Reducing tree decline of Casuarina equisetifolia in Guam through replacement of bacterial wilt infected trees and research into the bacterial microbiomes of trees and associated termites

Obj. 2 Yr 1-2: UH Res & UOG Res:  Research into the bacterial wilt pathogen to determine the origin of Guam’s infection and its genomic biology.

Methods: Obj. 2, University of Guam: research was conducted to find a method suitable for isolation of R. solanacearum from Guam’s ironwood trees. Though R. solanacearum could be detected from wood chips, drill shavings, water from root drill shavings, and stems and branches of trees, attempts to isolate from these same samples failed.  The only means by which Rs could be isolated was by streaking ooze that formed on disks taken from stems, roots, or large branches of infected trees onto selective medium. To enhance the production of ooze, slices were placed on a saturated paper-towel in a moisture chamber for 24 hrs. Once formed, the ooze was streaked on Engelbrecht’s semi-selective medium (mSMSA) (Engelbrech 1994). Colonies were re-streaked onto SMSA, which was followed by streaking onto modified Kelman’s tetrazolium chloride medium (TZC) before grow-out on TZC (Norman and Alvarez 1989). Once grow-out plates had well-isolated fluidal growth, a transfer loop or spatula was used to transfer colonies to 3 ml of sterilized tap water in a 6 ml sterilized screw cap glass vials. The inoculation process is repeated as necessary to create a cloudy suspension. Duplicate vials were prepared, one for storage at room temperature in Guam and one which was shipped to the University of Hawaii for further evaluation.

Information recorded at each tree site included: tree tag number, date of collection, GPS location, visual decline rating (DS), site condition, height, DBH, altitude, geology characteristics, occurrence, root exposure, and presence of termites and Ganoderma wood rot. Tree drill shavings were collected from a single 3-inch deep hole using a ¼ inch drill bit. Shavings were assayed for R. solanacearum using R. solanacearum-specific immunostrip tests kits manufactured by Agdia Inc. of Elkhart, Indiana, USA (catalog number: STX 33900 and ISK 33900).  Ralstonia solanacearum immunostrip testing was conducted on samples: either 1.5-3.0 mg of wood shavings or 80 µL of water from soaked drill shavings.

Methods: Obj. 2: University of Hawaii: water cultures from Guam were streaked multiple times onto TZC to ensure the cultures purity. A single colony from modified SMSA media was picked and mixed in 50 microliters of nuclease-free water. The colony was denatured for 10 minutes at 95^oC and centrifuged for two minutes. The colony was used as a template to do endpoint PCR with Ralstonia solanacearum species complex specific primers (developed during this research; Paudel et al, 2022). DNA extraction was performed with cultures found to be positive with PCR. The DNA was extracted using a Qiagen kit. The identity was further confirmed by sequencing dnaA gene region. Sequencing was done with both forward and reverse primers. To infer the diversity, evolutionary relationships and genealogy of the ironwood decline strains, three housekeeping genes (dnaA, gap, gyrB) and two virulence related genes (hrpB and egl) were used.

Engelbrecht, MC (1994). Modification of a semi‐ selective medium for the isolation and quantification of Pseudomonas solanacearum. ACIAR Bacterial Wilt Newsletter Vol. 10: 3-5

Norman D, Alvarez AM (1989). A rapid method for presumptive identification of Xanthomonas campestris pv. diffenbachiae and other xanthomonads. Plant Dis. 73: 654-658

Paudel S, Dobhal S, Lowe-Power T, Schlub RL, Hu J, Caitilyn A, Alvarez AM, Arif M (2022). “RSSC-Lineage Multiplex PCR” assay detects and differentiates Ralstonia solanacearum, R. pseudosolanacearum, R. syzygii and the R3bv2 subgroup. BioRxiv. https://doi.org/10.21203/rs.3.rs-1693987/v1

A more in-depth description of the materials and methods for objective 2 can be found in the thesis of Sujan Paudel, UH graduate student: http://hdl.handle.net/10125/70339

Obj. 3 Yr 1-2-3: UH Res., LSU Res., & UOG Res: Research into the flora of ironwood trees and the guts of termites.

   Obj. 3 Sub-obj. 1: UH Res. & UOG Res.: Research into the fungal flora of ironwood trees and their likely role in IWTD.

