Mass Trapping of Live Male Bactrocera dorsalis Hendel (Diptera: Tephritidae): Interaction between Trap Types and Period of Trapping

. Mass trapping of insects involves the use of specifc traps or devices that are designed to attract and capture a large number of adult insects, thereby reducing their population in the agricultural environment. Tis approach aims at disrupting the breeding cycle and proliferation of the target insect to minimize crop damage and economic losses. Tis study evaluated the efciency of three trap types (Ecoman, Tephri, and Bucket funnel) and period of the day (morning, afternoon, and evening) in mass trapping of live male Bactrocera dorsalis and the survival of such catch over a period. Te traps were deployed on selected trees following 3 × 3 factorial experiments. All the Bactrocera dorsalis caught in the traps were males. Ecoman traps caught the highest number of B. dorsalis in the mornings and evenings while Tephri traps caught the highest B. dorsalis in the afternoons. Tere was no signifcant diference in the survival of B. dorsalis in the three trap types after 24hours. Correlation analysis showed that climatic factors played a signifcant role in trap catches. Bactrocera dorsalis caught in the evenings recorded higher percentage survival in all the three trap types than those caught during the morning and afternoon. Te Ecoman trap was most efcient in trapping B. dorsalis . Hence, this study recommends the Ecoman trap for mass trapping live male B. dorsalis meant for further experiments such as irradiation studies.


Introduction
Insect trapping system is one of the most common tools for bio-systematics studies and biological control of fruit fies.Te efciency of trapping systems varies according to the trap type, type of lure and its concentration, population density of the target insect, weather conditions, and trap placement such as height from ground and orientation [1].Te mass trapping method represents a preventive control measure which is based on attracting and killing adult fruit fies before they get to the fruit to lay eggs.Te key advantage of the mass trapping method is the exclusion of fruit fies and the avoidance of whole canopy contamination by insecticides.Mass trapping methods can be applied by using traps of diferent constructions, which must be set on tree canopies or anywhere within an orchard.Te traps may contain diferent types of attractants and a killing agent which is usually an insecticide [2][3][4][5][6].It has been suggested that mass trapping has a better efcacy than bait sprays and has a lower cost of application [3,[7][8][9][10].Trap designs, including diferent colours and shapes, are essential to obtain a high efciency in fruit fy catches [11][12][13][14][15]. Afrotropical fruit fy pests in the genera Bactrocera, Dacus, and Zeugodacus (subtribe Dacinae) are known to respond to either methyl eugenol (4-allyl-1, 2-dimethoxybenzenecarboxylate) or Cue-Lure (4-(p-acetoxyphenyl)-2-butanone) [16,17].
Trapping a large proportion of the wild population of an insect pest is essential for the success of sterile insect technique (SIT) programs which require the release of sterilized feld-trapped adults to control wild populations [18].For SIT and many other control strategies such as gene drive techniques, cytoplasmic incompatibility techniques, and biological control that require the release of insects, it is essential to estimate the size of the wild population of the target insect before employing such strategies.Mark-releaserecapture is the most efcient method for estimating the size of wild populations.Te ability to trap live insects, release them in good condition to mix with the wild population, and be recaptured is also essential in SIT.Terefore, it is important to optimize the trapping of live insects as part of key pest management strategies.Te type of trap to be used for mass trapping of live insects is important as the capacity for capturing live insects may difer among diferent trap types.Hence, there is a need to evaluate diferent traps to determine their efciency for trapping live insects.Specifc trapping systems should be used depending on the objectives of specifc pest control programs.Factors including economic and technical feasibility, species of fruit fy, and the phytosanitary condition of the areas of concern could be considered before employing SIT [19].
In this study, three diferent traps for capturing males of Bactrocera dorsalis and the ideal period of the day for mass trapping live adult males meant for further studies were evaluated.Te study also investigated the infuence of weather conditions (temperature, humidity, and rainfall) on trap efciency.

