The use of unmanned aerial vehicle for controlling Oryctes rhinoceros beetle on Immature Oil Palm

Current constraints on labor supply and human error have triggered interest in using automated technology as an alternative to conventional operations. Hence, Minamas Plantation has embarked on an exploratory trial using unmanned aerial vehicles (UAV) in one of the labor-intensive works in the oil palm industry, which is the prophylactic fortnightly spraying of beetles on immature palms. The trial was conducted between October 2021 and January 2022 at PT Aneka Inti Persada. Three treatments, namely UAV, conventional, and no control, were investigated under two different environments (straight-line and terrace plantings). A randomized complete block design was adopted, with three replicates of immature oil palms planted in 2020. Prior to treatment implementation, all treated palms were georeferenced. The insecticide used was Cypermethrin (5.5% EC) at 20 ml/L of water. Insecticide application by point-to-point spraying was conducted for 4 seconds on each palm (250 ml solution/palm) at 0.5 – 1.5 m above the canopy. Knapsack spraying was performed as per estate practice. No significant difference in fresh damage was recorded between the UAV and knapsack spraying as compared to the control plots. Palms in control plots continued to be inflicted by beetles with more than 5% fresh damage throughout the trial period. The trial concluded that UAV technology has great potential for adoption in daily estate operations. The next step of the trial was to compare the cost-effectiveness between treatments. This study did not carry out the economic analysis due to the scale of the trial.


Introduction
The Oryctes rhinoceros beetle (ORB) is a major pest of oil palm and is known to cause severe damage to immature palms.The oil palm industry across Africa, Asia, and the Pacific region faced serious damage as the number of beetles increased.This has sparked interest in controlling and managing this persistent insects [1][2].Efforts to control Oryctes beetles have been carried out continuously, especially in immature palms.Chemical control measures are most effective for young palms against ORB.Chemical pesticides are used to prevent the adult beetles from damaging the spears and growing points [3].One of the control techniques used to manage the ORB population is insecticide application (Cypermethrin) to the growing point of the palm using knapsack spraying.Such operations require labor, which is becoming scarce and susceptible to how efficiency and chemical exposure.
The rapid development of unmanned aerial vehicles (UAV) in the industrial sector has led to the incorporation of this technology into daily plantation activities, such as spraying agrochemicals.It has been shown that the application of UAV or drones can reduce labor requirements and perform tare, which is difficult to perform manually.The application of UAV in agriculture can be deployed effortlessly to improve crop productivity.Its precise application can reduce the amount of water and agrochemicals used [4].According to Xin Ji (2021), several benefits of using drones to control Oryctes beetles are high terrain adaptability, more time efficiency, consistent spraying operation, low safety risk, and operator health [5].Coupled with remote sensing technology, the efficiency of drones can be significantly increased.The use of drone sprayers in oil palm plantations could be an essential tool for controlling pest outbreaks [6].Although until now there are still pros and cons regarding the utilization of drones among farmers.They believe that the use of knapsacks is more profitable and cost-effective, even though it is riskier to the health of operators and the environment [7].Therefore, this trial aimed to provide information on the effectiveness of drones for spraying insecticides to control the Oryctes beetle, compared to a knapsack sprayer.

Materials and methods
The trial was implemented on immature oil palm year of planted 2020, PT Aneka Inti Persada, Minamas Plantation, from October 2021 to January 2022.A trial design using a Randomized Complete Block Design (RCBD) comprised six treatments (Table 1).The treatments were sprayed using a drone and knapsack, whereas the palms in the control plot were left untreated.All tested in straight/flat and terrace planting areas.All treatments were carried out in 3 replicates.

Table 1.
Treatments in flat and terrace area.

Drone Description
The geo-reference point of each palm was created two days before treatment using a mapping drone from a technology provider and data processing for accuracy.The drone provided by the local technology provider was a D-16 RTK hexa-rotor UAV with a tank capacity of 16 L equipped with RTK Base for geo-location with 5 cm accuracy based on palm counting map input.

Pre-Treatment Assessments
Prior to treatment implementation, ORB fresh damage was assessed in all treatment plots.Spray coverage of drone sprayer drift onto measured palms was observed using water-sensitive paper and then quantified.To ensure a consistent spray volume, calibration was carried out to determine the time of spraying for 250 mL/palm using a full cone nozzle of 800 microns.Insecticide used was Cypermethrin 5.5% EC at 20 ml/L of water.

