Distribution, development biology and behavior of Dacus persicus associated with Calotropis gigantea in Sri Lanka

: Calotropis gigantea (Crown flower, Giant milkweed or Wara) is a native medicinal plant in Sri Lanka. It is recorded as an invasive plant in Australia, Brazil, USA, etc. Dacus persicus is recorded as a highly destructive monophagous pest of C. gigantea in Sri Lanka. Larvae of D. persicus feed on developing fruits and seeds and reduce the reproductive output of the plant significantly making it a suitable candidate for biocontrol. Therefore, the present study was aimed to investigate the distribution and reproductive biology of Dacus persicus to assess the potential as a biocontrol agent for Calotropis species. D. persicus distributed in six provinces in Sri Lanka. The duration of mating and ovipositing of D. persicus was 54 and 92 minutes, respectively. It laid eggs in the seed chamber of developing fruits and the fruit size is highly correlated ( p < 0.001, r = 0.990) with the number of laid eggs. Only one egg cluster of D. persicus found within a single fruit having 18.5 (± 0.85) eggs per cluster and the eggs hatched in 3 days. The duration of larval and pupal stages for D. persicus were 24 and 12 days, respectively. These results provide essential information needed in adopting D. persicus as a biocontrol agent of C. gigantea .


INTRODUCTION
Calotropis gigantea (Apocynaceae) (L.) W.T. Aiton. (Crown flower, Giant milkweed or Wara) is a tall plant or a shrub which grows up to 2.4 -3 meters (Singh et al., 2014). It consists of well developed, branched root system and all parts of the plant consist of whitish, thick latex (Kumar et al., 2013). It is native to India, China, Sri Lanka, Malaysia (Dhileepan, 2014), Bangaladesh, Pakistan, Indonesia, Camboidea, Thailand and Philippines (Islam et al., 2019) and also distributed in several countries in South Asia, the Middle East, Africa and in North and South America (Kumar et al., 2013). It is an exotic, invasive species in Australia, Virgin Islands of United States, Mexico and Brazil (Kumar et al., 2013;Amutha et al., 2018). Plants grow well on abandoned over-cultivated area; over-grazed grounds, roadsides, lagoon edges and disturbed sandy soil (Kumar et al., 2013).
Calotropis species are subjected to herbivory by several insect and mite species in its native range.
According to Al dhafer et al. (2011), in Western region of Saudi Arabia, 24 species of insects are associated with Calotropis procera (L.) W.T. Aiton while in Central region 99 species of insects are associated with C. procera. Though latex of the C. procera plant is considered toxic to insects (Al dhafer et al., 2011), large number of insect pests cause considerable damage to the plant (Saikia et al., 2015). Dhileepan (2014) recorded sixty-five insect species associate with Calotropis spp. in their native range. Among them, more than 50% of insects feed on leaves while others feed on flowers, stems, seeds, and fruits. Most of the insects associated with Calotropis species were recorded from India (Dhileepan, 2014). In contrast, there are few insect species found feeding on the Calotropis species in its introduced countries. Aphids, grasshoppers, and caterpillars of Danaus spp. are considered as common plant feeders associated with C. procera (Al dhafer et al., 2011). Danaus spp. are not considered as severely damaging pests in Australia, Hawaii, Fiji, Brazil, Jamaica, and Puerto Rico (Dhileepan, 2014). Most of the insects associated with C. procera are polyphagous while, Dacus persicus Hendel (Diptera: Tephritidae) is considered as a major insect pest specific to Calotropis species (Dhileepan, 2014).
Dacus persicus, commonly known as Aak fruit fly, is a monophagous insect of Calotropis species (Dhileepan, 2014). It is native to India, Sri Lanka, Iran, Pakistan, and Iraq. The closely related species, Dacus longistylus Wiedemann, is also a monophagous insect associated with Calotropis species. It is distributed in Sudan, Egypt, Israel, Saudi Arabia, Iran, Senegal, Kenya, Nigeria, Mali, Eritrea, Libya, UAE, and Niger (Dhileepan, 2014). D. persicus larva is one of the destructive seed feeders of Calotropis species (Sharma and Amritphale, 2008). Gravid D. persicus females lay eggs inside developing Calotropis fruits by penetrating the skin of fruit with its ovipositor (Parihar, 1984). Emerging D. persicus larvae feed and grow within the developing fruits. Infested fruits, rot and often drop prematurely. Fully developed larvae drop from the fruits into the ground and pupate in soil (Dhileepan, 2014). The damage caused by D. persicus drastically influence the reproductive output of the plant there by significantly reducing the plant dispersal and population growth.
When monophagous insect acts as a pest, normally it causes a considerable damage to the host plant. The ability to cause severe damage to their host plant, and being host specific are ideal characteristics that make the insect, the D. persicus as a promising weed biocontrol agent for Calotropis species in its introduced countries. In introduced countries, the plant is regarded as highly invasive and so far, there is no successful chemical or mechanical method to control the dispersal of the plant. Therefore, the weed biologists highly pay attention on classical biocontrol methods and focus on monophagous insects of Calotropis gigantea in their native range. As C. gigantea is a native plant in Sri Lanka and since no studies have been conducted on D. persicus of C. gigantea in Sri Lanka, the present study focuses on understanding damage caused to the host plant, distribution, mating, oviposition and life cycle of D. persicus in Sri Lanka with an aim to seeking the possibility of using it as a biocontrol control agent for Calotropis species in Australia and in other introduced countries where the plant is regarded as invasive.

