Basic Physiological Research on the Wing Flapping of the Sweet Potato Hawkmoth Using Multimedia

Large moths, which travel great distances in the natural environment, are highly likely to be repeatedly exposed to multiple types of agrochemicals. While LD50 values are standard indicators of acute exposure, chronic exposure is assumed as 1/100 of the LD50 value. Simple arithmetic indicates if an individual organism is exposed to 20 different types of agrochemicals, theoretical exposure to toxins polluting the environment is 1/100 × 20 = 1/5 of the LD50 dose. Our team has cast doubts on such a simplified method of calculating chronic toxicity levels based on LD50. Our study seeks to obtain physiological data of moth flight to screen basically healthy individual specimens for those exhibiting physiological anomalies in physiological function at low level exposures, with the goal of identifying an indicator that may replace 1/100 of the LD50 value.

Many reports of studies of flapping wings have focused on the theme “why insects can fly.” In recent years, military research has investigated insect flight mechanisms as potential propulsion mechanisms for micro air vehicles (MAVs). A mainstream trend in the fluid dynamics approach to the study of flight mechanisms applies numerical simulation techniques to analyze the lift provided by wings. These studies typically assume the steady-state wing theory, while taking into account the additional effects such as separation vortex motion [4]-[9].

The following equation gives the lift L produced by the translational motion of wings required to achieve airspeed U:

where

C is the coefficient of lift;

ρ is air density; and

S is the surface area of the wing.

This equation describes only a relationship that exists at a momentary snapshot of wing motion meeting a specific criterion (steady-state wing condition). Accounts based on pioneering studies and experiments suggest that for continuous wing flapping, the lift created by the second and third flap differs from that of the first flap. For this study, we decided to attach a monoaxial (mainly pitch axis) angular velocity sensor to a moth, albeit under restrictive conditions wherein the moth was suspended and constrained by a string.

We focused on angular velocity because we thought that due to its correlation with lift, we could observe muscular contraction by measuring angular velocity. This is because the measurement of the angular velocity has been technically completed in the past by measuring the force emitted by the flight of the whooper swan and the heart of the chicken (proportional to blood pressure) [10]-[12].

We gathered and analyzed physiological data of wing flapping for 100 hawkmoth (Agriusconvolvuli) specimens suspended in air using string (in a constrained state).

Number of moth specimens: 100 specimens of Agrius convolvuli (50 males and 50 females)

Mean body weight: 1.1 g

Mean dry weight: 0.28 g (approximately 75 % of the body weight is water)

Agrius convolvuli is the largest hawkmoth found in the Japanese islands. Its wingspan may reach 105 mm, often resulting in its being mistaken for a bird. Larvae feed on the leaves of plants belonging to the Convolvulaceae family, such as sweet potatoes, making them despised pests among farmers. They are bred for research purposes at the University of Tokyo and other institutions [13]-[15].

We selected the hawkmoth for our study due to its maximum payload, which is regarded to be around 0.3 g. This allows these insects to carry a micro radio transmitter while flying freely outdoors in experiments to be carried out after obtaining the appropriate radio transmission license from the Ministry of Internal Affairs and Communications.

As with vertebrates, surgical procedures on moths require anesthesia. Having discovered early that using nerve-block agents such as xylocaine acting directly on the nerves results in total loss of the moth’s ability to fly, we were compelled to examine other reversible procedures.

Reducing temperatures to 8–10 °C is known to slow moth metabolism, inducing a state of suspended animation. We placed a moth in a Tupperware® with 10 small holes for ventilation and stored the container in a commercial refrigerator. The moth became completely immobile when left in the refrigerator for approximately 60 minutes, giving us time to perform a 5-minute surgical procedure. After the procedure, the moth was gradually returned to room temperature and able to fly freely in about an hour.

Moths are covered by generous quantities of lamellar and piriform scales. Some procedures involve preparing moths by removing the scales using air sprays to expose the exoskeleton. Based on our extensive experience with surgical techniques, we chose to secure the field of view under a surgical microscope by wetting the surgical field with a diluted antiseptic agent in preparation for the surgical procedure.

The surgical procedure involves embedding multiple electrodes into the body of the moth. For each electrode, a small hole is formed in the dorsal surface of the hard, shell-like exoskeleton using a pin fixed to a celluloid plate. The hole measures approximately 5 mm deep and 0.2 mm in diameter. The procedure does not involve bleeding. After the hole is formed, a 0.18-mm enameled wire prepared beforehand is bent at a 90-degree angle at the tip to allow electrode insertion into the hole to a depth of approximately 4 mm. Solder is applied beforehand to the 2-mm section at the tip of the enameled wire.

The enameled wire must be fixed to the hole to prevent detachment during wing flapping. We applied small amounts of Aron Alpha® instant adhesive (Toagosei Co., Ltd.) for this purpose. (The team led by Prof. Ando used beeswax.) The instant adhesive has the advantage of solidifying together with the piriform scales surrounding the hole, which serve as auxiliary structures for fixation (Figure 1).