Insecticide Application
Insecticide application by point-to-point spraying was conducted for an average of 3.7 seconds for each palm (250 ml solution/palm) at 0.5 -1.5 m above the canopy.Knapsack spraying was implemented according to estate practice.Treatments were implemented in six rounds at biweekly intervals.2.4.2.ORB fresh damage.The symptoms of fresh damage observed were freshly broken spear, fresh fiber waste material, and the presence of the beetle on the spear (figure 1) [2].The total number of observed palms was 612 (20% of the total number of sprayed palms).It was carried out at 7 DAT and 14 DAT after each round.

Statistical Data Analysis
ORB fresh damage data were analyzed using Analysis of Variance (ANOVA) followed by Tukey test with a 95% confidence level.

ORB fresh damage pre-treatment assessment
Visually, palm damage caused by the ORB was almost evenly distributed.Table 2 shows that there was more than 66% palms damage at the trial site.They were dominated by old symptoms.It indicates that the population of the beetle was quite high.However, fresh damage remains below the economic threshold level.The ORB destructs palms by burrowing into the shoot and spear area.It leaves fiber waste material.Oryctes fresh damage on pre-treatment was low due to prophylactic Cypermethrin spraying conducted using a knapsack by the estate.

Point-to-point drone spraying
Oryctes rhinoceros beetle (ORB) control using a drone trial was initiated and implemented as an alternative to control ORB.The spraying drone D-16 and software for point-to-point spraying were developed and tested under the specifications of the local technology provider.There were some constraints on the use of drones in this trial.Initially, it was observed that the spray volume was not consistent for some palms due to the double tag.Adjustment (removing double tags) was perform so that the spray volume became consistent measured with the graduated cylinder.The drone was not equipped with automatic obstacle detection, therefore the spray height above the palm could not be adjusted automatically, particularly in the terrace area.The height of spraying was set to follow the highest palm.This matter caused the height of spraying to be slightly different among the palms.According to Hu et al., (2022), some features to consider in a drone sprayer include obstacle avoidance and terrain-following control technology.These technologies aim to improve the safety, reliability, and intelligence of drone operations as well as to maintain the crop canopy's relative height steady [8].Preparation to install the RTK base to connect to the drone slightly taking a long time (10 -30 minutes) and the signal depends on the weather and geographical conditions.The accuracy of spraying was interrupted by wind.In the first round, some tagged palm points did not fit to the center of the spear.Evaluation and adjustment were put on to remove all inappropriate results from this trial.

Spray coverage test using water sensitivity paper (WSP)
A spray coverage test was assessed to determine the nozzle type and accuracy of spraying right above the center of the spear.The WSP was used to collect spray droplets from the nozzles (Figure 2).This indicates droplet coverage on the surface of the canopy.The main objective of crop protection procedures is to improve the coverage of treated areas, and the essential task of technical spraying factors is to increase this quality [9].An initial stream nozzle of 1000 microns was used.However, it was adjusted to a full-cone nozzle of 800 microns to minimize the drift swath.This resulted in a smaller swath at less than 80 cm from 4 m above the ground compared to 120 cm.The full cone nozzle of 800 microns had more even spray coverage on the spear region and the first spiral of the frond, with an average of 83,8% while the stream nozzle had only 67,5% (figure 3).Therefore, a full-cone nozzle of 800 microns was selected to use.

Parameters requirement
Prior to insecticide spraying, several parameters were required to meet the minimum requirement (Table 3).The pre-application parameter requirement results showed that almost all the parameters were suitable for drone implementation in the trial plot.The item that did not fit was the use of Cypermethrin spray volume, which affected the number of palms.The spray volume per palm was 250 ml/palm owing to the robustness of the palm.Therefore, the number of palms sprayed decreased to 38 palms.It also affects the time of spraying to 3.7 seconds.The spray volume per mission was less optimal (only 9.9 L compared to the 16 L capacity).Include obstacle avoidance is one of the features that should be considered.

Time efficiency of drone spraying
The time duration recorded using a stopwatch includes all activities, i.e. installation of RTK Base, flight preparation, flying time, and downtime between flights.Drone conducted spraying for 38 palms per mission using a 9.9 L solution.Therefore, the mission per round was 77 flights.There were 78 flights on the 5th day due to the drone crashes.The drone was accidentally run out its battery.Therefore, it is important to ensure that all batteries are fully charged before use.It was suggested to optimize the battery use for a 15 L solution equal to 58 palms per mission.Therefore, the mission per round should include only 51 flights.Missions in flat and terrace areas were slightly affected by topography.Figure 6 shows a sample of one mission in the terrace area.H is the home base of the drone for the takeoff and landing.It took a longer time to complete the mission for hilly areas and windy conditions.According to Wang et.al.
(2023), spray drift is significantly affected by wind speed, nozzle type and nozzle coding [10].Drone spraying was affected by the wind when the velocity reached 10 km/h.This affected the missed spraying for some palms due to spraying was conducted at least 1 m above the palm.Some of the recorded velocities were 9.35 km/h, 11.15 km/h, and 11.51 km/h, therefore spraying was postponed for a few minutes.Nevertheless, the use of drone sprayers has the potential to be implemented in hilly areas, reduce chemical exposure to the operators, and allow precise spot spraying [10].