Field sampling of Calotropis gigantea within Sri Lanka
Field visits were conducted in 120 sites in nine provinces of Sri Lanka (Figure 1) from December 2014 to June 2015. Sampling was conducted at monthly intervals and each site was sampled only once. Roadside sampling sites were selected randomly at 30 minutes intervals while traveling on a vehicle with a speed of 50 km per hour. If a new site with C. gigantea was not observed after 30 minutes, traveling was continued until a site having C. gigantea is found. In every sampling site, occurrence data of C. gigantea distribution (GPS coordinates) was recorded. In every site, associated insect fauna of C. gigantea were observed within 30 minutes time duration. During the survey, insects associate with the plant were photographed. The insects including D. persicus were collected directly from various parts of the plant (leaves, flowers, flower buds, stems, and fruits) by hand-picking. 2-3 individuals of same species in each site were collected to small plastic vials for morphological identification. The specimens were deposited in the Department of Zoology laboratory, at the University of Ruhuna, Sri Lanka. Specimens were identified up to genus/species level under the guidance of Entomologists of Entomology Division, The Horticultural Crop Research and Development Institute (HORDI), .Gannoruwa, Sri Lanka

Life cycle studies
In parallel to field studies, D. persicus was reared in the laboratory to study the mating and oviposition behavior, and durations of larval and pupal development. Male and females of D. persicus were collected from the study sites from Southern Province (directly by hand picking and placing them into small plastic vials) and transported to University of Ruhuna. D. persicus adults were maintained in transparent plastic boxes covered with insect-proof wire mesh material on top. Adult D. persicus was fed with bee honey and sugar solution.

Mating and oviposition behavior
To study oviposition behavior of D. persicus, fresh C. gigantea fruits having developing whitish seeds (fruit length 4.2 -10.7 cm/ fruit width 2.68 -7.64 cm) were placed in rearing cages. The fruits that were selected for oviposition were labeled and observed whether more than one fruit fly oviposit on a single fruit. The oviposition duration of each fruit fly was recorded. Mating behavior of D. persicus was recorded by observing activities of pre-mating, mating and post mating under the natural in the field. Similarly, under laboratory conditions,time duration for pre-oviposition, oviposition and post-oviposition behavior of D. persicus were observed by the naked eye.

Extraction of eggs of D. persicus
Two hundred and fifty (250) C. gigantea fruits having whitish-developing seeds (fruit length 4 -11 cm/ fruit width 2 cm) were collected monthly from 11 selected sites and transported to laboratory. Egg clusters of D. persicus were extracted from infected fruits under laboratory conditions. Before extraction, the maximum width, and the maximum length of the C. gigantea fruit were measured using a vernier caliper. The number of egg clusters per fruit and eggs per cluster were counted in the lab under Wild Heerbrugg stereo microscope (under the power of 15 × 40). To find any correlation between the volume of oviposited C. gigantea fruits and the number of laid eggs per fruit, the following equation was used.
4/3πab 2 where, As D. persicus lay eggs in immature, developing oval shaped C. gigantea fruits, an assumption was made as all the studied fruits were prolate ellipsoid in shape and the above equation was applied to gain the volume of C. gigantea fruit.

Larval development
Some of the extracted eggs, larvae, and pupae were placed back to same fruits and were reared to adults under laboratory conditions at 27 -30 • C. In high moisture conditions, mature fruits were rotten with live larvae within the fruit. In such conditions, these larvae were transferred to seed chamber of fresh fruits and allowed them to complete larval development and facilitate pupation. The replacement of rotten fruits with fresh fruits, was done until larvae turn into pupae.
During monthly samplings, C. gigantea fruits oviposited by D. persicus (n =100) were collected randomly from field sites. They were kept under laboratory conditions and different larval stages (until final larval instar) were extracted periodically by opening C. gigantea fruits.
To determine the number of larval instars, larval stages were placed on 70% alcohol (Ghafoor, 2011). In order to determine the larval stage, the maximum length of the head capsule of each larva of D. persicus was measured using a calibrated ocular micrometer in a binocular dissecting microscope.