   Obj.3 Sub-obj. 2: UH Res. & UOG Res.: Research into the bacterial flora of ironwood trees and their likely role in IWTD.

Methods: Obj. 3: Sub-obj. 1 & Sub-obj. 2: University of Guam: In 2021, ironwood trees from four locations across Guam were surveyed for suitability for the bacterial and fungal microbiome study. The trees and their sites were evaluated for several characteristics. A tree was classified as damaged if it had been cut into, drilled into, or otherwise damaged by humans (microbiome possibly or likely compromised); and undamaged if it had never been cut into, drilled into, or otherwise damaged by humans (microbiome uncompromised). Damage caused by natural phenomena (storm damage, water damage, or any damage not caused by direct human contact with the tree) was not considered. Trees were also examined for the presence or absence of termites, and presence or absence of Ralstonia solanacearum. Effort was made to choose an equal number of R. solanacearum positive and R. solanacearum negative trees for inclusion into the study. The size and health status of the trees were not specific as the main objective was to identify 30 ironwood trees which were damaged by human activity and 10 ironwood trees which were undamaged by human activity.

The following materials were used to execute the collection of woody ironwood tissue for the University of Hawaii

• GPS (Garmin or Cell Phone)
• Flagging Tape
• Tree tags
• Copper Nails
• Insight LH Precision Laser Rangefinder with Hypsometer
• Measuring Tape
• Tree Log Data Sheet
• Machete
• Hammer
• Wide chisel (or scraper)
• Nitrile gloves
• 1/4 in and 5/16 in drill bits
• Cordless drill
• Agdia Rs Immunostrip test strips and buffer filled bags
• Paper funnels
• Glass vials
• Ziploc bags
• Cooler

At each site, healthy and unhealthy trees were tested for Rs (R. solanacearum) using Agdia immunostrip tests. Drill shavings were collected from beneath the bark layer with a ¼ inch drill bit approximately 1.5 inches into the trunk of the tree. Approximately 0.2 grams of drill shavings were placed in Agdia test buffer extraction pouch. Testing continued until approximately five Rs+ and five Rs- trees were identified at each site. In some situations, Rs+ trees were unable to be identified at one site, so additional Rs+ trees were identified at remaining sites to make up the difference. If the tree had been uncompromised before testing for Rs, the test drill hole was filled with a sterilized stainless-steel screw, thereby conserving the integrity of the microbiome of the tree. After trees for inclusion in the study were identified, the following tree information was collected: tree tag number, tree GPS, location, tree disease severity, level of human impact, height, diameter at breast height, presence or absence of sporocarps, site condition, altitude, termite activity, and suspected termite species (Nasutitermes sp., Cototermes sp., or Microcerotermes sp.)

Conditions in the field were kept as sterile as possible: nitrile gloves were worn when preparing the tree and collecting the samples to avoid human skin cells in samples, machete and chisel used to clear off the bark were cleaned with Clorox wipes, all drill bits were autoclaved or flamed in the field, and all vials were autoclaved and wrapped sterilely to avoid contamination in transport. All samples were stored in the refrigerator (4°C) until shipment. Samples were shipped out the same day they were collected and overnighted to the University of Hawaii.

To collect woody ironwood tissue samples for the microbiome analysis, a protocol was developed to collected both shallow (0-3 inches) and deep samples (4-7 inches).  At each depth, two combined samples were collected at 50 cm above the ground and on each side of the tree.  Prior to drilling, a clean Clorox-wiped scraper was used to scrape off outer bark. When the bark was nearly removed, the scraper was cleaned with Clorox wipe again. Scraper was then used to slice off additional layers until the bark layer was easily separated from sapwood (Figure 1). If the exposed tissue appeared to be living, samples of drill shaving were collected, if not, another location on the tree was chosen. 