Fresh damage assessment
Oryctes fresh damage was under control in treatments using a drone and manual knapsack.Both treatments showed significant results compared to the control, especially at 14 days after treatment (DAT) in the 2nd round, 7 and 14 DAT in the 3rd round, 14 DAT in the 4th round, 7 DAT in the 5th round, and 7 & 14 DAT in the 6th round.The results of ORB damage as tabulated in Table 5.Thus, ORB control using knapsack spraying could suppress fresh damage.It might suppress the infestation up to zero accidents.However, inconsistently also showed an 8% infestation rate.The expected time of using a knapsack to spray each palm at the appointed time was not gained.In contrast to Roudhiyah and Saepuloh [11], spraying of Cypermethrin using a knapsack at biweekly intervals could protect the immature palms from Oryctes damage below threshold level of 5%.This means that the Cypermethrin solution was not applied at the same volume to each palm.While in drone spraying, the time of spraying was similar for each palm at approximately 250 ml for an average of 3.7 seconds.The efficiency of manual knapsack spraying is subject to the performance compared to the measured and consistent in drone spraying [5].ORB population based on fresh damage in both flat and terrace areas was quite different (figure 7), indicating that high incidents occurred at different times.Oryctes fresh damage was observed to be higher in the flat area, while in the terrace area, it remained stable at 5% up to 7 DAT for 4th round.Fresh damage increased in both areas from 14 DAT in the fourth round, then decreased after 14 DAT for the fifth round in the flat area, and after 7 DAT for the fifth round in the terrace area.Kamarudin (2005) stated that there are significant relationships between the O. rhinoceros population, the environment and physical characteristics of its habitat such as cover crop, moisture, soil pH, and rainfall.Rainfall affects beetle populations, especially in flat areas [12].Rainfall data were collected from a rain-gauge ombrometer installed near the trial plot.According to Mohamad (2016), the relationship between rainfall and Oryctes beetle population is very strong (as shown in the flat area graph, Figure 7).However, rainfall was not the primary factor affecting the dynamic population of Oryctes rhinoceros (as shown in the terrace area graph, Figure 7) [13].

Conclusions
From the efficacy trial, control of Oryctes rhinoceros beetles using a drone was as effective as manual spraying.The main advantages of drones compared to manual knapsack spraying are precision spraying to all palms and the ability to deliver the desired amount of chemicals to the target.Future work is needed to determine the cost of large-scale commercial trials.

Acknowledgements
The authors wish to thank Crop Protection and Geospatial & Remote Sensing Unit team, Minamas Research Centre, and PT Aneka Inti Persada, Pinang Sebatang Estate for their continuing support and cooperation in this study.We also thank the local technology provider.

3 2. 4 .
Assessments Parameter 2.4.1.Time efficiency.The parameters observed during insecticide application were the volume used and time of spraying of each mission.Time efficiency to control the specified area was recorded and compared to estate practice.

Figure 1 .
Figure 1.Oryctes fresh damage (a) fresh broken spear, (b) fresh fiber waste material, and (c) the presence of the beetle on the spear.

Figure 4 .
Figure 4. Drone accurately sprays on to center of the spear.

Figure 5 .
Figure 5. Pilot controlled drone from distance comparing manual knapsack direct spraying to each palm.

Figure 6 .
Figure 6.Flight map spraying drone on terrace area.

Figure 7 .
Figure 7. Oryctes infestation in flat and terrace areas against rainfall.
[1] Wood B J 2002 Pest control in Malaysia's perennial crops: a half century perspective tracking the pathway to integrated pest management Integrated Pest Manag.Rev. 7 173-90 [2] Chung G F 2012 Effect of pests and diseases on oil palm yield Palm Oil 6th chapter p163-

Table 2 .
Pre-treatment of ORB attack prior the trial.

Table 3 .
Pre-application parameters requirement result.

Table 4 .
Time efficiency of drone spraying.

Table 5 .
ORB fresh damage rate after insecticide application.
a Means that do not share a letter are significantly different (α=0,05).