Study Location.
A major mango (Mangifera indica L.) growing area (Manya Krobo District) in the Eastern Region of Ghana was selected for this study due to its history of being a home to many mango varieties, both local and exotic.Fruit fy sampling was performed in Power of Trinity Farm (POT) (6 °6′12″N; 0 °0′7″W) (Figure 1) from March 2019 to June 2019.Te POT farm houses a mixture of Keitt and Kent varieties of mangoes.All mango trees in the farm had exceeded the economic fruit-bearing age of seven years at the time of data collection.

Trap Types.
Trapping was performed according to International Atomic Energy Agency (IAEA) guidelines [19] with few modifcations.Tree diferent traps, that is, the Ecoman fruit fy trap (Ecoman Biotech, Beijing, China), Tephri trap (SORYGAR, Madrid, Spain), and Bucket funnel trap (Insect Science, Tzaneen, South Africa), were evaluated to ascertain their efciency and efectiveness in collecting a large number of adult males of Bactrocera dorsalis.Captured B. dorsalis males from each trap type were tested for their survival under laboratory conditions (i.e., 25 ± 1 °C; 75 ± 5% RH; 12D:12L photoperiod).
Te Ecoman fruit fy trap (Figure 2(a)) is a vertical cylinder, with a height of 17 cm and an average diameter of 5.83 cm at the top entrance.Te total volume of the trap is about 449 ml.It has a white translucent bottle and a black cap (0.65 cm high and 0.70 cm wide) which can be unscrewed to facilitate servicing.Te black cap is domeshaped, with four spiral entry points (each 0.13 cm in diameter).A plastic pin (0.6 cm in height) attached to the inside of the cap serves as a holder for the attractant.Te trap was hung on a hook on top of the dome-shaped black cap.
Te Tephri trap (Figure 2(b)) is similar to a McPhail trap.It is a vertical cylinder with a height of 10.6 cm and a diameter of 11.6 cm at the base and can hold up to 1,119 ml of liquid.It has a yellow base and a clear top, which can be separated to facilitate servicing.Tere is a platform inside the top to hold attractants.A nylon thread, placed on top of the trap body, was used to hang the trap on tree branches [19].
Te Bucket funnel trap (Figure 2(c)) consists of a tapered upper yellow pane (the funnel), white lower collection bucket, green lid, white caps, and green pheromone basket/ cage.It is also a vertical cylinder, with a height of 12.6 cm and a diameter of 12.8 cm.It can hold up to 1,621 ml of liquid.Te green lid has two holes on top with a thread for hanging the trap.
Te traps were hung on the mango trees using a nylon thread.Grease was applied to the frst one-third proximal part of the thread near the branch to prevent ants from preying on the captured B. dorsalis.Traps were rotated monthly to prevent the location of a trap from interfering with its performance.

Attractant.
Methyl eugenol (ME, Scentry Biologicals, Inc, Billings, MT), a known attractant of adult males of B. dorsalis over long distances, was placed in a slow-releasing polymeric gel form [20,21].No killing agent was added to the attractant in the traps because the captured fies needed to be kept alive and their survival monitored.To avoid contamination from other odour sources, only new traps were used.Te applicators' hands were also covered with disposable latex gloves during trap handling and placement.