Sampling began with the shallow woody tissue sample: drill shavings from a 3-in deep hole were collected using a 12 in by 1/4 in drill bit. At least 2 g (20 mL) of shavings were collected into a sterile glass vial by using a sterilized paper funnel (Figure 2). Vial with shavings was immediately placed on ice in the field for transport back to the lab. After the shallow woody tissue sample was collected, a 5/16 in by 12 in drill bit was used to enlarge the diameter of the hole and to increase its depth to 4 in.  This would minimize the shavings contaminating the deep sample. For the deep sample, a sterile 12 in by 1/4 in bit was used to collect a total of 20 mL of drill shavings from the same drill holes, but at a depth of 4-7 inches into the tree. Again, shavings were collected into a sterile glass vial and vial with shavings was immediately placed on ice in the field for transport back to the lab. It should be noted that the depth of the deep sample was not allowed to exceed the radius of the tree. In these cases, depth was noted in the comments section.

In the same area of the tree (50 cm above ground) where the bark was removed for microbiome samples, additional holes were drilled and shavings collected for conductivity measurements. Conductivity is a measure of water's capability to pass electrical flow. This ability is directly related to the concentration of ions in the water. It is a significant predictor of percent wetwood within a tree. Greater conductivity is associated with greater % wetwood. Drill shavings were collected to the depth of 4 inches on both sides of the tree. Multiple holes were drilled with a 5/16 drill bit until approximately 75 mL of shavings were collected into a clean ziplock bag. The bag was placed on ice and transported to the lab. Once at the lab, samples were grinded until small enough to pass through a window screen (square opening of 3 mm). Five grams of sieved sample was then placed in a clean glass vial and placed in the freezer for storage 1 to 90 days. Ground shavings, which were powder-like and allowed to thaw out if necessary, were poured into 50 mL of distilled water in a 125 mL flask. The mixture was placed on a hot plate, brought to a boil, boiled for 5 minutes, then removed. While hot, mixture was then immediately run through filter paper. Filtrate was allowed to cool to room temperature before measuring conductivity. Values were recorded as microSiemens per centimeter (μS/cm at 25 °C). Conductivity level was then classified using the value: 0= slight (<301); 1= low (301-574); 2= moderate (575-849); 3= high (>849).

Six of the forty trees were chosen to be cut down (felled) for collection of additional microbiome samples: root sample, 150 cm felled sample, 300 cm felled sample, and 450 cm felled sample. Criteria included trees being undamaged (never having been drilled into, cut, or otherwise damaged by human activity), and half Rs positive/ half Rs negative. Two roots were sampled for each tree. When possible, each root had to be at least 2 inches in diameter. For each root, the spot where the root starts to go underground was marked (this might be directly at the trunk base or a foot or more away). Starting from the mark, the root was uncovered for a distance of 15 inches. Soil was removed from the exposed root using water and a soft scrub brush. At the 12 in distance, a drill sampling area was created by removing the bark with a disinfected machete. Using a 1/4 in drill bit, drill shavings were collected to a depth of 2 inches into each root by using a sterilized weigh boat to catch the shavings. The combined shavings were transferred into a glass vial until a volume of 20 ml was reached. Vial was immediately placed on ice in the field. Additional shavings were collected in a ziploc bag and tested for Rs using an Agdia immunostrip test.

After collecting the root sample, the trees were cut down (felled). The outer bark was removed from one side of the tree at distances of 150 cm, 300 cm, and 450 cm from the tree’s base using a disinfected machete (Figure 3, Figure 4). At each distance, drill shavings were collected to a depth of 2 inches using a sterilized weigh boat (Figure 4). The shavings were transferred into a glass vial until a volume of 20 ml was reached and then placed on ice to transport to the lab.

Figure 1. An exposed area of sapwood used for microbiome sampling.

Figure 2. Using a 12 in long by 1/4 in wide drill bit, drill shavings were collected into a sterile glass vial by using a sterilized paper funnel.

Figure 3. Felled tree showing 150 cm, 300 cm, and 450 cm collection sites cleaned and bark removed.

Figure 4. The 150 cm felled sample is collected using a sterilized weight boat (two people in front), while the 300 cm felled sample collection area has bark removed to sapwood in preparation for sample collection (one person in back). 

Methods: Obj. 3, Sub-obj. 1 & Sub-obj. 2: University of Hawaii:

Genome sequencing of Ralstonia strains for evolutionary analyses: selecting ironwood strains from current and past Guam collections, their DNA was extracted from a half loopful of pure bacterial colonies, grown overnight on DPA (dextrose 5 g/L, peptone 10 g/L and agar 17 g/L) at 26 ± 2°C, by using the Genomic-tips 500/G kit (Qiagen) according to the manufacturer’s instruction. DNA concentration of the seven samples was measured using a Qubit 4 fluorometer (Thermo Fisher Scientific, Life Technologies, Carlsbad, CA); additionally, the DNA quality was assessed in a 1% agarose gel electrophoresis.

