/ml and from the inoculated checks ranged from 3 to 36 cfu/ml. The identity of each yellow colony on KCH-S as Xcc was tested with Agdia test strips. No resistant Xcc colonies were detected from the grove samples with canker lesions. Positive (resistant) Xcc were only detected from the checks that had been spiked with grapefruit leaves inoculated with a streptomycin resistant strain. Recovery of streptomycin resistant colonies validated the protocol for isolation and detection of resistant Xcc from field leaves with other bacteria present. Conclusion: No Xcc resistant to streptomycin were recovered from four commercial groves locations nor were resistant Xcc recovered from the trees in a trial at Ft. Pierce that received four sprays of FireWall or Firewall plus two sprays of Kocide in 2013. The trees in Ft. Pierce had a history of 33 sprays (11 per season) of FireWall or FireWall+ Kocide from 2009-2011. Hence, Xcc in this location have a 4 year history of high frequency exposure to streptomycin without development of resistance thus far.
Objective 1. To define the role of chemotaxis in the location and early attachment to the leaf and fruit surface. Strains of Xanthomonas citri subsp. citri (Xcc) and other Xanthomonads sense signals from the host which facilitate the location of leaf entry points that are specific for each bacterium-host association. In addition, there are differences in swimming as well as surface motility among the wide and narrow host range strains of Xcc. Differential biofilm formation was observed in vitro and in planta among the Xanthomonas strains assayed. Minimal medium XVM2 significantly contributed to biofilm formation for every strain analyzed. Biofilm production of wide host range Xcc did not differ between minimal XVM2 and the nutrient rich LB medium. In contrast, narrow host range Xcc strains formed much more biofilm in XVM2 medium. In addition, results in planta showed that the Xanthomonas stains assayed produced much less biofilm on non-host leaves or fruits than the respective host plant. Furthermore, SEM of biofilms onleaf and fruit surfaces revealed different structure biofilm aggregates betwenn narrow and wide host range Xcc strains. Large differences in swimming motility between wide and narrow host range strains were also found. It appears that narrow host range strains of CBC partially compensate for the lack of biofilm formation with the greater aggregation in the appropriate niche and higher swimming motility. Objective 2. To investigate bifofilm formation and composition and its relationship with bacteria structures related to motility in different strains of Xcc and comparison to non-canker causing xanthomonads. Two assays were performed to determine extracellular DNA presence in biofilm matrix. DNase (SIGMA) was added to bacterial culture in XVM2 media at 0, 24, 48 and 72 hours incubation. In such assays, effect of the DNAse was clear again at the initial stages of biofilm formation (0h), confirming the role of the DNA at the first stage of the biofilm for Xcc and X. alfalfae subsp. citrumelonis. Effect of the DNase in Aw and A* strains differed according to time of incubation. No clear effect of the DNAse was found in Xanthomonas campestris pv. campestris. To visualize DNA in biofilms, bacterial aggregates on glass surface were stained with Propidium iodide and SYTO9. Fibers visualized by violet crystal were stained, confirming the fibers contain DNA. Simultaneously matrix proteins were stained with biofilm SPYPRO (Invitrogen). Most of the fibers have been identified by these two methodologies as containing a mix of DNA and proteins. Transcription of genes fimA (XAC3241), fimA (Xac3240) (Pilus type IV), fliC (flagelin), fleN (flagella regulator), pilA (fimbria), motA (flagella motor), rpfF and rpfB (quorum sensing signal) were analysed on dried and inundated biofilm as well as the planktonic bacterial stage after 72 h incubation in both LB and XVM2 culture medium. Three independent assays were started and from one of them RNA was already extracted and amplified by RT-PCR. RNA was obtained from aggregates produced at the bottom of culture flask after or before the dry process as well as aggregates of the air-liquid interface. Analysis of the results is under progress but some preliminary infers could be concluded. The assay revealed that in wide and narrow host range Xcc strains, fimA (XAC3241) expression was higher in early stage of the biofilm formation as compared to after biofilm establishment. This gene was particularly upregulated in XVM2 medium compared with the biofilm or planktonic stages of the bacteria. When planktonic and biofilm bacteria were compared, differential expression of flagella genes (fliC, fleN, motA) was found between the wide and narrow host range strains when incubated in LB medium but not in XVM2 medium. transcription of the genes’.were analysed on dried and inundated biofilm as well as planktonic bacteria stage