Fruit Fly Sampling and
Monitoring.An area of 12,141 m 2 within the 32,375 m 2 farm was demarcated as the sampling area for fruit fy collection.Te mango trees were selected systematically to cover the area uniformly.Within the sampling area, three blocks were demarcated 10 m apart.Within each block, a total of 15 trees, 9 m apart, were systematically selected and tagged.Fifteen traps (one per tree) were deployed on the selected trees at a height of 2.0 m above ground depending on the architecture of the tree [22].Te deployment of traps followed a 3 × 3 factorial arrangement.Te 3 × 3 factorial arrangements multiplied by the three blocks resulted in twenty-seven experimental units for the sampling area (12,141 m 2 ).Te 15 traps per block consisted of fve Ecoman traps, fve Bucket funnel traps, and fve Tephri traps.Methyl eugenol polymeric gel was used as an attractant in the traps.Te traps were placed in semishaded and upwind parts of the canopy at 6:00 am on 2 Psyche: A Journal of Entomology sampling days.Te traps were left in the feld for 3 hours each in the morning, afternoon, and evening.Flies caught between 6:00 am and 9:00 am were designated as morning catches.Tose caught between 12:00 noon and 3:00 pm were designated as afternoon catches and those caught between 4: 00 pm and 7:00 pm were designated as evening catches.Te catches of each trap for each designated period were carefully emptied into a cage.Te fies were provided with enzymatic yeast hydrolysate, sugar (three parts yeast: one part sugar), and distilled water soaked in cotton wool in a small vial inside the cage [22].Another cotton wool soaked in water was placed on top of each cage to keep the catches hydrated and the cage humid.While on the feld, each cage was labeled with the respective trap type and period of catch and placed under shade.At the end of the day, the catches were transported under a temperature condition of 20 °C to the laboratory for further studies.Te fies were monitored for 30 days under laboratory conditions for survival or Psyche: A Journal of Entomology mortality.An artifcial diet (three parts yeast: one part sugar) and water were provided to fruit fies that were confrmed to be male B. dorsalis ad-lib.Fruit fy sampling was replicated three times over the periods of March, May, and June 2019.

Taxonomy and Identifcation of Captured Fruit Flies.
Te captured fruit fies were confrmed as adult males of Bactrocera dorsalis based on morphological characteristics using taxonomic keys developed by the African Fruit Fly Initiative [22].Te fies were viewed under a dissecting microscope (GX Microscopes, GT Vision Ltd, Sufolk, UK) at a magnifcation of 20×.Nontephritid fies were identifed to order or family levels but were not used in this research.Samples of the identifed insects were deposited at the Radiation Entomology and Pest Management Centre under the Biotechnology and Nuclear Agriculture Research Institute of the Ghana Atomic Energy Commission.

Climatic Data. A TinyTag data logger (Gemini Data
Logger Ltd, West Sussex, UK) was installed in the orchard to collect climatic data.It was launched to collect the temperature, relative humidity, and dew point readings every 60 minutes.At the end of the day, it gives the daily summary.Te daily summary of the temperature and relative humidity was then converted to mean monthly summaries.Te rainfall reading was collected using PS, PP Rain Gauge (Shreeji Instrument, Gujarat, India).Te rain gauge gives a weekly reading which is converted into mean monthly summaries.Te rainfall pattern in the study area was bimodal.Tus, two peaks of rainfall events occurred in the study area, one in the major cropping season (beginning from May to July) and the second in the minor cropping season (beginning from September to November).

Statistical Analysis.
Te number of B. dorsalis adult males captured by the traps was subjected to analysis of variance (ANOVA).Te efect of trap type and period of the day on the number of B. dorsalis male catches and the percentage survival were subjected to Fisher's test.Data were input into Microsoft Excel to generate nine samples for the trap type and period of the day, representing the sample size (n) used in ANOVA for the single factor efect using Genstat software [23], by selecting a general treatment structure to run the 3 × 3 factorial experiment.Data were logtransformed to normalize the initial distribution of raw data collected for males of B. dorsalis catches and the percentage survival of fies in traps before performing ANOVA.Correlation and regression analyses were performed between the trap catches and climatic data measured during the study.

Response of Adult Male Bactrocera dorsalis to Trap Types.
All adult B. dorsalis caught were males.Te trap catches showed that the Ecoman trap was the most efcient in catching large numbers of adult males of B. dorsalis (Figure 3).Te Tephri trap also caught more male B. dorsalis compared to the Bucket funnel trap.Te weekly catches were signifcantly diferent among the trap types (F (2,15) � 26.44, p ≤ 0.001) (Figure 3).
Tere were no signifcant diferences in the survival of adult male B. dorsalis in the three trap types under study during a 24-hour study period (F (2,14) � 0.08, p ≤ 0.924).However, numerically more male B. dorsalis survived in the Ecoman trap followed by the Bucket funnel trap and Tephri trap (Figure 4).
Survival of captured B. dorsalis adult males in Ecoman traps was highest in the evening captures followed by the afternoon captures before the morning captures albeit not statistically diferent (Table 1).A similar trend was observed in the captures by the Bucket funnel trap.In the Tephri trap, the highest survival was in the evening, followed by the morning before the captures in the afternoon (Table 1).Tere were no signifcant diferences in the percentage survival of B. dorsalis males that were captured in the three trap types at the three diferent periods of the day after 24 h in the insectary (F (2,14) � 0.10, p < 0.979) (Table 1).

