The Hawaii Protocol for Scientific Monitoring of Coffee Berry Borer: a Model for Coffee Agroecosystems Worldwide

Note: A Spanish translation of the protocol is provided as Supplementary File 1.

1. Define Sampling Zones within Coffee Fields

1. Survey the perimeter of the coffee field to be monitored using a global positioning system (GPS) instrument. Import the field coordinates into a global information system (GIS) and generate a map of the coffee field.
2. Divide the field into "zones" (i.e., polygons), each about 335 m². These will be used to ensure a systematic random sampling design across the field.

2. Create a Data Collection Application in an Electronic System

1. Using an electronic data collection platform, build a data collection application comprised of the following interlinked databases: Traps, Zones, Site Service, Weather Stations, and Management.
   Note: These databases will be used in all subsequent steps of the protocol for the collection and organization of data.
2. For the Traps database, create fields for 'site name', 'trap number', 'deployment date', 'field technician name', 'deployment photo', and a link to the GPS coordinates for each trap.
3. For the Zones database, create fields for 'site name' and 'zone number', with a link to the georeferenced site map displaying each zone.
4. For the Site Service database, create fields for 'site name', 'date', 'field technician name', and 'site notes'. Within the Site Service database create nested databases comprised of the following.
   1. Include Trap Service to record trap number (with a link to relevant trap deployment record in the parent Traps database), photograph of trap catch, and trap count.
   2. Include Zone Service to record phenology photographs, berry infestation assessment (total number of green berries, infested green berries, green berries with B. bassiana, and raisins), and a link to the relevant zone record in the parent Zones database; this record also includes GPS coordinates for each sampled tree.
   3. Include Weather Station Service to record site name, date, data download, and battery check.
   4. Include Berry Dissection to record lab technician name, date, and CBB position (AB or CD) and mortality category (alive, dead by other causes, or dead by Beauveria bassiana) for each dissected berry.
5. For the Weather Stations database, create fields for 'site name', 'station number', 'deployment date',' field technician name', 'deployment photo', and a link to the GPS coordinates for each weather station.
6. For the Management database, create fields for 'site name', 'date', and 'type of management practice'.

3. Prepare and Deploy Traps for Monitoring CBB Movement

1. Determine the number of traps needed to monitor CBB movement in each field.
   NOTE: Trap density per field should approximate 5 traps for small fields (~0.5 ha) and 10 traps for large fields (~1 ha)¹⁵.
2. Using a thumbtack, make a series of drainage holes above the fill line in each trap collection cup to avoid dilution of the kill solution by rainwater. Assemble the funnel traps according to the manufacturer's instructions.
3. Prepare 1 L of kill solution comprised of 200 mL of propylene glycol and 800 mL of water. Next, prepare an attractant mixture comprised of a 3:1 solution of methanol:ethanol. Pour 40 mL of the attractant into plastic semi-permeable bags (2 mil, 3 inch x 4 inch) and place in a container for transport.
   NOTE: Semi-permeable bags have been shown to perform better than open vials in attracting CBB, and also require less frequent visits to refill lures due to lower elution rates¹⁶.
   CAUTION: Methanol and ethanol are highly flammable liquids, are toxic if inhaled or ingested, and are skin and eye irritants. These chemicals should be handled in a well-ventilated room while wearing gloves, eye protection, and protective clothing.
4. Deploy the traps by randomly distributing them across the field. Place traps 0.5 – 1.5 m above the ground, and clear of the aisles. Stakes may be used to effectively secure traps between trees. Write the site name and trap number with permanent marker on each trap for future identification.
5. Fill trap collection cups with 100 mL of the glycol kill solution and screw the cups tightly into place. Attach a paper clip to each attractant bag and use the paper clip to hook the bag to the center of the trap.
6. Using a mobile device equipped with the electronic data collection platform, navigate to the Traps database and create a new deployment record comprised of the site, date, trap number, and a photograph of the trap.
   NOTE: The location of the trap within each field is automatically recorded via GPS on the mobile device.