The isolated DNA were sent for illumine (NovaSeq) sequencing. Short-read sequencing of the seven Ralstonia strains was performed using Illumina NovaSeq system at the UC Davis Genome Center. DNA libraries were prepared using Seqwell plexWell LP384 Library Preparation Kit with 10 ng gDNA required for Illumina sequencing platform (seqWell, Beverly, MA). The prepared library was amplified with 8 PCR cycles, analyzed using Bioanalyzer 2100 (Agilent, Santa Clara, CA), and quantified by Qubit 4 Fluorometer Instrument (LifeTechnologies, Carlsbad, CA), and equimolar library pool was quantified and sequenced using qPCR with a Kapa Library- Quant kit (Kapa Biosystems/Roche, Basel Switzerland). Sequencing was run with paired-end 150 bp reads, using an and Illumina NovaSeq system. The sequences have been received.

The genomes were also sequenced using Oxford Nanopore MinION. DNA libraries were prepared using the Barcoding Kit (Oxford Nanopore Technologies Inc) with 400 ng of high-quality genomic DNA. The libraries were sequenced in a flow-cell version R10 and run for 72 hours. Sequencing was monitored in real time using the MinKNOWN software version 4.0.20 (Oxford Nanopore Technology). The generated FAST5 sequence files from Nanopore sequencing were base called using MinKNOWN software version 4.0.20 (Oxford Nanopore Technology). Hybrid assemblies were generated using both short- and long-reads, and evolutionary analyses were conducted. Core genome based evolutionary analyses was performed using ClonalFrame and fastGEAR software; total 17 genomes were used for analyses.

Bacterial flora Standardized DNA isolation method from ironwood plant: ironwood seeds sent by Dr. Robert Schlub in 2020 were sown for standardizing experiments and protocols. The critical and most important factor while sowing the ironwood seed was to place the seed horizontally 2-3 cm deep in the conical pots. The seeds germinate in 2-3 weeks depending upon the temperature condition. Ironwood seed germinated in 2 weeks under greenhouse conditions. The germination rate was lower comparatively to seeds sown in the month of January.

The modified protocol for DNA isolation from ironwood wood tissues was standardized.

DNA extraction from wood (Sapwood and heartwood) was quite difficult due to presence of higher quantity of secondary metabolites phenolic and lignin compounds. Therefore, after experimenting 8-10 different protocols we were able to finalize the protocol as follows:

Preparation of CTAB buffer (250ml):

1. NaCl (2.5M) – Weigh 14.625 gm and dissolve in 100 ml of autoclaved distilled water
2. Add Tris (1M) – 25 ml
3. Add EDTA (0.5M) – 10 ml
4. Add 5 gm of CTAB and stir until dissolved
5. Make the volume up to 250 ml with autoclaved distilled water
6. Adjust pH- 8.0 with conc. HCl
7. Preparation of 10% PVP. Add 10 gm of PVP and dissolve in 100 ml of distilled water in a beaker. Filter using 0.22 μM filter membrane. Take a syringe of 10 ml, intake 8-9 ml of prepared solution at a time, fit the tip of syringe to 0.22 μM filter membrane without touching the surface of membrane, adjust the nozzle of filter to 50 ml centrifuge tube and gently press the syringe. Stored at room temperature.