We initiated an Illumina NextGen DNA sequencing project comparing ‘Sun Chu Sha’ mandarin (HLB tolerant) and ‘Duncan’ grapefruit (HLB sensitive) infiltrated with Candidatus Liberibacter asiaticus flagellin 22 (CLas-flg22) peptide and water control. Between 42 and 63 millions of pare-end reads were generated from our six cDNA libiaries per genotype (three replicates for CLas-flg22 treatment and control, respectively). Using Citrus clementina genome as the reference, an average of 45% of reads for each sample were uniquely mapped reads. It was shown that over 700 genes were differentially expressed in ‘Sun Chu Sha’ mandanrin due to the treatment of CLas-flg22. In contrast, only 1 gene is differentially expressed in ‘Duncan’ grapefruit. Interestingly, in our other project comparing the effect of flg22 of citrus canker causal bacteria Xanthomonas citrus subsp. citri (Xcc-flg22) showed that a much more extensive transcript reprogramming was triggered in both ‘Sun Chu Sha’ (over 2600 genes) and ‘Duncan’ (over 1300 genes), suggesting a weaker defense eliciting ability of CLas-flg22 than Xcc-flg22’s. Functional analysis of CLas-flg22 affected genes in the two citrus genotypes is underway.
The objectives of this proposal are 1) to conduct a statewide survey of tangerine and tangerine hybrid groves to determine the proportion of strobilurin resistant Alternaria alternata isolates along with the identification and characterization of resistance-causing mutations; 2) establish the baseline sensitivity of Alternaria alternata to the SDHI class fungicide, boscalid and characterize field or laboratory SDHI resistant mutants to determine the likelihood of SDHI resistance development in Florida tangerine production and 3) Develop an accurate and rapid assay to evaluate sensitivity to DMI fungicides. During this quarter we accomplished: ‘ Baseline sensitivity of Alternaria alternata population to boscalid manuscript is on-going ‘ Structure of SDH-subunits A,B,C, and D has been determined in baseline population ‘ Manuscript “QoI-resistance stability in relation to pathogenic and saprophytic fitness components of Alternaria alternata from citrus” was submitted ‘ Preliminary planning for adapting rezasurin assay for use with DMI fungicides underway
In collaboration with Mike Irey, we completed evaluation of a population of transgenic lines for their resistance to Clas. Transgenic trees were kept in a greenhouse containing CLas+ Asian Citrus psyllid (ACP) adults and trees were evaluated every 6 months for HLB for 2.5 years. In addition, psyllids were tested periodically for the presence of CLas. Although we had transgenic trees from four transgenes (LIMA, AttacinE, CEAD and NPR1) showing resistance to HLB, the results from NPR1 were the most promising. 27% of the trees containing the 35S-NPR1 construct and 57% of trees with the phloem specific AtSUC2-NPR1 construct survived and were PCR- after 30 months of inoculation. The HLB-free trees were moved to the SG field site under MTA and DPI petition (many in poor condition due to severe psyllid damage). Additional clones of the promising transgenic lines are currently being propagated for additional field testing in the Dunwoody Grove of Southern Gardens. In addition, Several newer NPR1 lines are already in the field test at the Picos Farms (USDA) and results there are promising as well. Our current goal is to combine transgenes that function by completely different mechanisms as to have a back-up to prevent the pathogen from overcoming single gene resistance in the field. We have produced 40 new transgenic plants of Hamlin cominging NPR1 with the successful AMP genes CEME or CEMA. We have successfully developed a heat inducible Cre/loxP site specific based recombination system for efficient excision of antibiotic resistance genes in citrus. In our construct, the nptII gene under the control of the NOS promoter and the Cre recombinase gene driven by either a soybean heat shock protein (hsp17.5E) promoter or a Arabidopsis thaliana small heat-shock protein (HSP20) gene promoter were flanked by two loxP recognition sites in direct orientation. An anthocyanin biosynthesis gene from Vitis vinifera (VvMYBA1) was placed outside the loxP sequence. Transformation efficiencies were similar using either soybean or the Arabidopsis promoter. Anthocyanin activity analysis on transformed Carrizo citrange (Citrus sinensis x Poncirus trifoliata) demonstrated that approximately 30-40% of transformation efficiency could be obtained following Agrobacterium mediated transformation and heat shock treatment. Molecular analyses have demonstrated that 100% selectable marker gene deletion occurred in all regenerated plants expressing anthocyanin. We have completed building a RES type structure in our greenhouse#7. Transgenic trees have reached the top of the structure and have been bent downwards. We did not observe any flowering this past year; however, downward growing branches on many of the trees have completely lost their thorns, indicating a rapid reduction in juvenility. Since this greenhouse is heated during the winter, flowering induction may be inhibited.