Data analysis
Data analysis was done using Minitab 16 Statistical software package. Mean values and standard errors related to life history data of D. persicus was obtained. Correlations were developed for C. gigantea fruit size and numbers of eggs within the fruit.

Distribution of D. persicus in Sri Lanka
D. persicus was restricted to certain districts and it was not recorded in all districts where the plant was available. D. persicus was recorded in the coastal as well as in the inland areas ( Figure 1). However, there is no significant difference (Chi Sq -1.63, p value -0.201) in the presence of D. persicus in between coastal and inland sites. There were no records of D. persicus in Western province although the plant was available. There were no records of C. gigantea in Central Province, and therefore, D. persicus was absent in the Central region of Sri Lanka (Figure 1).

Temporal variation of D. persicus in Southern Province
D. persicus was recorded in all three districts of Southern Province. Comparatively higher abundance of D. persicus was recorded in Galle district while lowest abundance was recorded in Hambantota district (Figure 2). It is statistically also confirmed a difference in the mean abundances of D. persicus (p = 0.002) at least in two districts of Southern Province. Peak abundance of D. persicus was recorded during July to October while lowest abundance was observed from November to December. However no significant difference in mean abundances of D. persicus (p = 0.281) during study period.

Temporal variation of D. persicus across the sites of Galle, Matara and Hambantota
D. persicus was recorded in all sites except Kathaluwa site (Galle district) and Devinuwara (Matara district). Dadalla was the only site, where fruit flies were recorded throughout the year with high abundance (Table 1).

Life cycle studies of D. persicus
Mating and oviposition behavior of D. persicus D. persicus mated any time during the day (Figure 3). They mostly paired in mornings and rarely paired in afternoons (Figure 3). Gravid females of D. persicus oviposit only in developing fruits with immature, milky white seeds (Table 2). Female D. persicus highly attracted (51.62%) for immature fruits having 1-20 cm 3 volume. Interestingly, their preference for selection of a fruit for oviposition reduces a-Maximum length of Calotropis fruit / 2 b-Maximum width of Calotropis fruit / 2    into half when fruit size is doubled (Table 2). Statistically also proved that there is a strong correlation (r = -0.933, p = 0.020) between the volume of fruit and fruit selection for oviposition by D. persicus. Similarly, when volume of the fruit increases the amount of laid eggs laid also increase (Table 2) indicating a strong correlation (r = 0.968, p = 0.007) between the number of eggs of D. persicus within the Calotropis fruit.
D. persicus place eggs inside the seed chamber of C. gigantea fruit by directly piercing the fruit by its ovipositor (Figure 4). Field observations revealed, there was a great competition for selection of fruits for oviposition. Few females (4 -5 individuals) showed walking on such fruits and aggressive behavior (chasing fruit flies away from the fruit) among female fruit flies were observed. In such conditions, 2 -3 females of D. persicus oviposited in the same fruit. Even though, a female fruit fly was already in the oviposition, process, other (one or two) fruit fly females start their oviposition within the same fruit. The latter ones also expend approximately same time for the oviposition process, as initially oviposited fruit fly. Even though multiple oviposition of D. persicus females on the same fruit was seen; only one egg cluster per fruit was recorded. Personal experience revealed that, the prevailing egg cluster was closer to the oviposition punch of the initial female fruit fly.

Eggs, larvae and pupae of D. persicus
The eggs of D. persicus were pale whitish, delicate, elongate, slightly curved, tapering towards either side and arranged similar to a bunch of bananas (Figure 4). After 2.87 (± 0.62) days of oviposition, the first instar larvae of D. persicus was emerged. The developing larvae were creamy white in colour with brownish head capsule with mandibles and prolegs. D. persicus consisted of three larval instars. Survival of eggs was recorded as 64.38%.
The pupa of D. persicus was cylindrical in shape but rounded at both ends and dull creamy white in color with horizontal rings like ridges (Figure 4). D. persicus emerged from pupa at daytime was pale brownish in color ( Figure  4). Under laboratory conditions, about 59% of larvae reached pupal stage, and about 50% pupae emerged as adults. In laboratory conditions, newly emerged individuals represented the sex ratio as 1:1. In the field studies, it was recorded as female abundance was higher than males.