Modified protocol for DNA isolation from greenhouse ironwood plants:

1. Pre-heat 5 ml of CTAB buffer in water bath (65°C) in 50 ml centrifuge tube. Once it is heated, add 500 μl of 10% PVP and 10 μl of β-mercaptoethanol.
2. Weigh 500 gm of plant tissue and grind in an autoclaved mortar pestle using Liquid Nitrogen. Grind well and make it in a powder form.
3. Mix well the grinded tissue in the 5 ml of CTAB buffer prepared in step1.
4. Use a wide borer tip (make a cut of around 1 cm on 1 ml tip using blade) to pour 500 μl solution from step 3 into bead beater tubes.
5. Grind in the bead beater for 2 minutes at maximum speed.
6. Incubate at 65°C for 1 hour and 30 minutes in a water bath. Vortex the tubes for 5 seconds after a 20-minute interval.
7. Take out the tubes from the water bath and let it cool down for 2-3 minutes under.
8. Add 5 μl of RNase (100mg/ml) and mix well. Incubate the tube at 37°C and 140 rpm for one hour in room 312 incubator with shaker.
   • *Following steps to be done on DNA isolation workbench. Take aliquots of buffer solutions required for DNA isolation.
   • *Buffers used in the following steps are taken from Qiagen DNeasy Plant Mini Kit (Catalog no. 69104)
9. Add 130 μl of AP1, mix well with pipette and incubate in ice for 5 minutes.
10. Centrifuge the tubes at 20 X1000 g for 5 minutes.
11. Pipet the lysate into a QIAshredder spin column placed in a 2 ml collection tube.
12. Centrifuge for 2 min at 20 X1000 g.
13. Transfer the flow-through into a new tube without disturbing the pellet if present. Add 1.5 volumes of Buffer AW1 and mix by pipetting.
14. Transfer 650 μl of the mixture into a DNeasy Mini spin column placed in a 2 ml collection tube. Centrifuge for 1 min at ≥6000 x g (≥8000 rpm). Discard the flowthrough. Repeat this step with the remaining sample
15. Place the spin column into a new 2 ml collection tube. Add 500 μl Buffer AW2, and centrifuge for 1 min at ≥6000 x g. Discard the flow-through.
16. Add another 500 μl Buffer AW2. Centrifuge for 2 min at 20,000 x g. Note: Remove the spin column from the collection tube carefully so that the column does not come into contact with the flow-through.
17. Transfer the spin column to a new 1.5 ml or 2 ml microcentrifuge tube.
18. Add 25 μl Buffer AE for elution. Incubate for 5 min at room temperature (15–25°C).
19. Centrifuge for 1 min at 8000 g.
20. Repeat step 17
21. Store DNA in aliquots at -20°C. 

The current study focused on the collection of ironwood samples associated with ironwood decline, along with phenotypic metadata, and 16S rRNA (V3) and ITS (ITS1 and ITS2) amplicon-based microbiome analyses.  

This research involved a large-scale sampling (101) of ironwood tree samples from 5 different regions on Guam, a geographically isolated and unexplored island with unique climatic conditions, with the aim of resolving ironwood decline-Ralstonia disease complexity and revealing endophytic microbial interactions within the host system. The comparison of endophytic microbial community associated with different factors such as Ralstonia infection, root rot fungus, disease severity, termite activity, etc. considered in this study, will not only pave a way to resolve disease complexity but also will provide an insight into the complex interactions occurring among endophytic microorganisms sharing the same niche and their role in modulating pathogen response. Three characteristics were purposefully replicated, one being the detection of Ralstonia using immunostrips. Roughly half of the trees tested positive for Ralstonia and half tested negative. The other two characteristics replicated were location and tree disease severity.   

The samples from Guam were processed immediately for DNA isolation using the modified method. The first run was with an in house 16S primer set to check the quality. The DNA isolated from 101 samples was sent to the Microbial Genomics and Analytical Laboratory (MGAL) core facility at the University of Hawai'i at Mānoa for 16S and ITS library preparation. Community analysis was performed using EzBioCloud. The Kruskal-Wallis H test (α=0.05) was used for statistical comparison of alpha diversity identified by bacterial and fungal richness (total number of genus) and diversity index (Simpson’s index) associated with A) Ralstonia infection and B) sample type. 

A more in-depth description of the materials and methods for this objective can be found in the Diksha Klair, University of Hawaii Master of Science thesis: https://hdl.handle.net/10125/103907

Obj. 3: UH Res., LSU Res., & UOG Res: Research into the flora of ironwood trees and the guts of termites.

   Obj.3 Sub-obj. 3: LSU Res. & UOG Res.: Research to determine if termites carry ironwood bacteria and thus, might be responsible for their movement.