Transformations of citrus plants with the FLT-antiNodT fusion protein expression construct are now underway. The transformations are being performed at the Citrus Transformation Facility at the University of Florida Citrus Research and Education Center at Lake Alfred, FL. The FLT-antiNodT expression cassette is being introduced into ‘Duncan’ grapefruit by Agrobacterium tumefaciens – mediated transformation. Transformations were begun on November 5, 2013. The transformation construct includes a green fluorescent protein (GFP) marker. Within a few weeks of initiating the transformation experiment, clusters of green-fluorescing citrus cells were observed, indicating that the FLT-antiNodT fusion protein transformation vector was working. This was important, because we had encountered significant difficulties in the development of the FLT-antiNodT fusion protein transformation vector. All indications are that the transformation vector is working as expected, which is good news. Transgenic shoots were successfully regenerated that are green-fluorescing and presumably carry the FLT-antiNodT fusion protein expression cassette in their genome. As of December 13, 2013, micro-grafting of transformed shoots onto recipient plants was begun. Successful regeneration of shoots containing the FLT-antiNodT fusion protein expression cassette is critically important. Sometimes certain proteins and transgenes might interfere with the normal plant biology in such a way that regeneration of normal plant tissues cannot occur. We are pleased to report that this has not been a problem so far for this project. All indications are that we should be able to obtain transgenic plants suitable for testing for HLB resistance. This project has gone a bit more slowly than initially anticipated. All steps of the project took more time than initially predicted, from antibody development and cloning to the development of the transformation construct. It is likely that we will take at least until August, 2014, to obtain transformed plants that are suitable for HLB resistance testing. We will also need time to do laboratory experiments characterizing the expression of the FLT-antiNodT fusion protein in the transgenic plants. Therefore, we fully anticipate needing to request a minimum of one year no-cost extension to this project. A no-cost extension of the project will be necessary to achieve the goals of the project fully.
This research is directed at determining the activity of Non-antibiotic tetracycline analogs against the causative agent of HLB. Through chemical changes within the family of tetracycline derivatives it has been determined that specific derivatives harbor increased potency against surrogate strains of Liberibacter used to assess antibioitic compounds in vitro. In a several series of derivatives analogs were found to possess 2 to 3 orders of activity against the test strain compared to oxytetracycline, and EPA registered antimicrobial agent, and a 1st generation tetracycline useful against Liberibacter and other alpha-proteobacteria. In these studies, conducted at the University of Florida, separate compound series have emerged, with limited or no activity against human Gram positive and negative pathogens, allowing their classification as Non-antibiotic compounds. Additionally, clinically used tetracyclines, doxycycline and minocycline, had no activity against the surrogate strain, delineating specific structure-activity properties in these non-antibiotic tetracyclines. Other relationships exist between potency and activity in vitro, and further synthetic studies will be done to further increase activity against the causative agent of HLB. Other studies using non-antibiotic tetracyclines have demonstrated that specific compounds and series are able to be formulated using specific systems for delivery and application to the tree bark systems, showing enhanced penetration and distribution throughout the citrus tree via phloem transport. In future studies 2 compounds have been readied for field trials, with current enhanced formulations, and other and more potent compounds will be synthesized and examined for enhanced anti-HLB activity.