Interaction between Trap Catches and Period of the Day.
Tere were signifcant diferences in the interaction between the trap types and the period of the day (F (4,15) � 6.69, p ≤ 0.003).In the morning, Ecoman traps caught higher numbers of B. dorsalis adult males than Tephri and Bucket funnel traps.Again, in the morning the number of B. dorsalis males caught by Tephri and Bucket funnel traps was not signifcantly diferent (Table 2).In the afternoon, there were no signifcant diferences in the mean catches by Ecoman, Tephri, and Bucket funnel traps (Table 2).In the evening, Ecoman traps were efcient in catching large numbers of male B. dorsalis compared with Tephri and Bucket funnel traps.However, the number of B. dorsalis males caught by Tephri and Bucket funnel traps was not signifcantly different (Table 2).Moreover, in the Ecoman traps, there were signifcantly more catches in the morning than in the afternoon.However, the catches in the afternoon were not signifcantly diferent from those in the evening and the catches in the morning were also not signifcantly diferent from those in the evening (Table 2).In the Tephri traps, the catches in the morning and afternoon were not signifcantly diferent but both were signifcantly higher than the catches in the evening (Table 2).Te catches in the Bucket funnel traps were not signifcantly diferent among all three periods of the day (Table 2).

Response of Adult Males of Bactrocera dorsalis to Period of the Day.
Tere were signifcant diferences in the mean number of adult male B. dorsalis caught at the diferent periods of the day (F (2,15) � 9.24, p ≤ 0.002).Te number of B. dorsalis males caught in the morning was signifcantly higher than the numbers caught in the afternoon and evening.However, the catches in the afternoon compared to those in the evening were not signifcantly diferent (Figure 5).Psyche: A Journal of Entomology Tere were signifcant diferences in the percentage survival of B. dorsalis adult males at diferent periods of the day within a 24-hour period (F (2,14) � 8.83, p ≤ 0.003).A signifcantly higher number of males of B. dorsalis survived in the evening than in the morning.Similarly, a signifcantly higher number of the fies survived in the traps set up in the afternoon than in the morning.However, there was no signifcant diference between the fies caught in the afternoon and evening (Figure 6).

Interaction between Trap Catches and Weather
Parameters.Te average rainfall, temperature, and relative humidity for the frst trapping period in March 2019 were 45 mm, 31 °C, and 72%, respectively.Te average rainfall, temperature, and relative humidity for the second trapping period in May 2019 were 145 mm, 30 °C, and 77%.In June, the average rainfall, temperature, and relative humidity for the third trapping period were 195 mm, 28 °C, and 80%, respectively.
Correlation analyses between adult males of B. dorsalis trap catches and weather parameters showed no signifcant diferences.Temperature (r � 0.6638; p > 0.0668), relative humidity (r � 0.6192; p > 0.0754), and rainfall (r � 0.6182; p > 0.0760) were positively correlated with Ecoman traps.Even though Ecoman trap catches had a positive correlation with climatic factors, there were no signifcant diferences among them.Tere was a strong positive correlation observed between the weather parameters and the B. dorsalis adult male catches for the Tephri trap: temperature (r � 0.7766; p < 0.0138), relative humidity (r � 0.7220; p < 0.0281), and rainfall (r � 0.7196; p < 0.0138).A similar correlation was observed for the catches in the bucket funnel traps.Tere was a strong and signifcant correlation between catches in the bucket funnel traps and temperature (r � 0.7286; p < 0.0404) as well as relative humidity (r � 0.7001; p < 0.0354).On the contrary, the correlation between catches in the bucket funnel traps and rainfall was not signifcant (r � 0.6705; p > 0.0688).