Fig. 1. Inserting and fixing electrodes for EMG.

Download Original Figure

The muscles involved in wing flapping are the dorsal longitudinal muscle (DLM) for the downstroke and the dorsal ventral muscle (DVM) for the upstroke. These two muscles are called power muscles, but indirect muscles. This is because the curved plate on the thoracic back is deformed and the wings are moved based on the lever principle.

The 3rd axillary muscle is responsible for wing retraction and the subtrochanteric muscle of the wing for abduction and downstroke, these two are called steering muscles, direct muscles those are linking with wings. In this study, we intended to measure two opposing muscles associated with the highest potentials (DLM and DVM) in Figure 2.

Fig. 2. Anatomical conceptual diagram of muscles involved in wing flapping.

Download Original Figure

We recorded EMG and angular velocity (force) of wing flapping together with high-speed video of the moth suspended in air and constrained by string. Figure 3 presents a conceptual diagram of the experiment. The angular velocity sensor is grounded via the pin, which also serves as the fourth electrode (neutral point) of the EMG amplification circuit. The wiring to the angular velocity sensor consists of three 0.18-mm enameled wires, connected perpendicular to the pitch angle so as not to interfere with pitch angle measurements.

Fig. 3. Conceptual diagram of experiment.

Download Original Figure

The moth is suspended in air and constrained with an enameled wire. However, it can still flap its wings and the counteraction of the flapping motion will manifest as vertical vibrations of the moth’s abdomen to retain balance. This movement can be measured on the thorax and can be recorded as pitch angle using the angular velocity sensor.

Since the neutral point is grounded, the zero potential of the operational amplifier is adjusted manually after inserting the electrode. Once this is adjusted, the output should remain within the range of −5 to +5 when the amplification factor (gain) is 134 and fall within the range of the A/D conversion element. Based on the frequency characteristics provided in the specifications for AD623, when the GB product is 800 kHz and gain is 134, the upper limit of the band frequency will be 70 kHz. Since 70 kHz is far higher than the sampling rate of 20 kHz of the A/D converter, using the AD623 should pose no issues for this amplification circuit with respect to circuit design.

Fig. 4. Circuit diagram of EMG with two instrumentation operational amplifiers (OP AMP AD623) [16].

Download Original Figure

An angular velocity sensor, which relies on the principle that a vibrating body with angular velocity will exhibit the Coriolis force, was used to measure pitch angle. This compact piezoelectric vibrating gyro sensor features a simple cap-base structure measuring approximately 0.1 cc in size and contains a ceramic bimorph vibrator. This gyro sensor was welded perpendicularly on the pin fixed to the moth. The pin doubled as a neutral point for the EMG amplification circuit. The LF411 operational amplifier was used with an amplification degree of 13 and zero bias being controlled using a variable resistor (Figure 5).

Fig. 5. Circuit diagram for angular velocity measurement using the Murata ENC03R [17].

Download Original Figure

In Figure 5, the pin inserted into the dorsal thorax is welded to the enameled wire and grounded and also serves as the fourth electrode of the EMG amplification circuit.

We recorded data to a PC using a 3-channel analog-to-digital converter with a USB connection at a sampling rate of 20,000/sec and 16-bit resolution potential variation (y axis). We used Microsoft Excel to analyze the linear data, using programs for calculating EMG amplitude and PP intervals created with VBA.

The data files for wing flapping for 100 moths consisted of recordings lasting 49 seconds at a sampling rate of 20 kHz and 24 seconds at a sampling rate of 40 kHz. Since we processed the data in Excel, recording duration was restricted by the maximum number of rows permitted in an Excel sheet. The data obtained is explained below.

Figure 6 shows a characteristic 1-second section taken from 3-channel data recorded for 49 seconds.

Fig. 6. Digital conversion performed at sampling rate of 20,000/sec and 16-bit potential variation (y axis) for 1 second.

Download Original Figure

The data presented in Figure 6 corresponds to a 1-second section illustrating the transition from a phase in which the moth slowly begins to flap its wings to the active flapping phase. As the amplitude of the pitch angle increases, a transition from a double-peaked waveform to a single-peak waveform occurs. The baselines (DC components) for both the EMG1 (for the dorsal longitudinal muscle) and the EMG2 (for the dorso-ventral muscle) switches to a decreasing trend at the transition. We believe this mode change in the DC component will be useful in assessing the relationship between the movement of the muscle as a whole and the depolarization of the inserted electrodes. The DC component cannot be observed with the AC amplification circuit due to capacitive coupling.

Figure 7 represents an enlarged 100-msec section of the plot above. We confirmed two peaks in the downstroke muscles, attributed to the difference in timing of the wingbeats for the left and right wings. The potential variation is most likely generated by the dorsal longitudinal muscle.