Methods: Obj. 3, Sub-obj. 3, University of Guam:

Termite Collection Methods: from 2019-2021 termites were collected on the island of Guam. For each ironwood tree with an active infestation, general data was recorded, including: Tree tag number, date of collection, GPS location, visual rating, site condition, height, DBH, altitude, geology characteristics, occurrence, root exposure, collection area of termite ( i.e. mud trail, above or below ground colonies), and presence or absence of Ralstonia solanacearum (tested using Agdia Immunostrip buffer test pouch). A minimum of 21 active termites (15 workers/6 soldiers) per tree were collected directly from the infestation site with a general aspirator and immediately transferred into vials. For each tree, the samples including soldier and worker termites which were partitioned into 10 mL of 70% ethanol (for morphological identification) and 10 mL of 95% ethanol (for Illumina sequencing). Each individual vial was labeled by tree number in the field, placed in a vial box, then placed on ice in a cooler. The vials were further processed through labeling by clinic number, then in a cardboard vial box with the lid closed and placed in the laboratory freezer.

Termite-sampled ironwood trees were tested for Ralstonia solanacearum by selecting two locations on the stem breast height above the soil line. An electric drill with a 1.5 in. sterilized drill bit was used to create the initial break in the periderm, then a 5/32 in. sterilized drill bit was used to collect drill shavings by drilling a 4 cm deep hole in the tree. The shavings were collected in plastic bags and subsequently labeled. To reduce the chances of secondary infection by other insects or pathogens, the two holes were plugged with a sterilized Flat Head #12 (7.32) stainless steel screw. The shavings were then brought back to the University of Guam Cooperative Extension Plant Pathology Lab and tested for Rs using the Agdia® Inc. Ralstonia solanacearum (Rs) immunodiagnostic strips. The test was performed on the same day the samples were collected. Following the manufacturer’s instructions, 0.15 g of the drill shavings were placed in BEB1 sample extraction bags. The sample was allowed to set three minutes before inserting the immunostrip.

Drill shavings were collected for conductivity determinations. In the case of LSU, the bark was removed just to the point that the live tissue layer appeared (phloem) before being drilled and shavings collected. In the lab, tissue samples were grinded and sieved through 1/8 in galvanized hardware cloth. Five grams of sieved tissue was then placed in a clean glass vial and placed in the freezer for storage. Processing method included the following: the entire 5 g of sieved tissue was transferred into a 125 mL flask containing 50 mL of steam-distilled water, shaken, and placed on a hot plate. At first sign of boiling (approximately 2 minutes later), mixture was boil rapidly for an additional 5 minutes. While hot, using a Vortex-genie 2, the sample was vortexed vigorously for one minute at the medium setting. Afterwards, it was placed in an ice bath and cooled to 60°C, then filtered through Whatman #3 paper by pulling a vacuum. After the filtrate cooled to room temperature, its conductivity was measured using a Hach 51800-10 sensION 5 Waterproof Conductivity Meter and recorded as microSiemens per centimeter (μS/cm at 25 °C). Conductivity level was then classified using the value: 0= slight (<301); 1= low (301-574); 2= moderate (575-849); 3= high (>849).

Soil Collection Methods: Using the same ironwood trees that termites had been collected from, in 2022 five core soil samples were collected per tree and combined for processing. Using a clean, disinfected handheld soil core sampler with an inner diameter of ¾ inch, five soil cores were collected into a labeled zip lock bag. The collection of five samples was roughly equally spaced around the tree and 3-5 ft from the tree base. Without disturbing the underlying soil, any grass, weeds, “needles”, and other debris were removed so that the soil sampling area would be bare. The soil core was pushed as deep as possible into the soil but no deeper than 10 cm. Between trees, the core sampler was wiped clean and disinfected in 10% Clorox. All tools were either autoclaved or thoroughly cleaned with a 10% solution of Clorox. To eliminate DNA contamination of soil samples for ETOH preservation, nitrile gloves were worn, and between tree samples gloves were either cleaned with Clorox wipe or a new pair used.

