Seeing green: phototactic responses for management of Rove beetle Paederus fuscipes

Paederus fuscipes Curtis (Coleoptera: Staphylinidae) is a medical pest due to its attraction to bright light sources from human residential premises. This study aims to investigate the phototactic response on light wavelengths and colour preferences by P. fuscipes. Paedarus fuscipes showed positive responses to all types of light sources (fluorescent, LED, and incandescent) presented in the light and dark choice test, which indicates their attractiveness towards light presented in human habitation. However, most beetles showed a higher preference towards fluorescent and cool white LED. Paedarus fuscipes showed consistent preferences for white, blue and green, with the exception of red. Remarkably, green light elicited the strongest positive phototactic response compared to other colours. Overall, our results showed that P. fuscipes beetles prefer the range of 550 nm wavelength. Warm light colours such as yellow (warm white LED) with red illumination can be recommended to avoid infestation and in innovative management of Paedarus.


Introduction
Rove beetles, Paederus spp.(Coleoptera: Staphylinidae) are a genus of small insects of 2-8 mm in size.At least 650 described species worldwide [1] and universally scattered around tropical and temperate countries with P. fuscipes is the predominant Paederus spp.found in the rice field ecosystem in West Malaysia [2,3].Notably, P. fuscipes are renowned for having a unique hemolymph pederin that causes a lesion on human skin [4,5] known as Paederus dermatitis or dermatitis linearis.
Adult Paederus usually appears in the evening after dusk and is attracted to the artificial light of a light bulb [6].This positive phototactic behaviour could be the cause of dermatitis linearis outbreak cases in human settlements, particularly for those residing close to the agricultural fields [7][8][9].Unlike blister beetle from the family Meloidae that spontaneously bleed, Staphylinid beetles need to be crushed to provoke the release of the potent fluid [10].Although considered a pest to humans, P. fuscipes play an important ecological role as natural predators of other pest insects [2].However, these beetles can cause a significant cutaneous injury via their toxic vesicant fluid, therefore considered as a problematic pest.
Managing Paedarus can be a complex task due to the distinct habitats of adults and larvae.The larvae are typically deposited deep within the soil, making them challenging to control.The larvae of Paedarus infesting lawns have traditionally been managed by using soil insecticides since the 1970s and 1980s.During this period, short-residual organophosphates and carbamates were commonly employed [11].These products demonstrated their highest effectiveness when applied shortly after the hatching of eggs.However, these control methods have negative impact towards nontarget organisms [12] including these curative insecticides that faced restrictions or were discontinued for turf usage during the 1990s due to environmental concerns and the implementation of the 1996 Food Quality Protection Act [11].In human habitation areas, only adults are typically encountered since they exhibit flying behaviour.Controlling adult Paedarus beetles often relies on the use of insecticides that are also effective against mosquitoes.Insecticides commonly used for mosquito control, such as pyrethroids and organophosphates, can be applied to areas where Paedarus adults are active.These insecticides are chosen for their broadspectrum effectiveness against various flying insects, including adult Paedarus beetles.According to Mhalu and Mandara [13], reducing the use of artificial light and using UV light traps might be helped to control an infestation of rove beetle.
While basic light attraction studies have been conducted on beetles, in recent years, there are multiple new types of lights have been introduced.Different light-emitting bodies emit different wavelengths that are almost indistinguishable from the human eyes.Generally, insect vision ranges between 340 to 700 nm at the violet to the red domain [14].Thus, insects are capable of detecting light sources such as ultraviolet (UV) and are able to differentiate colours ranging from violet to red domain in the visible light spectrum [15][16][17].Although, most insect species have trichromatic eye vision showing a preference towards colour between 360 and 550 nm range, which comprise the UVA, blue and green spectral regions [18].
The fact that most nocturnal insects are attracted to artificial light with shorter light wavelength is well known [19].Thus, UV light is known as a unique wavelength perceived by most insect vision [20].However, little is known about what types of light sources strongly attract most P. fuscipes dispersal flight towards human settings.The importance of visual cues based on their spectral sensitivity functions and colour preferences of this rove beetle species is lacking.Therefore, it would be crucial to determine the wavelength of the light spectrum of commercially sold light bulbs that is most visible to the insect range of vision.Our objectives were to determine the spectral range to which P. fuscipes are most receptive, from blue to infrared spectra.Herein, we investigate the colour preferences in P. fuscipes with three phototaxis experiments: (1) light and no light (dark) conditions, (2) light preferences based on three different types of light sources (fluorescent, cool white LED and incandescent light bulbs) and (3) the colour preferences based on four different colours of LED light sources (white, blue, green and red) using a Y-maze experimental design.