Fig. 7. Signals associated with DLM and DVM and PP intervals of EMG.

Download Original Figure

Between these peaks, there is a signal for the upstroke, which we believe is associated with the dorso-ventral muscle.

Figure 8 shows the data for 100 EMG amplitudes and the corresponding amplitudes of angular velocity. The correlation coefficient of the regression curve is R = 0.792, indicating generally high correlation (does not intentionally exclude outliers). This demonstrates that greater muscular strength correlates with greater angular velocity, suggesting that continuous EMG measurements can be used for physiological functional evaluations of moth flight.

Fig. 8. Correlation between EMG and angular velocity.

Download Original Figure

Next, we analyzed frequency fluctuations. Figure 9 plots the data for 100 PP intervals and the amplitude of angular velocity for the corresponding intervals. The PP interval here is similar to the RR interval in ECG and represents the correlation between angular velocity and the interval between wingbeats.

Fig. 9. 100 EMG amplitudes and corresponding angular velocity amplitudes.

Download Original Figure

The correlation coefficient for the correlation between the fluctuation for the 100 PP intervals and the amplitude of angular velocity is R = 0.379 shown in the Figure 10. The accuracy of the regression curve is relatively poor (does not intentionally exclude outliers).

Fig. 10. Correlation between 100 PP intervals and angular velocity amplitudes.

Download Original Figure

Although angular velocity is a factor that contributes to the fluctuation of the lift provided by wing flapping, our investigation showed a poor correlation between PP interval fluctuations of flapping motion and the fluctuation of the lift, suggesting that the former does not directly contribute to lift. This experiment failed to identify the physiological mechanism responsible for the PP interval fluctuations of flapping motion.

The experiment proved a correlation between angular velocity in the pitch direction of insects and EMG signals. Previous studies have analyzed insect flight using cameras and wind tunnels or EMG itself. However, these studies have been unable to establish a correlation between pitch angle and motor physiological data (corresponding to the EMG signal amplitude in this study). Our experiment corroborates this correlation and demonstrates the potential usefulness of EMG in functional evaluations of moths. Needless to say, physics supports the existence of a linear correlation between pitch angle and lift provided by wings. For agrochemicals, the exposure limits for chronic low-level exposure have been defined as 1/100 the LD50 value, the dose at which half of the exposed population dies in the case of acute exposure. No current standard for insects is based on physiological functional evaluations.

Even at doses far below 1/100 of the LD50 value, a decrease observed in EMG potential may indicate an onset of functional disorder. We believe this points to the potential for novel screening methods capable of evaluating the effects of repeated exposure to multiple agrochemicals. The application of such a method would not be limited to agrochemicals. It could also be used to assess exposure dose in radiation exposure incidents, such as the Fukushima-Daiichi nuclear power plant accident, based on functional evaluations of insects [18]-[24]. After the Fukushima-Daiichi nuclear power plant accident, professional journals reported numerous observations of anomalous white spots (depigmentation) on the wings of butterflies of the Lycaenidae family.

Government-specified exposure dose limits are calculated from the estimated total lifetime dose over 70 years and do not take into account functional evaluations of insects. As with LD50 values, calculations of these dose limits are not based on physiological evidence.

We confirmed a correlation between EMG amplitude and amplitude of angular velocity. What factor accounts for the fluctuations observed in PP intervals, a frequency component? We believe the answer will prove crucial in producing a parameter for screening functional disorders caused by chronic low-level exposure.

In human ECGs, the RR interval reflects the status of autonomic nervous system regulation. Similarly, we believe that fluctuations in the PP interval for wing flapping of the moth may reflect the state of regulation. We believe chronic low-level exposure to agrochemicals may manifest as anomalies in Δf fluctuations.

In the present experiment, we constrained the moths to keep them from ascending or flying forward, a factor that may have inhibited their free flapping motions. Additionally, our focus was on the two large muscles (DLM and DVM). We may have failed to make proper assessments of the movement of the two other relatively minor muscles, which are known to be associated with hovering and rotation in flight. Thus, the cause of Δf fluctuations for these muscles may only be identified through monitoring of EMG during free flight.

In the next study, we expect to investigate the fluctuations of the two steering muscles (the 3rd axillary muscle and the subtrochanteric muscle) and fluctuations of PP-Interval, which are directly linked to the wings.

For subsequent experiments, we devised a transmitter using microchips and batteries for wrist watches (1.5 V) and carried out free-flight experiments after obtaining appropriate licenses for radio transmission from the Ministry of Internal Affairs and Communications. The experiment carried out by Prof. Ando’s team, which preceded ours, achieved 2-channel transmission using an 80 MHz band FM radio transmitter [13]. We plan to pursue research on Δf fluctuations to analyze the role of the remaining two muscles as part of efforts to establish a screening method for chronic low-level exposure.