Damage levels of D. persicus
Larval development of D. persicus took place within the Calotropis fruit. With respect to larvae of Aak weevil (P. farinosa) (Wijeweera et al., 2020), D. persicus larvae voraciously fed on developing seeds of C. gigantea fruit causing severe damage to seeds. The present study revealed that all larval stages feed on seeds and destroy (100% of the seeds) the infected fruit. When it was about to pupate, larvae consume all the seeds of infected fruit. During this stage inner seed chamber appear as a decaying sponge and such kind of fruits can be easily identified through its external features such as pale yellowish, malnourished and stunted nature. However, adult D. persicus were harmless to C. gigantea as they feed on naturally occurring sugary compounds and under laboratory condition, they feed on artificial sugar solutions.

DISCUSSION
There is a close association between D. persicus and its host plant as it provides shelter, food, oviposition sites and mating grounds (Aluja and Liedo, 2013). When considering monophagous pests, they only survive if the host plant is available. Similarly, in this study also, the distribution of D. persicus is closely associated with its host plant C. gigantea (Figure 1). Dadalla was the only site where D. persicus appeared throughout the year. It may be due to resource availability around the sampling area. D. persicus was recorded in all sites of Hambantota district. It may be due to climate suitability of the district for survival of D. persicus.
Life history studies reveal that D. persicus oviposit on developing fruits. This may be due to two major reasons. Developing fruits are easy to penetrate and ensure the placement of eggs in the inner seed chamber and the middle fibrous layer. On the other hand, developing fruits having immature seeds; a suitable food source for newly   Width of a pupae (in mm) 0.29 ± 0.004 (n-60) emerged larvae with delicate, developing mouthparts. It is recorded that female gravid D. persicus are more attracted to soft fruit morph than hard fruit morph of C. gigantea due to the high penetrability of the ovipositor into soft morph fruits than hard morph (Sharma and Amriphale, 2007). Male D. persicus associates with developing fruits, during the oviposition period of females. It may be due to easy accessibility to females for the mating process and territory marking on suitable host fruits for facilitating females for oviposition (Aluja and Liedo, 2013). Pre-and post-oviposition behaviors of D. persicus are very similar to those observed in D. longistylus (Parihar, 1984).
Dissected fruits reveal that D. persicus eggs are in the inner seed chamber of the fruit or sometimes among seeds. The newly emerged fruit fly larvae contain weaker mandibles which are not strong enough to penetrate the inner pericarp layer of the fruit. To avoid the barrier, female D. persicus penetrate the inner pericarp layer by their long ovipositors and ensure the eggs are placed in the seed chamber. Although 2 -3 D. persicus females oviposit on a single fruit, dissected fruits contain only one egg cluster per fruit. It may be due to pseudo-oviposition of female fruit flies. Similarly, several females of Toxotrypana curvicauda Gerstaecker (a fruit fly), oviposit on single fruit but no multiple clusters of eggs in the fruit (Aluja and Norrbom, 1999). In addition, studies of Tephritid fruit flies reveal that they mark the oviposited fruit using host marking pheromones. It signals the other female fruitflies that the fruit is already used for an oviposition (Scolari et al., 2021). It may a possible reason of having only one cluster of eggs per Calotropis fruit. The eggs of D. persicus are longer (1.35 mm) than D. longistylus (1.00 mm) eggs (Parihar, 1984) . Similarly, the pupae of D. persicus (0.65 mm -length, 0.29 mm -width) are larger than D. longistylus (0.45 mm -length, 0.2 mm width) pupae (Parihar, 1984).
The third larval instar of D. persicus pupates in soil by forming burrows at 3 -5 cm deep (Sharma and Amriphale, 2008). Newly emerged D. persicus adults are lighter in color and unable to fly. However, within 15 minutes of emergence, they were able to fly.
D. persicus reared in the laboratory had a 1:1 sex ratio of female and male. In field observations, the female fruit fly abundance was higher than males. Normally female fruit flies have to choose a proper mate, mate with the selected partner, bear their eggs until maturation, find a suitable fruit to oviposit and lay their eggs. For a male fruit fly, their reproduction role is limited only to find a mate by competing with other male fruit flies and mating. Contribution of female fruit fly to generate new progeny is comparatively higher than male flies. For the effectiveness of the process, female fruit flies should live longer life. The observation of higher abundance in female fruit flies than male may occur due to the long lifespan of female fruit flies.

CONCLUSIONS
The findings of the present study provide detailed information on distribution, intensity of damage to the host plant, mating, oviposition and life cycle stages of D. persicus in Sri Lanka where no known records are previously available. As D. persicus feed on all the seeds (100%) of infected Calotropis fruit, it greatly influences on reproductive output and ultimately dispersal of the plant. Therefore, the present study will provide essential information for further studies of D. persicus in Calotropis spp. to seek the possibility of using it as a biocontrol agent.