Collection of P. fuscipes in the rice field
Adults of P. fuscipes beetles were freshly collected a week prior to the experiment from an established colony in Telok Air Tawar rice field (5°29 9.3171 N, 100°23 1.3012 E) at mainland Penang, Malaysia.Beetles were captured using 12-volt battery-powered blacklight traps (CDC Fay-Prince Blacklight Trap, Model 812, John W. Hock, Florida, USA), each incorporating a near-ultraviolet (300-400 nm) LED which was placed along the rice field levee starting at 20:15 until 22:15 h, for use in the phototactic response experiment.

Rearing method of P. fuscipes
Adult Paederus were brought back to the Medical Entomology Laboratory, Vector Control Research Unit, School of Biological Sciences, Universiti Sains Malaysia and reared at the photoperiod of 12h:12 h (L: D).Paedarus was reared in a plastic container sized 16 cm weight × 16 cm in width × 7.0 cm in height.They were provided with cat food and a moist cotton ball as a food source.Paederus fuscipes colonies were kept consistent with the natural light-dark cycle to prevent disruption to the normal circadian rhythm.An hour before the experiments were conducted, each beetle was housed separately in an individually numbered plastic container measuring 5.6 cm height × 4.6 cm width and kept in darkness in between experimental trials.

Y-maze design
Paederus fuscipes light preferences were tested using a Y-maze test.The Y-maze experiment was adapted from Fadzly and Burns [21] with some modifications to the size and colour of the Y-maze.The Y-maze was designed using black polyurethane plastic with 5 cm width × 5 cm height × 20 cm arm length in size, with an angle of 60°apart between both arms (Figure 1).The base of the Y-maze is 25 cm in length.Black polyurethane plastic was used to give the impression of "dark" during nighttime as they are nocturnal insects.Clear plastic was used to cover only the arm's part to permit sufficient lighting and as illumination "windows" for observation to track Paederus beetle's movement during each experimental trial.

Paederus fuscipes attraction to different types of light sources
The light preferences experiment was conducted using three different types of light sources; (1) compact fluorescent light (Philips, 8W); (2) cool white light-emitting diode (LED) (Valens LED lightings, 5W) and (3) incandescent (Philips, 40W).All three light bulbs produced between 400 and 500 lumens (lm).This type of lighting was chosen for this experiment due to the fact that these types of lights are commercially sold and commonly used by residents.
The first experiment was designed to make P. fuscipes choose between two light conditions; (1) with a light source and (2) no light source (dark).The Y-maze test was conducted by placing a fluorescent bulb at only one end of the maze (arms site) for the light condition, while at the other end of the maze arm was covered with black polyurethane plastic for the no-light condition.The experiment was repeated using two other types of light (LED and incandescent) as the light source.Each experiment was conducted separately, and the left or right direction was also randomized to minimize direction selection bias.The choice experiment between the light source and no light source was replicated five times.For each replicate, ten insects were exposed for each experimental trial.A total of 150 experiments were run to represent three types of light sources between light and no light, repeated five times with each replication consisting of ten individual insects.This is to avoid the pseudo replication effect.
Only a single individual of P. fuscipes beetle was tested at a time, and a new individual will be used for each of the following experiments.Paederus fuscipes was introduced at the opposite end of the Y-maze and facing the light and dark stimuli.We considered that P. fuscipes made a decision after the beetle moved beyond the Y-maze intersection.If there was no response after one minute (i.e.P. fuscipes did not move towards either one of the intersections of the two arms), the trial was stopped, and the results were excluded.The experiment was considered successful (1) if P. fuscipes moves towards the light area and failure (0) if it moves towards the dark area for the analysis purposes.
For the second portion of the experiment, we wanted to test the preference between (1) fluorescent and LED bulbs, (2) fluorescent and incandescent bulbs and (3) LED and incandescent bulbs.Different types of light sources we placed at each end of the Y maze arm.There were three combinations, and for each combination, it was replicated five times (using ten insects for each replicate to avoid pseudo replication, totalling up to 150 experiments).The experimental procedures follow the previous experiment.However, for this experiment, each type of light source that P. fuscipes chose was recorded as 1 and the other is marked as 0.