Once the soil was collected into the labeled zip lock bag, the soil was immediately mixed by shaking the bag. Transferred from the bag using a sterilized metal spoon, small aliquots of soil were sieved through a ¼ in galvanized steel hardware cloth (screen) into an autoclaved paper cup. By folding the edge of the paper cup, small amounts of sieved soil were then poured into a pre-weighted and labeled 15 ml plastic ultra-high performance centrifuge tube containing 8 ml of 95% ETOH. Soil was added until a total volume of 11 ml was reached. To standardize conditions and maximize mixing, the tube was shaken frequently during the process. The tubes were immediately placed on ice and transported back to the lab. In the lab, the tubes were re-weighed to determine grams of field soil added to each tube. Afterward, sample tubes were stored in a freezer until shipment to LSU.

Also taken back to the lab was the leftover unsieved soil in the zip lock bag (approx. 80 ml), for soil-moisture determination (no need to refrigerate). In the lab, this soil was sieved with the same sieve type used for the centrifuge sample soil. Approximately 50 g of sieved field soil from each tree was placed in a weighed aluminum pan and dried at 105°C for at least 48 hrs. Dish was removed from the oven and allowed to cool. Once cooled, the dish was re-weighed and the difference from the pre-drying weight was calculated, which calculates the weight of the water in the original sample. This information was used to calculate soil moisture (dry wt. basis). Percent moisture content of soil is equal to weight of water in soil sample divided by the weight of dry soil.

Spiking Soil Methods: As a means to determine the suitability of the LSU soil sampling procedure to detect Ralstonia and various wetwood bacteria in a soil sample, 6 soil samples were spiked with ooze from slices of tissue from selected trees (positive for R. solanacearum species complex (RSSC) and Nasutitermes takasagoensis (Nt)). To determine the impact of soil on the detection of bacteria in the ooze, straight ooze was also sent. Similar to the soil collection procedures, precautions were taken to keep things clean. This included wearing Nitrile gloves and disinfecting with 10% Clorox solution or Clorox wipes.

Trees from which termite and soil samples were previously collected for LSU and which were positive for R. solanacearum species complex (RSSC) and Nasutitermes takasagoensis (Nt), were processed and ooze collected. In the field, slices of wood from the tree were collected and placed into disinfected plastic tubs containing sterile water and a few sheets of sterile paper towels (roughly 6). Water level was at least 1 cm deep and did not exceed ½ the thickness of the wood slice. To reduce evaporation, tubs were covered with lids or plastic wrap. To increase the odds of collecting an adequate amount of ooze, several slices were collected. Slices ranged in thickness, from 1 to 4 inches. Slices were collected mainly from trunks; however, large lower branches and/or roots were occasionally sampled. At the time of cutting, slices were scraped clean of most of the saw dust debris using a disinfected machete. The chain saw was cleaned between trees by operating in a bucket of water and then 10% Clorox solution for roughly 30 seconds. Slices were either placed in the lab or outside the lab on a table in the shade.

At the same time tree slice samples were collected, a soil sample consisting of 5 soil cores was collected in a zip-lock bag using the same procedure as outlined previously. This sample was kept out of direct sunlight and transported back to the lab, where it was stored on the lab bench until ooze was collected.

The tree slices were examined over the next 2 days. If total ooze production was deemed inadequate (less than 0.5 ml), slices might be left for additional time. In some cases, additional wood slices were collected from the same tree. In such a case, all the slices would be processed together after 72 hrs from when the first slices were collected. Ooze was collected using a vacuum pump fitted with a 1 ml pipet tip. The aspirate was collected in a centrifuge tube that had been cut to size and placed in the vacuum flask. Areas of ooze production on the cut sections were vacuumed until 3 ml of ooze was collected. If aspirate was less than 3 ml, drops of sterile distilled water were added to ooze production areas and the vacuuming process continued until 3 ml was reached. After ooze was collected, the presence of Rs was confirmed by testing the combined scrapings from all the pieces with an Agdia Rs-specific Immunostrip test. From the 3 ml of ooze suspension, a 0.06 ml subsample was removed and place it in a 0.5 ml Eppendorf tube containing 0.44 ml of 95% ETOH.

The soil sample that was set aside on the lab bench at the time the slices were collected was sieved using the same procedure as before (refer to soil collection methods). While avoiding contamination with skin cells, 10 g of sieved soil was placed into a new zip-lock bag. The remaining ooze suspension after taking the subsample (2.94 ml) was thoroughly mixed with the soil and then that spiked soil was poured into a pre-weighted and labeled 15 ml plastic ultra-high performance centrifuge tube containing 8 ml of 95% ETOH and stored in a freezer until shipment to LSU.

