EConomiCal EnviRonmEnTal SamplER DESignS FoR DETECTing .

flow and allow use with a 12-volt, 35 amp-hr gel cell battery (Interstate Batteries, catalog no.DCM0035, 80/unit). Total cost of the Active Sampler is 200/unit. The sampler is mountedhorizontally on a vertical section of ¾ inch electrical conduit with hose clamps, or attached to atree trunk or branch with a bungee cord. A standard 1 inch X 3 inch microscope slide (Catalogno. 12-550-343, Fisherbrand frosted microscope slide, Fisher Scientific, Thermofisher Scientific,Waltham, MA) is coated with a thin layer of silicone grease (Catalog no. 335148, vacuumgrease, Beckman Coulter, Brea, CA) with a cotton swab, covered with a strip of cellophane tape(Scotch 810 Magic Tape) so that the writing block on the slide remains exposed, and thencoated again with a second layer of grease. The first coat of grease makes removal of thecellophane tape easier for DNA extraction, while the second coat creates a sticky surface thatcaptures fine particulates that are pulled over the surface of the slide by the 12-volt fan.Location and date information can be written on the exposed writing block. The slide is thenplaced at a 45 angle inside the cutout end of the ABS pipe (Figure 1) and held in place byfriction so that airflow produced by the fan passes over the greased side of the slide and outthe back of the sampler. By mounting the sampler horizontally rather than vertically, the ABSpipe provides the greased slide some protection from rain.The sampler can be powered with a 12-volt, 35 amp-hr battery for up to seven days withoutbeing recharged. Battery life can be extended by connecting a small 10 watt, 12-volt solar panel(Catalog no. SLND 00542, Solarland, Ontario, CA) to the battery and mounting it on theelectrical conduit beneath the sampler with a panel mount (Catalog no. COLO 00688, Solarland,Ontario, CA). Since airflow generated by the fan is constant, total volume of air passing overthe slide can be quantified for any desired period of time by measuring air flow through thesampler (Figure 2).Passive Environmental SamplerPassive Environmental Samplers were designed to hold up to four greased microscope slides.Slides are supported in a small wooden box that holds the greased slides at a 45 angle. Thebox fits into a short section of 8-inch diameter galvanized ductwork that is attached to a sheetmetal reducer (8–7 inches) (Figure 3). The ductwork assembly is supported by a 5/16-inchthreaded rod that passes top to bottom through the sheet metal reducer. To allow the samplerto spin freely in the wind stream, the threaded rod passes through two 3/8-inch T-Nuts that areheld in place by 5/16-inch nuts with a nylon washer (Figure 4). A sheet metal fin attached tothe sheet metal reducer at the back of the assembly and secured in place with aluminumflashing allows it to rotate and face into the prevailing wind, eliminating the need for anexternal power source. Two versions of the sampler can be constructed, depending onprevailing wind speeds. For light winds, the sampler can be supported on a 5/16-inch threadedrod and made with an unreinforced sheet metal fin to make it more responsive to minorchanges in wind direction and velocity. For windy locations, the sampler can be supported on a3/8-inch threaded rod with ½-inch T-Nuts, and made with a fin that is reinforced with a strip ofheavier gauge metal to reduce flexing and metal fatigue. Since Passive Environmental Samplersare non-quantitative and rely entirely on prevailing winds for airflow across the greased slides,they are most useful for detecting presence/absence of frass containing Ceratocystis over any3

given sampling period. Passive samplers can be constructed from materials available in localhardware stores for 50/unit. A full list of materials and description of how the sampler isassembled is available on request.Figure 1. Active Environmental Sampler. The Active Environmental Sampler is a modified CDC(Centers for Disease Control) Gravid Mosquito Trap that contains a single greased slide at thecut-out end of the sampler (arrow). The slide is mounted at a 45 angle and held in place byfriction with the sides of the tube. A 12-volt fan behind the slide pulls a steady stream of airover the sticky, greased surface.4