Paederus fuscipes attraction to different coloured LED bulbs
This experiment was conducted to test the specific colour preference of LED lights.We used smart LED lights (Philips HUE, 8.5W).The wavelength (colour) can be manipulated via an App using a Samsung Galaxy Tab.Using the App, the LED lights were set to emit normal (white), blue, green and red light.As with the previous experiment, we put different light sources at the end of the Y maze.There were six combinations, and each combination was replicated five times (using ten insects for each replicate to avoid pseudoreplication, totalling up to 300 experiments).Similarly, the chosen light is recorded as 1 and the other is recorded as 0.
To coincide with their natural circadian rhythm, all trials took place after sunset at 1930 h in the laboratory.The room was kept in total darkness with blackout curtains during the trials to reduce excess light bias.The Y maze was monitored by the researchers using an Eveready Red light headlamp.

Spectrographic analysis
The spectrographic analysis was conducted to determine the spectral irradiance of the fluorescent, incandescent and LED.We measured the spectral irradiance using Ocean Optics Jazz, a portable spectrophotometer.Irradiance was measured with a cosine-corrected sensor and a D65 (normal daylight) light as a reference.Spectra were calculated at 5 nm intervals from 300 to 700 nm with SpectraSuite software.

Statistical analysis
The Y maze choice test was analysed using a binomial test.However, binomial tests could not test for the differences in the different types of light sources.Further analysis using one-way analysis of variance (ANOVA) was conducted to examine the (1) types of light sources (fluorescent, cool white LED and incandescent) and (2) coloured lights (white, blue, green and red) used that strongly induces most P. fuscipes attraction towards the artificial light sources.A post hoc multiple comparisons were conducted to determine the differences in light selection by using Tukey's honestly significant difference (HSD) test at α = 0.05.Prior to statistical analyses, all data were checked for normality of distribution at the 0.05 significance level by using the Shapiro-Wilk Normality test.Data were transformed to Log10 to conform to the normal distribution pattern.The one-way ANOVA was performed after the confirmation of normality and homogeneity tests.All analyses were conducted using SPSS version 26.0 (SPSS Inc., Chicago, IL, USA.)

Paederus fuscipes attraction to light and dark conditions
Paederus fuscipes responded significantly to all three types of light sources presented with the light (fluorescent, cool white LED, incandescent) compared to the dark (no light) source (P < 0.05; Figure 2).The response rate of P. fuscipes was seen higher towards the light sources.The binomial test showed that P. fuscipes significantly preferred light conditions (P = 0.000, Table 1).Among the types of light sources tested, the phototactic response of P. fuscipes was strongest towards the fluorescent light with a mean response rate of 8.4 ± 0.51 individuals.Paederus fuscipes attraction towards other light sources followed closely by incandescent and LED light by means of 8.2 ± 0.37 and 7.8 ± 0.49 (mean ± SE), respectively.

Paederus fuscipes attraction to fluorescent, LED and incandescent light sources
Once P. fuscipes were given choices between two types of light sources, P. fuscipes selectively chose fluorescent over incandescent light (P = 0.000; Table 2).
Although LED was selectively chosen more than incandescent light, the binomial test showed that it was not statistically different due to a weak response rate (P = 0.065; Binomial).Interestingly, fluorescent and LED light sources were chosen at equal rates (P = 1.000;Binomial; Figure 3).Overall, one-way ANOVA indicated a significant difference in the relationship between P. fuscipes response rate for different types of light sources used during experimental trials (F = 13.34,df = 2, 297, P = 0.00).Post hoc analysis using Tukey's test revealed that P. fuscipes response was significantly higher on fluorescent and LED than incandescent light sources (P < 0.05; Figure 4).The phototactic response of P. fuscipes was highest on fluorescent (31.50 ± 6.60), followed by LED (28.50 ± 3.50), and incandescent (15.00 ± 3.00).