Response of Adult Male Bactrocera dorsalis to Trap Types.
Adult fruit fies are normally monitored with traps containing attractants [19,24,25].Te choice of trap for mass trapping depends mainly on the type of attractant to be used [19].In the present study, adult males of B. dorsalis were collected at three diferent periods during the day (i.e., morning, afternoon, and evening) in mango orchards using three diferent trap types with methyl eugenol as the attractant.Methyl eugenol was chosen because it is a known attractant of adult males of B. dorsalis, the target pest [17,20].Males of B. dorsalis collected in the evening with the baited traps had a higher survival rate for the frst 24 hours after trapping than those collected in the morning and afternoon.Tis could be due to the more favourable weather conditions in the evening.Te evenings had relatively low temperatures (28 ± 1 °C) and high humidity (77 ± 5%) that might have aided the survival of the captured fruit fies in the traps.
Te mean percentage survival of B. dorsalis male catches in the Ecoman trap was 84%, followed by Bucket funnel trap (76%) and Tephri trap (74%).Motswagole et al. [26] and Choi et al. [27] reported 16.7 °C to 34.9 °C as possible optimum temperatures for the survival of fies during the period Means followed by the same letter within the column are not signifcantly diferent.Psyche: A Journal of Entomology of capture.Te success of the mass trapping strategy depends on the efciency of traps and lures for collecting live insects [28].Te traps used in the experiments posed little or no harm to the environment because no killing agent was added to the traps [29][30][31][32].Methyl eugenol is highly attractive but very specifc in attracting fruit fies in the Bactrocera complex including B. dorsalis.In fact, earlier studies have shown that methyl eugenol is very efective in mass trapping Bactrocera species in mango orchards [33,34].Our fndings suggest that the best trap to capture live adult male B. dorsalis for further studies is the Ecoman trap.

Response of Captured Male Bactrocera dorsalis to Period of the Day.
In all the three types of traps, more than 80% of B. dorsalis survived in the traps set in the evening.Tis implies that, it is advisable to set the Ecoman trap in the evening for mass trapping of live adult males of B. dorsalis since it had the highest percentage survival of the fies.Our fnding is in line with those of Siddiqui et al. [34] who stated that fruit fies exhibit a wide range of diurnal activities.Measures to control this pest should be adopted during the morning and evening hours.Tis is premised on the fact that the fruit fies were found to be most active at 10:00 am−11:00 am and that adult fies spent much of the day resting on other plants in the vicinity of host crops [35].Large populations of adults were found on host plants before 8:00 am and after 5:00 pm.Te observations in this study are in close agreement with those reported in earlier studies which suggested that B. cucurbitae (Coquillett) and B. dorsalis are active in the morning [36].In the current study, B. dorsalis showed an activity peak between 7:00 am and 8:00 am.Moreover, Rizk et al. [1] reported high mean catches of peach fruit fies, Bactrocera zonata (Saunders) between 5 am and 7 am, which is usually their mating period.

Interaction between Trap Type and Period of the Day.
Tis study has demonstrated that the number of male B. dorsalis captured is infuenced by trap type and period of capture.Te Ecoman trap captured and retained a large number of adult male B. dorsalis in the mornings and evenings.Te construction of the Ecoman fruit fy trap is such that the entry holes into the traps are spiraled, preventing trapped fies from escaping.Te Tephri trap used in this study was a modifed McPhail trap.Te entrance holes were around the top of the yellow base, and an invaginated opening was in the bottom.Tis design probably allowed captured fies to escape since no killing agent was incorporated.Clearly, fndings from this study have demonstrated that Tephri traps are not suitable for capturing live fies intended for further studies e.g., irradiation studies.Te Bucket funnel trap had a wide space between the upper yellow pane and the white bucket.Tis window allowed captured fies to escape easily.Terefore, the Bucket funnel trap is also not suitable for retaining captured fies if no killing agent is incorporated.Findings from this study support earlier studies that suggest that trap designs, including diferent colours and shapes, are essential to obtain a high efciency in fruit fy catches [11,12,14,15].Similarly, several studies have reported that tephritid fruit fy traps vary in efectiveness depending on their size, colour, shape, and the particular olfactory attractant used [12-14, 31-33, 37-39].Te type of trap is important in mass trapping.Tis study and others have demonstrated that the Ecoman trap is good for collecting large numbers of Means followed by diferent letters within columns are signifcantly diferent at p < 0.003 (Fisher's test).Psyche: A Journal of Entomology B. dorsalis males due to its trapping efciency [40].If this trap is used in the right period of the day as demonstrated in this study, it will be very efcient in mass trapping live adult males of B. dorsalis for studies or programs that require live fies.