Figure 2. Quantification of air flow across the surface of a greased slide in an ActiveEnvironmental Sampler. Airflow through the sampler is measured with a hand-heldanemometer. Volume of air exposed to the surface of the greased slide (S) is calculated asheight (h) of the slide X slide width X airflow distance (d) in one sec. Total volume sampled inone hour was calculated by multiplying by 60 to obtain volume/min and multiplying again by 60to obtain volume/hr. Volume of air sampled by the Rotorod was calculated as described byFrenz and Elander (1996) and Frenz et al. (1996) where Volume (rod area) X (path diameter)X (motor speed) X 𝜋 X (time sampled).5

Figure 3. Passive Environmental Sampler. The sheet metal sampler (A) holds a wooden box (B)with four greased microscope slides that are mounted at a 45o angle. The sampler is placed atthe end of a 10 ft. section of ¾-inch electrical conduit so that it can rotate freely in theprevailing wind (A). As wind passes across the surface of the greased slides, fine particulatematter is captured on their sticky, greased surfaces.Figure 4. The Passive Environmental Sampler is constructed around an 8 inch to 7-inch sheetmetal reducer (A) that is supported by a 5/16-inch threaded rod that passes top to bottomthrough the 8-inch diameter end of the reducer. To allow the sampler to spin freely, thethreaded rod passes through two 3/8-inch T-Nuts (B) that are held in place by two 5/16-inchnuts with a lock washer (C). A nylon washer (arrow) is placed between the 5/16-inch nuts andthe T-Nut to reduce friction.6

Sampler DeploymentPassive Environmental Samplers can be deployed on vertical 10 ft. sections of ¾-inch metalelectrical conduit. The threaded rod that supports the Passive Environmental Sampler fits intothe top of the section of conduit and rests on a 5/16- or 3/8-inch nylon washer that issupported by a 1½-inch fender washer at the top of the conduit (Figure 3). The sampler is thenfree to spin full 360o and face into the prevailing wind. By making a simple modification insampler design to place the threaded rod in the center of the sampler and the threaded end ofa 5/16-inch turnbuckle at each end, the sampler can be suspended at different heights aboveground in the forest canopy (Figure 5). Active Environmental Samplers, by contrast, can beattached at any height on the conduit with hose clamps or on tree branches or trunks with anelastic bungee cord.Electrical conduit can be supported by guy lines and stakes or can be attached to a metal orwooden fence post with heavy duty plastic zip ties or metal clamps designed for electricalconduit. Passive Environmental Samplers can be lifted into position or removed from the top ofthe electrical conduit with a “sampler picker” made from a 5 ft. section of metal conduit, withfour, 18-inch lengths of twisted, 14-gauge galvanized fence wire that are glued into the end ofthe conduit with construction adhesive and shaped to form an open basket that can support thesampler as it is lifted into position (Figure 6).Figure 5. Passive Environmental Sampler modified for deployment in the forest canopy. Notethat the sampler has been moved to the center of the threaded rod. Turnbuckles added to theends of the rod allow multiple samplers to be hung on the same line at different heights in theforest canopy.7

Figure 6. Passive Environmental Sampler “Picker”. The picker has four, 18-inch lengths oftwisted 14-gauge galvanized fence wire that are glued into the end of a section of conduit withconstruction adhesive and shaped to form an open basket that can support the sampler as it islifted on or off of a supporting pole.Sampler Evaluation Under Controlled Laboratory ConditionsSampling efficiency of Passive and Active Environmental Samplers was compared to acommercial sampling device (Rotorod Model 20) to determine number and range of particlesizes that each sampler can detect under relatively controlled conditions. Two 24-inch diameterelectric fans were placed on either side of the back of a 5.5 x 2.3 x 3 m room and oriented sothat wind flow was directed at a 45 angle against the rear wall. This allowed a circularcirculation pattern to form in the room with air currents traveling to the rear along the ceilingand returning to the fans along the lower half of the room. Combined wind speed in the centerof the room measured approximately 1 m/second with a hand-held anemometer. Since aircurrents differed slightly in speed and direction at different heights in the room, two, 2.4 m tallmasts constructed of ½-inch metal electrical conduit and supported horizontally by one sectionof ½ inch PVC (polyvinyl chloride) conduit were placed in the center of the room. Masts werefreestanding and supported by placing the ends into a ¾-inch PVC pipe that was immobilized8