Paederus fuscipes attraction to white, blue, green and red LED light sources
In a colour preference experiment, the binomial test showed that P. fuscipes was generally preferred light from shorter wavelengths.For the choice selection between blue (440 nm) and green (520 nm) LED light colours, P. fuscipes showed no significant preferences (P = 0.203).Subsequently, leaving the green light  unchanged, with another choice of a white light source, a considerable insignificant result was obtained (P = 0.888).There were also no significant differences between blue and white (P = 0.672).
However, when a red-light source (620 nm) was paired with the blue light bulb, P. fuscipes selectively chose the blue light over the red light (P = 0.000).Binomial tests also showed that P. fuscipes preferred green light to red light (P = 0.000) and white light to red light (P = 0.000).
On the whole, P. fuscipes response to different wavelengths of light colour was significant as indicated by one-way ANOVA (F = 35.14,df = 3.60, P = 0.000, Table 3).The red light showed the least preferred light by P. fuscipes compared to other available light colours with a mean of 8.33 ± 0.67 (Figure 5).Paederus fuscipes were observed to show higher preference towards LED green light with a mean of 31.67 ± 4.98 individuals.However, post hoc analyses revealed that green, white and blue light colours have equal chances to be attracted by P. fuscipes beetle more than red light (Tukey's HSD test, P < 0.05).Overall, based on colour attraction, P. fuscipes beetles were attracted to green > white > blue > red.

Light colour wavelength
Spectrographic results (Figure 6) show the spectral properties of the fluorescent, cool white LED and incandescent bulbs.Fluorescent light showed multiple peaks within the RGB (Red, Green, Blue) wavelength.The LED (Valens) emits a lower reflectance intensity peaking at 450 and 550 light wavelengths within both blue and green spectral wavelengths.Conversely, incandescent bulbs radiated light at 580 nm wavelength onwards, ranging from yellow to red spectral wavelength.
All LED (Philips) coloured light bulbs used in the colour preferences study of P. fuscipes showed peak sensitivity at blue (λ max = 455 nm), green (λ max = 550 nm) and red (λ max = 635 nm) respectively.The normal white LED light emits wavelengths all across the visible light spectrum (Figure 7).