This is a continuing project to find economical approaches to citrus production in the presence of Huanglongbing (HLB). We are developing trees to be resistant or tolerant to the disease or to effectively repel the psyllid. First, we are attempting to identify genes that when expressed in citrus will control the greening bacterium or the psyllid. Secondly, we will express those genes in citrus. We are using two approaches. For the long term, these genes are being expressed in transgenic trees. However, because transgenic trees likely will not be available soon enough, we have developed the CTV vector as an interim approach to allow the industry to survive until resistant or tolerant trees are available. A major goal is to develop approaches that will allow young trees in the presence of HLB inoculum to grow to profitability. We also are using the CTV vector to express anti-HLB genes to treat trees in the field already infected with HLB. We have modified the CTV vector to produce higher levels of gene products to be screened. At this time we are continuing to screen possible peptide candidates in our psyllid containment room. We are now screening about 80 different genes or sequences for activity against HLB. We are starting to test the effect of two peptides or sequences in combination. We are attempting to develop methods to be able to screen genes faster. We are also working with other groups to screen possible compounds against psyllids on citrus. Several of these constructs use RNAi approaches to control psyllids. Preliminary results suggest that the RNAi approach against psyllids will work. We are screening a large number of transgenic plants for other labs. We are beginning to work with a team of researchers from the University of California Davis and Riverside campuses to express bacterial genes thought to possibly control Las. Since we are testing about 80 genes for induction of resistance or tolerance to HLB in citrus, we changing our focus of building new constructs to controlling psyllids until we have more conclusion from the peptides under screen. We recently examined all of the peptides constructs for stability. The earliest constructs have been in plants for about nine years. Almost all of the constructs still retain the peptide sequences. A recent advance is that has greatly speeded up our screen is that we now can estimate when plants become infected with HLB and can tell whether a peptide is working more quickly.
Transformations of citrus plants with the FLT-antiNodT fusion protein expression construct are continuing at the Citrus Transformation Facility at the University of Florida Citrus Research and Education Center at Lake Alfred, FL. The FLT-antiNodT expression cassette has been introduced into ‘Duncan’ grapefruit by Agrobacterium tumefaciens – mediated transformation. Plants resistant to the kanamycin selection marker and expressing the green fluorescent protein have been regenerated successfully into plantlets. Plants are developing and appear to be healthy and normal, without any deleterious effects of the transgene noticed in the plant regeneration process so far. We anticipate that 1-foot tall plants will be available to begin testing as early as August of 2014. We are requesting a no-cost extension to the end of 2014 to have enough time to test the HLB resistance status of the trangenic plants.
This project is focused on the optimal deployment of a superinfecting Citrus tristeza virus (CTV)-based vector as a tool to be used in the field to prevent existing field trees from the development of the HLB disease and to treat trees that already established the disease. In order to provide protection against HLB, the superinfecting CTV vector will be carrying an anti-HLB gene (i.e. a gene of an effective antimicrobial peptide). The majority of trees in Florida are already infected with some CTV isolates. The main question for us is how these pre-existing isolates would affect the establishment of infection with the superinfecting vector and, thus, expression and the production level of an anti-HLB polyprotein. Several sets of experiments in which we are examining how preexisting infection with different CTV strains affects the ability of the superinfecting CTV vector to infect and get established in the same trees are ongoing. We are also examining the levels of multiplication of the superinfecting CTV vector in trees infected with different field isolates of CTV. We first graft-inoculated sweet orange trees with the T36,T30 and/or T68 isolate of CTV, singly or in mixtures (these isolates were propagated in our greenhouse) as well as with CTV-infected material obtained from the field trees (FS series isolates). Real time PCR analysis protocol is being optimized for quantification of multiplication of CTV genotypes in the inoculated trees. Trees with developed CTV infection along with uninfected control trees were challenged by graft-inoculation with the superinfecting vector carrying a GFP gene. The latter protein is used as a marker protein in this assay, which production represents a measure of vector multiplication. The trees are now being examined to evaluate level of replication of superinfecting virus. Tissue samples from the challenged trees are observed under the fluorescence microscope to evaluate the ability of the vector to superinfect trees that were earlier infected with the other isolates of the virus. Levels of GFP fluorescence are monitored and compared between samples from trees with and without preexisting CTV infection. Real time PCR quantification is also being employed to these tests. In these experiments we are using different citrus rootstock/scion combinations in order to find combinations that would support the highest levels of superinfecting vector multiplication and thus, highest levels of expression of the anti-HLB protein of interest from this vector. These combinations include trees of Valencia and Hamlin sweet oranges and Duncan and Ruby Red grapefruit on three different rootstocks: Swingle citrumelo, Carrizo citrange, and Citrus macrophylla.
(863) 956-8817
(863) 956-5894
700 Experiment Station Road
Lake Alfred, FL 33850
Email Us