Morning
Studies have also demonstrated that adult fruit fies can be attracted to specifc chemical lures during the daytime [41].Although Ecoman traps can retain captured fruit fies very well, the fies should be released in cages within a relatively short period after capturing due to inadequate ventilation in the trap.Findings from this study show that traps set in the afternoons have low catches.Tis could be because of the high temperature during that period of the day and that male B. dorsalis fnd suitable refuges away from the heat of the sun.

Interaction between Trap Catches and Weather
Parameters.Abiotic factors infuence the capture of fruit fies by baited traps.In this study, there were positive correlations between the three trap types and climatic factors such as temperature, relative humidity, and rainfall.Tis implies that an increase in temperature, relative humidity, and rainfall increases the number of fruit fies that are captured in traps and vice versa.
Ecoman trap catches had a positive correlation with climatic factors although there were no signifcant diferences among them.Te Ecoman trap did not allow any captured fy to escape.Te other trap types allow fies the possibility to escape.Similar fndings were reported in a study in which the number of fruit fies captured with Cue-Lure-baited traps correlated positively with all three abiotic factors, i.e., temperature, humidity, and rainfall [42].Variations in weather conditions play a vital role in the multiplication, growth, development, and distribution of insects.Tese also infuence their population dynamics [43].
Tephri traps on the other hand exhibited a strong positive correlation with all the climatic factors recorded.Tis supports the fndings of Khan et al. [44] who stated that weather parameters have a signifcant efect on the population dynamics of fruit fies.Temperature and rainfall were reported to be the most important factors infuencing the population dynamics of fruit fies [45].
For the Bucket funnel traps, both temperature and relative humidity had a signifcant positive correlation except rainfall, which had a positive but nonsignifcant correlation.Bana et al. [45] and Patel et al. [46] demonstrated a positive correlation between temperature, relative humidity, and rainfall on fruit fy catch.Even though earlier fndings have demonstrated a similar relationship between fruit fy infestation and weather factors, Adzim et al. [47] reported that in the coastal grassland area, there was a negative correlation between B. dorsalis for both rainfall and temperature.

Conclusion
Te three trap types evaluated had diferent shapes, colours, and designs, which made them unique in collecting large numbers of B. dorsalis adult males.Survival was higher for males of B. dorsalis trapped in the evening in all the three trap types.However, the Ecoman trap proved to have a higher trapping efciency than the Tephri trap and Bucket funnel traps.Ecoman traps are therefore ideal if one is interested in mass trapping live adult males of B. dorsalis or for programs that require live fies e.g., irradiation studies.Climatic factors infuence the catches by the traps diferently.Terefore, when installing traps intended for mass trapping of live fruit fies, the optimum temperature, relative humidity, and rainfall should be considered to boost the efciency of the traps.On the average, the survival rate of adult males of B. dorsalis in the traps is highest for Ecoman, followed by Bucket funnel and Tephri trap.Tis study provides useful knowledge for the development of trapping systems that require live fies.

Figure 3 :Figure 4 :
Figure 3: Mean catches of adult males of Bactrocera dorsalis in three diferent trap types.

Figure 5 :Figure 6 :
Figure 5: Mean catches of adult males of Bactrocera dorsalis at diferent periods of the day.

Table 1 :
Percentage survival of adult males of Bactrocera dorsalis in three trap types after catches in 24 hrs on a mango plantation.

Table 2 :
Trap catches of adult males of Bactrocera dorsalis at diferent periods in the day.