Discussion
Our results showed that most P. fuscipes were observed to move towards the light condition regardless of light types; fluorescent, cool white LED or incandescent compared to the no light (dark) region.Furthermore, P. fuscipes showed greater preferences for fluorescent and cool white LED lights compared to incandescent light.This positive response suggests that P. fuscipes are strongly attracted to light.From this result, we can postulate that artificial light sources from human residential premises are a possible attractant source for P. fuscipes' infestation [22] Hence, the usage of artificial light for illumination at human residential premises during nighttime had caused P. fuscipes to start congregating around the housing area, mainly nearby the light sources [23].This inadvertently caused a significant rise in the risk of dermatitis linearis through an act of human response by trying to kill or remove the beetle physically once the beetle accidentally landed on human's skin [24].The previous study also mentioned that most individuals affected with dermatitis linearis were found close to artificial light sources at night [25].
Certain light wavelengths were found more attractive to P. fuscipes beetle compared to others.The present work on the coloured light experiment revealed that P. fuscipes responded positively to the green wavelength (λ max = 550 nm) but showed minimal response to the red wavelength (λ max = 635 nm) of the visible light spectrum.Fluorescent and all LED light bulbs used in this study emit lights throughout the visible spectrum ranging from 400 to 700 nm.Electromagnetic radiation in this range of wavelengths is called visible light or simply light visible to both humans and insects' perception  [26].However, the tungsten filament in an incandescent bulb produces electromagnetic radiation with longer wavelengths extending from the nominal red edge of the visible spectrum at 700 nm onwards, which mostly emit light in the infrared bands [27].
Advancement in the light industry has produced LED light bulbs, replacing the traditional lighting technologies such as fluorescent and incandescent light sources.However, with the development of the new LED light, our results showed that P. fuscipes were similarly attracted to both LED and fluorescent lights compared to lights emitted by the tungsten filament incandescent bulbs.According to Proctor et al. [28] and Cohnstaedt et al. [29], most insects have little or no sensitivity to red and infrared regions at the higher end of the light spectrum.This is because incandescent emit light most strongly in the infrared spectra and weaker in the visible light [29].Our results thus fulfil the statement above, which showed that fewer beetles were attracted to the incandescent light bulb.
In a more recent study conducted by Lima et al. [30], about 15.22% of individuals of the Paederus species were collected near incandescent lights.According to Cohnstaedt et al. [29], approximately 94% of heat or infrared radiation is emitted by a clear tungsten incandescent bulb, while the amount of light emitted was only 6%.Thus, the amount of current passing through the glowing filaments of the incandescent bulbs generated an increasing amount of heat which may have attracted P. fuscipes towards the incandescent bulb.Therefore, we postulate that the attraction towards incandescent light bulbs might be solely by the heat generated rather than the spectral properties and probably of the UV radiation produced by the incandescent light.While only 6% of the total light emitted is visible, UV radiation makes up a sizable portion of the remaining 94% [31].
Moreover, a previous study has shown that P. fuscipes contact with humans at night was particularly due to their attraction towards fluorescent and incandescent lights [32].However, the majority of dermatitis linearis cases occurred when fluorescent lighting was used during nighttime instead of incandescent light.According to a study conducted by Gopal [9], about 73% of the affected patients used fluorescent light at night, while the affected 27% used incandescent light for lighting.In another example, 268 factory workers in China were diagnosed with dermatitis linearis due to favourable fluorescent lighting conditions [33].In Malaysia, outbreak cases were frequently reported in residential premises with fluorescent lighting [8,22].The study by Lima et al. [30] showed that Paedarus species in Brazillian savanna are also attracted to the fluorescent and incandescent lights incredibly black light during nighttime.Here, the current study also found that P. fuscipes is primarily attracted to light emitted by the fluorescent bulb.
Essentially, insects are capable of distinguishing colours using their photoreceptor cells as each insect has specific spectral sensitivities [16].Although very little is known about P. fuscipes visual properties, we found that green, white and blue light colours were considerably more attractive to beetles compared to red light.However, P. fuscipes exhibited a greater preference for LED green light with maximal peak sensitivity at 550 nm in the colour preference study.Even though there was no statistical difference between green and blue light colours, the present study showed that P. fuscipes higher preference for green instead of blue was highly unexpected considering that most nocturnal insects are attracted to lights of lower wavelengths such as in the blue region of the visible spectrum emitted by the LED light.
Preferences for the light of certain spectral properties are associated with behavioural contexts of insects.According to Briscoe and Chittka [16], the physiological needs of each insect species are to determine the range of wavelengths that are attractive or repulsive.This behavioural feature offers great potential for individual evolutionary adaptation.Light sources that exert the strongest nocturnal insect attraction emit shorter light wavelengths with high UV emission [34].Jang et al. [35] stated that the initiation of insect dispersal flight is mainly elicited by the shorter light wavelength of the visible light spectrum.Short-wavelength photoreceptors play a significant role in insect's flight orientation [36].Although P. fuscipes are drawn to the heat than incandescent bulbs produced, once beetle was given a choice between two types of light sources, incandescent light was least attractive.Prominently higher response rates and orientation proportion were observed towards fluorescent light followed by the cool white LED light source.In which, our current study suggested that higher light intensity and shorter light wavelengths emitted by cool white LED and fluorescent lights are major factors in attracting P. fuscipes.As P. fuscipes responded strongly to the green and blue hue emitted by these two light sources, we strongly agree with other authors that insects are attracted to lights of shorter wavelengths.This proposes that the spectral reflectance that peaks within the 400-550 nm may be particularly attractive for P. fuscipes beetles.
In our present study, the spectrographic analysis showed that the spectral properties of the normal white LED reflect lights from three monochromatic sources consisted of the blue, green and red wavebands.Hence, it is apparent that P. fuscipes are drawn more towards the white compared to the blue LED light because they responded positively to the green wavelength emitted by the white LED.According to Steigerwald et al. [37], white LEDs were produced by using a mixture of various light colours to produce the white colour.As a result, light reflected from the white LED bulb is composed of all the colours of the visible spectrum.
Insects were assumed unable to detect light in the red spectral-domain greater than 600 nm [28].However, the present results showed that P. fuscipes could still perceive light within the 600-650 nm regions, despite the fact that red LED being the least attractive to P. fuscipes beetles.Our observations concur with Ashfaq et al. [38], who found that red light is unattractive for most insect orders.According to Prokopy and Owens [15], besides receiving light emitted by the sun during sunrise and sunset, insects also receive reflected light from beneath the soil or from the surrounding vegetation.Paederus fuscipes which are found widely distributed in the rice field areas [2] are exposed to radiation transmitted by plants.Chlorophyll is responsible for the dominant reflectance of green hue between 500 and 580 nm range (λ max = 550 nm).Therefore, most insects from the plant canopy have apparently conferred a degree of preadaptation to the colour green [15].This theory is supported by Briscoe and Chittka [16] who specified that the environment may have shaped an insect's photoreceptor spectral sensitivity.Hence, Paederus strong preference for green light is hypothesized to be related to its presence in the rice field surrounded by the greenery vegetation.
Other than that P. fuscipes are known for their ecosystem service as they serve as biological agents of insect pests in the rice field areas [2].Manley [2] indicated that the most preferred prey for the adult beetles were the green leafhoppers, whereas the Nephotettix virescen green leafhoppers are dominantly found in Asian tropical rice fields [39].Thus, as P. fuscipes receptors are maximally sensitive to the colour green, it is advantageous for this beetle to detect their food source, the green leafhoppers, easily.