4. Service Traps

1. Upon arrival in the field, navigate to the Site Service database within the electronic system and create a new site service record comprised of the site name, date, and field technician name.
   NOTE: The initial trap service is conducted two weeks after trap deployment, and every two weeks thereafter. If higher-resolution trap catch data is desired, a weekly trap servicing may be done, although we note that bi-weekly sampling is sufficient to capture general movement trends throughout the season (Figure 1).
2. Locate the trap in the field. Place a fine-mesh hand sieve (mesh size 0.8 – 1.0 mm) onto a plastic container and pour the kill solution from the collection cup through the sieve. Transfer the kill solution back to the collection cup and swish the liquid around to ensure that all the CBB are removed from the collection cup.
3. Within the new site service record, navigate to the Trap Service database and create a new trap service record. Enter the relevant trap number, and photograph the sieve with the site name and trap number in the background. Save the photograph to the trap service record.
4. Using a spoon or metal spatula, scoop all insects into a vial filled with 70% ethanol. Label the vial with site, date, and trap number.
5. Refill the collection cup with fresh kill solution and screw back onto the trap. Once per month wash out the collection cup with soapy water, rinse, and replace with fresh kill solution. Also replace the attractant and bag once per month or as needed.

5. Service Zones for Plant Phenology

1. Within the site service record, navigate to the Zone Service database and create a new zone service record. Select a sampling zone from the site map in the linked Zones database.
2. To avoid sampling bias, randomly select a tree from within the zone by casting eyes downward so only the bases of the trees are visible. Standing in front of the selected tree, randomly choose a lateral branch around chest height. Clip a ruler onto the selected branch, making sure that the ruler does not block any of the reproductive parts (nodes, buds, flowers, fruits) from the camera's field of view.
3. Take a single photo ensuring that the ruler and the entirety of the target branch are visible. Take a second photo of the whole tree; try to get as much of the mid-level canopy in the photo as possible. Save both phenology photos to the Zone Service record.

6. Service Zones for Damage Assessments of Green Berries

1. If the branch used for phenology appears to have >30 green berries, count the number of berries on the branch that are at least pea size (~0.6 cm) and larger, and are green to light yellowish-green in color (BBCH scale 77 – 85¹⁷). Enter this number into the Zone Service record.
   NOTE: If the branch used for phenology appears to have <30 green berries, haphazardly select a lateral branch at about chest height from a tree in the target zone with >30 green berries visible. Do this from a distance to avoid bias in selection.
2. Also in the Zone Service record, enter the number of green berries infested by CBB on the branch. Infested berries will have a small hole that is typically located in the central disc; CBB may or may not be visible in the hole.
3. Enter the number of green infested berries with visible white B. bassiana fungus. The fungus may be seen on the CBB and/or surrounding the entrance hole.
   NOTE: Further assays may be needed to identify fungal species if this is of special interest.
4. Enter the number of raisins (dried berries) on the branch. This information can be used to understand relationships between management practices (e.g., strip picks) and CBB infestation.
5. Collect three infested green berries from the branch; these will be taken back to the lab and dissected to assess CBB position within the berry.
   NOTE: Green infested berries may be acquired from other branches within the zone if the branch used for the damage assessment has <3 infested green berries.
6. Place infested green berries in a plastic container and label with the site and date. Store containers in a cooler on ice until they can be transported back to the lab.
   Note: Ideally, berries should be dissected within 1 – 3 days of collection to ensure maximum survivorship of CBB.
   1. Store berries (as needed) in the lab at 14 °C for up to three days with little to no mortality (S. Fortna & R. Hollingsworth, personal communication).
7. Repeat steps for phenology and berry damage assessments in each sampling zone.
   NOTE: Approximately 25 branches should be sampled for large farms (~1 ha), and ~15 branches should be sampled for small farms (~0.5 ha). For dissections, 75 green infested berries should be collected from large farms, and 45 from small farms on each sampling date. During some parts of the year, it may not be possible to collect this number of berries. In this case, try to collect a minimum of 50 green berries for large farms (~1 ha) and 30 green berries for small farms (~0.5 ha).