with cement in 10 X 10 X 12-inch cinder blocks. Paired samplers (two Passive, two Active, andtwo Rotorod Model 20) were attached to the conduit adjacent to one another and rotatedbetween the three vertical positions (140, 180, and 220 cm) above the floor for each trial(Figure 7). The paired Active Environmental and Rotorod Samplers were powered by two 35amp/hr gel cell batteries (one Active Environmental and one Rotorod Model 20 per battery).Active Samplers held one greased microscope slide that was prepared as described earlier withScotch tape and silicone grease. Passive Environmental Samplers each held four greasedmicroscope slides. The Rotorod Model 20 Sampler held two square lucite rods, 20 mm inlength and 5 mm in diameter that were smeared with silicone grease. All slides and rods wereprepared inside a positive pressure PCR workstation (Airclean 600, Airclean Systems,Creedmoor, North Carolina) to reduce chances of extraneous contamination.Figure 7. Release of microspheres under controlled conditions in a closed room with circulatingairflow. Paired samplers were rotated between three vertical positions in a series of trials.Paired Rotorods (red box, bottom), Active Environmental Samplers (blue box middle), andPassive Environmental Samplers (purple box top) were rotated between three vertical positions(140, 180, 220 cm from floor) in a series of trials. Airflow was adjusted by releasing soapbubbles into the room.9

A 100 mg mix of polyethylene microspheres (Catalog Number CPMS-0.96, Cospheric LLC, SantaBarbara, California) ranging in size from 12–160 µm diameter (Mean 80.0 28.2 µm, Median 73.8 µm, Mode 60.9 µm) was carefully poured into a disposable 10 ml plastic serologicalpipette (Catalog no. 13-678-14A, Fisher Scientific, Thermofisher Scientific, Ipswich, MA) thatwas modified into a 10 mm diameter plastic pipe by removing the tapered tip from one end andthe cotton plug from the other. Microspheres were blown from the pipette into the rear of theroom with compressed air and allowed to circulate for each 4-hour trial. After each trial, slideswere removed from Active and Passive Environmental Samplers and stored in a closed slide boxuntil examined. Metal rod holders were removed from the Rotorod Samplers and stored inplastic bags until rods could be removed with tweezers and examined on a microscope slide.Paired samplers were rotated to a new vertical position, microscope slides and rods werereplaced, and trials were repeated until each pair of samplers was operated in each of the threevertical positions at least twice, for a total of eight trials.Slides and lucite rods were examined with a 10X objective through an Olympus BH50compound microscope (Olympus Life Sciences, Tokyo, Japan). Total number of captured beadsand diameter of up to 50 beads was measured on each slide or lucite rod with CellSenssoftware version 1.9 (Olympus Life Sciences, Tokyo, Japan). The flat surface of each squarelucite rod was examined, even though most particles were collected on the two leading sides ofthe rod. Particle counts were square root transformed to normalize data and reduce skewness.Variance in particle counts for each sampler type was compared with a Levene’s Test forequality of variance. Differences in particle counts between each of the three sampler types wascompared by one-way ANOVA. To determine if sampler types collected different sized particles,differences in variance were compared with a Levene’s Test for equality of variance and meanparticle diameters for each sampler type and a random sample of 1,000 microspheres from thestarting material were compared by one-way ANOVA.Sampler Evaluation Under Controlled Field ConditionsTo compare samplers under controlled field conditions, Xyleborus spp. frass was collected fromWaipunalei Ahupua a near the town of Laupahoehoe and Waiakea Forest Reserve on Hawai iIsland from January–February 2018. About 1.5 g of frass was collected into each of 14, 15 mlFalcon tubes (Catalog no. 05-527-90, Fisher Scientific, Thermofisher Scientific, Waltham, MA)from the lower trunk of symptomatic trees (Figure 8). To insure homogeneity, each tube offrass was transferred to individual Whirl-Pak bags (Catalog no. 01-812-6C, Fisher Scientific,Thermofisher Scientific, Waltham, MA), mixed thoroughly, and returned to their respectivecollection tubes. To test for Ceratocystis infection, 15 mg of frass from each Falcon tube wastransferred to 2-ml sterile screw cap tubes (Catalog no. 02-681-375, Fisher Scientific,Thermofisher Scientific, Waltham, MA) containing six, 3-mm zirconium beads (Catalog no.BAWZ 3000-300-23, OPS Diagnostics, Lebanon, NJ). Samples were extracted with NucleoSpinPlant II DNA extraction kit (Catalog no. 740770.250, Macherey-Nagel, Bethlehem, PA) followingthe manufacturer’s protocol. Briefly, frass samples were disrupted with 300 µl of PL2 buffer and10 µl of RNAse A (10 µg/µl) in 2.0 mL tubes with zirconium beads with a FastPrep 5G10