Brief Methods: Obj.3, Sub-obj. 3, Louisiana State University (termite analysis):

Methods were devised for meta-analysis to answer the hypothesized question: Are Nasutitermes takasagoensis, Coptotermes gestroi and Microcerotermes crassus, which represent the major termite species associated with IWTD, vectors for Ralstonia and other pathogenic bacteria causing IWTD? 

To test this hypothesis, it was essential to collect termite samples in equal numbers that tested negative or positive for Ralstonia in Guam as this would service as the replicated set in order to (1) describe the bacterial taxa associated with workers attacking ironwood trees in Guam to test if termites carry plant pathogens, (2) test for relations between the tree-, plot-, and location-related factors associated with ironwood trees attacked by workers and microbial diversity of those worker samples, (3) determine if termites prefer feeding on parts of ironwood trees with low pathogen content compared to high pathogen content, and (4) determine whether R. solanacearum bacteria are ingested and survive in the termite gut. The Quantitative Insights into Microbial Ecology (QIIME2) pipeline (Caporaso et al. 2010, Estaki et al. 2020) version 2021-4, accessible on a server of the Hubbard Center for Genome Studies at the University of New Hampshire, was used to perform sequence data analysis. Demultiplexed sequencing reads were obtained from University of New Hampshire after Illumina sequencing in FASTQ format. 

Forty-five termite samples were collected in 2019-20 by the University of Guam from healthy as well as sick ironwood trees present on 14 distinct locations on the island of Guam. Geology related (location, parent material classification, site management), tree-related (tree DS, tree health, presence or absence of Ralstonia, altitude classification) and plot-related (plot average DS, plot average health, stand maturity estimate, percentage of trees with termites in the plot, percentage of dead trees in the plot) parameters were recorded. The samples including soldiers and worker termites were partitioned into 70% ethanol (for morphological identification) and 95% ethanol (for Illumina sequencing) and were shipped to Louisiana State University. In the attached report by LSU, a full description of the materials and methods for the following topics can be found: (1) description of the factors collected from ironwood tree plots by University of Guam; (2) Morphological species identification of termites; (3) DNA extraction; (4) Primer selection; (5) DNA amplification and sequencing; (6) Bioinformatics analysis; (7) termite feeding experiments. 

A more in-depth description of LSU’s methods for years 1-2 can be found in their progress report: LSU WSARE R&E Progress Report 4.1.21 - 3.30.22

A more in-depth description of LSU’s methods for year 3 can be found in the thesis of Garima Setia, LSU graduate student: https://digitalcommons.lsu.edu/gradschool_theses/5699/

Methods: Obj.3, Sub-obj. 3, Louisiana State University (soil analysis): A portion of each soil sample was first taken from the ethanol suspension and dried on a clean weighing boat. For each sample 250mg was used for DNA extraction with Qiagen DNeasy PowerSoil Pro Kit. Briefly, the samples were transferred into the PowerBead Pro Tubes with lysis buffer (CD 1) and homogenized by Benchmark Scientific BeadBlaster Tissue Homogenizer. The suspension was then processed and purified following the kit handbook. The DNA concentrations were quantified by Invitrogen Qubit 4 Fluorometer with dsDNA BR assay kit and 2.5 µl/ng DNA per sample was shipped to the University of New Hampshire for next-generation sequencing. The V1-V3 hyper variable region of the bacterial 16S rRNA gene was amplified from the DNA samples using one forward (27F) and two reverse primers (519Rmod and 519Rmodbio) to capture a broad range of biodiversity. QIIME2 was used to analyze the demultiplexed sequencing reads obtained from UNH after Illumina sequencing accessible on a server of their Hubbard Center for Genome Studies. Forward reads were subjected to further processing with removal of primers sequences and truncation to 251 nucleotides by DADA2. The rarefaction, alpha diversity (faith-pd, ASV richness, evenness and Shannon) and beta diversity (permanova and permdisp) were then conducted according to the tree-related factors and the visualization was completed by QIIME2 View and R package iNEXT. Taxonomical assignment was performed by comparing each sequence to the SILVA reference database.