Conclusions
In conclusion, our study would help to enhance integrated pest management programmes as an alternative for controlling P. fuscipes beetle's infestation in human residential areas at night.Like any other kind of insect, a shorter light wavelength is also attractive to P. fuscipes beetles.Hence, cool white and daylight bulbs should be avoided and replaced at every entryway with warm light colours or light sources that emit higher light wavelengths.Not only by using warm lights such as yellow (warm white LED) with red illumination are less attractive to most insect species, but they also appear as decorative lighting devices.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.The Y-maze diagram of 5 cm in width, 5 cm in height, 20 cm arm length in size, with an angle of 60°apart between both arms and base of 25 cm in length.

Figure 2 .
Figure 2. The response rate of P. fuscipes to two light conditions in the light and dark choices test among fluorescent, LED and incandescent light sources using binomial test design.Scale bar with the same letters showed no significant differences (P > 0.05).

Figure 3 .
Figure 3.The response rates of P. fuscipes towards two light source choices between fluorescent vs incandescent, LED vs incandescent, and fluorescent vs LED lights using binomial test design.Scale bar with the same letters showed no significant differences (P > 0.05).

Figure 4 .
Figure 4.The cumulative response rate of P. fuscipes towards three types of light sources among fluorescent, LED and incandescent light sources using binomial test designs.Scale bar with the same letters showed no significant differences (Tukey's HSD; P > 0.05).

Figure 5 .
Figure 5.The response rate of P. fuscipes to four LED coloured light sources using binomial Y-maze test design.Scale bar with the same letters showed no significant differences (Tukey's HSD; P > 0.05).

Figure 6 .
Figure 6.The light wavelength of fluorescent, LED, and incandescent bulbs using spectrographic analysis.

Figure 7 .
Figure 7. Spectrographic analysis of light wavelength for LED white, blue, green and red lights.

Table 1 .
Paederus fuscipes preferences toward three different types of light sources (fluorescent, cool white LED, and incandescent) in binomial test experiment 1; a light and no light (dark) choice condition.
Note: Significant values are in bold.

Table 2 .
Paederus fuscipes preferences toward three different types of light sources (fluorescent, cool white LED, and incandescent) in binomial test experiment 2; a light and light choice tests.
Note; Significant value is in bold.

Table 3 .
Paederus fuscipes preferences toward four different LED light colours (blue, green, white, and red) in binomial test experiment 3.