homogenizer (MP Biomedicals, Santa Ana, CA) for two, 60 sec intervals at 6.5 m/s. The sampleswere heated at 65 C for 10 minutes between the two homogenization steps. Other extractionsteps followed manufacturer recommendations for the NucleoSpin Plant II kit. DNA was elutedwith elution buffer provided in the kit in a final volume of 100 µl. Screening for C. lukuohia andC. huliohia by qPCR followed methodology described by Heller and Keith (2018).Eleven of 14 batches of frass were then soaked in 5 ml of 70% isopropyl alcohol for 30 minutesto kill Ceratocystis spp. propagules and then oven dried at 70 C for 48 hours. All frass sampleswere screened by qPCR for C. lukuohia and C. huliohia DNA (Heller and Keith 2018) aftertreatment with alcohol. About 30 mg of frass from each tube of treated and untreated frass wasalso carrot baited to check for fungal viability with a single carrot bait per tube (Moller andDeVay 1968).Figure 8. A. Frass (arrow) produced by ambrosia beetles during excavation of natal galleries ina tree infected with Ceratocystis. The frass is frequently found on and under bark near theentrance of galleries. B and C. Frass particles produced by Xyleborus spp. are rectangular orcylindrical in form and measure approximately 500 µm in length. Bar 500 µm.11

Release of beetle frass under controlled field conditions was done at facilities provided by U.S.Department of Agriculture, Agricultural Research Service, Daniel K. Inoye Pacific BasinAgricultural Research Center (USDA-ARS-DKI-PBARC) in Hilo. The experiment was conducted ina 6 x 3 x 15 m screened (1 mm mesh) outdoor cage with two “Heat Buster” fans (ModelQBX4223, Triangle Engineering, 6.1 amps ½ HP) placed at one end of the cage to create aconstant air flow from one end of the cage to the other. Samplers were set up so that tworeplicates for each sampler were run in each trial. Horizontal PVC piping was connected to threevertical support poles seven meters from the fans and 1.5 m from the ground. Samplers wereattached 0.5 m apart on the horizontal PVC pipe and arranged so that those on the left side ofthe cage were mirror images of those on the right side of the cage, e.g. Rotorod , Active,Passive on left and Passive, Active, Rotorod on right (Figure 9). Samplers were systematicallymoved between each trial so that samplers were operated in each location on the horizontalPVC pipe three times. Twelve 15-minute trials were carried out during a 6-hour period wherethe average wind speed was 1.2 m/s with a range of 0.7–2 m/s. For each trial, 300 mg ofpooled treated or untreated ambrosia beetle frass was shot into the wind column from behindthe fans using an air compressor and a 10 ml serological pipette as described for laboratorytrials. After each trial, slides and rods were collected, samplers were rotated, and new slidesand rods were replaced. A total of nine trials were conducted using frass treated with isopropylalcohol, for a total o

flow and allow use with a 12-volt, 35 amp-hr gel cell battery (Interstate Batteries, catalog no. DCM0035, 80/unit). Total cost of the Active Sampler is 200/unit. The sampler is mounted horizontally on a vertical section of ¾ i