7. Count the Number of CBB in Each Trap

1. Place a coarse-mesh hand sieve (mesh size ~1.5 mm) over a plastic container, and empty the beetles from the collection vial into the sieve. Use a wash bottle filled with water to get all the contents out of the vial.
2. Use the wash bottle to spray the contents in the sieve, forcing as many small insects through the sieve as possible. This permits larger insects and debris to be separated out from the small beetles in the sample, and limits inaccuracies in volumetric estimates of CBB. Discard the large insects and debris and rinse out the sieve.
3. Place a fine-mesh hand sieve (mesh size ~1.0 mm) over a second plastic container, and pour the contents of the first container into the fine-mesh hand sieve.
4. If there are more than several hundred CBB, skip ahead to step 7.6. If there are less than several hundred CBB, place the fine-mesh sieve onto a paper towel to remove excess water. Turn the sieve upside down and tap all the contents onto a clear plastic lid. Spread the beetles around with a paintbrush if they are clumped together, and allow them to sit until dry.
5. If there are less than several hundred CBB, use a fine-tipped paintbrush or similar implement to line the beetles up into rows that are several beetles wide, and begin counting under a light microscope. Count the total number of beetles and separate into "CBB" and "other" categories.
6. If there are more than several hundred CBB, transfer the CBB from the fine-mesh sieve to a 10-mL syringe using a metal spatula. Place the ejector column into the syringe and press downward gently until a slight resistance is felt, while being careful not to crush the beetles. Record the volumetric value on the syringe.
7. Count 200 beetles from the volumetric sample using the protocol described above. Use the following equations to determine the number of CBB versus other beetles in the sample.
   1. Estimate CBB count using:
      Total CBB estimate for the sample = (# CBB ÷ 200) x ∂ x (mL measured in syringe).
      NOTE: Here ∂ = the number of insects/mL. It is recommended that an estimate of ∂ be made for each region; on Hawaii Island a value of 1033 was measured.
   2. Estimate other count using:
      Total "other" beetle estimate for the sample = (# Other ÷ 200) x ∂ x (mL measured in syringe).
8. When the trap count has been completed, navigate to the relevant Trap Service record and enter the number of CBB and other beetles.

8. Score Phenology Photographs  

1. Export the coffee phenology photographs from the data collection application. Open the photograph and locate the branch with the attached ruler. For this branch score the following.
   1. Score the number of nodes (attachment points of the leaves to the branch).
   2. Score the presence or absence of immature buds, mature buds, candles, open flowers, and pin heads.
   3. Score the number of pea sized green berries, immature green berries, mature green berries, berries showing a color break, fully ripe berries, and raisins.

9. Dissect Berries to Determine CBB Position

1. Take the infested green berries out of cold storage and allow them to warm to room temperature for 10-15 min before proceeding with berry dissection. This recovery time is important so that CBB can be accurately assessed as alive or dead.
   Note: Dissection of the infested berries allows the position of adult CBB to be determined. Position AB indicates the female has initiated penetration into the berry but has not reached the endosperm; position CD indicates the female has entered the endosperm¹³.
2. Using a scalpel or similar implement, make a slice through the berry parallel to the central disc as a preliminary assessment of beetle position. Next, make a series of shallow slices perpendicular to the central disc and around the entrance hole to determine if the CBB is in the AB or CD position.
3. Subdivide the AB and CD categories into "alive", "dead by Beauveria bassiana", "dead by other causes", and "beetle missing". If it is unclear if the adults are alive or dead, zoom in with the microscope and watch the legs for movement.
4. Place counted individuals into a dish with water or alcohol. This helps to keep track of what has been counted, and prevents adult beetles from escaping into the lab.
5. Once the dissections for a site are completed, navigate to the Berry Dissection database in the relevant site service record and enter the total number of CBB in each category.
6. Place dissected samples in a container and freeze for 72 h before disposal.

10. Service Manual Weather Stations

Note: Weather stations requiring manual data download may be serviced bi-weekly or monthly to download data and ensure all sensors are working properly. Weather variables that are important to consider for understanding CBB biology may include rainfall, humidity, air and soil temperature, solar radiation, photosynthetically active radiation (PAR), soil moisture, and wind speed/direction.

1. Locate manual weather stations in the field. In the electronic system, open the relevant Site Service record and navigate to the Weather Station Service database. Create a new weather station servicing record that is linked to the relevant weather station deployment record.
2. Use a waterproof shuttle to directly connect the data logger and laptop for data download. Make a note in the weather station servicing record that data was downloaded.
3. Once data has been off-loaded, manually re-launch the solar and temperature/humidity loggers to ensure they have the correct settings (rainfall logger does not need to be re-launched). Check battery level and replace as needed. Make a note in the electronic system that this has been done.
4. After returning to the lab, add the most recent data to the full weather stitches and update the metadata record.

11. Record Management Practices

NOTE: Information on management practices may be used to understand patterns in CBB activity and population sizes. Relevant management practices may include (but are not limited to): spraying B. bassiana fungus, spraying pyrethins or other insecticides, pruning, weed management, strip picking, cherry picking, removing raisins from the ground, etc.

1. In the Management database, create a new management record with the site name, date and type of management practice conducted.