Laboratory breeding of two Phortica species (Diptera: Drosophilidae), vectors of the zoonotic eyeworm Thelazia callipaeda

Background Some species of drosophilid flies belonging to the genus Phortica feed on ocular secretions of mammals, acting as biological vectors of the zoonotic eyeworm Thelazia callipaeda. This study describes an effective breeding protocol of Phortica variegata and Phortica oldenbergi in insectary conditions. Methods Alive gravid flies of P. oldenbergi, P. variegata and Phortica semivirgo were field collected in wooded areas of Lazio region (Italy) and allowed to oviposit singularly to obtain isofamilies. Flies were maintained in ovipots (200 ml) with a plaster-covered bottom to maintain high humidity level inside. Adult feeding was guaranteed by fresh apples and a liquid dietary supplement containing sodium chloride and mucin proteins, while larval development was obtained by Drosophila-like agar feeding medium. The breeding performances of two media were compared: a standard one based on cornmeal flour and an enriched medium based on chestnut flour. All conditions were kept in a climatic chamber with a photoperiod of 14:10 h light:dark, 26 ± 2 °C and 80 ± 10% RH. Results From a total of 130 field-collected Phortica spp., three generations (i.e. F1 = 783, F2 = 109, F3 = 6) were obtained. Phortica oldenbergi was the species with highest breeding performance, being the only species reaching F3. Chestnut-based feeding medium allowed higher adult production and survival probability in both P. oldenbergi and P. variegata. Adult production/female was promising in both species (P. oldenbergi: 13.5 F1/f; P. variegata: 4.5 F1/f). Conclusions This standardized breeding protocol, based on controlled climatic parameters and fly densities, together with the introduction of an enriched chestnut-based feeding medium, allowed to investigate aspects of life history traits of Phortica spp. involved in the transmission of T. callipaeda. Obtaining F3 generation of these species for the first time paved the road for the establishment of stable colonies, an essential requirement for future studies on these vectors in controlled conditions. Graphical abstract


Background
The Phortica genus (Diptera: Drosophilidae) includes around 130 species distributed worldwide, especially in mountain areas of the Oriental Region, where many species occur [1]. Most of the species of the subgenus Phortica (which includes P. variegata, P. foliiseta P. magna and P. omega complexes, and P. varipes group) have adapted
It is of paramount importance to breed colonies of arthropods in controlled conditions. From the early 1900s to now, breeding of several invertebrate taxa (e.g. moths, mosquitoes, beetles, marine copepods and fruit flies) has been pivotal to study subjects in many fields of science, including evolution, ecology and physiology [14]. In medical and veterinary contexts, the establishment of arthropod colonies is pivotal to investigating their life cycles, genetics, behaviour, interaction with vectorborne parasites and vector competence or susceptibility to insecticides [15]. Several rearing protocols have been developed for studying vectors of medical and veterinary concern such as mosquitoes, sand flies, tsetse flies [16][17][18] and P. variegata, for which a single protocol has been described in literature [19].
This study aims to describe a novel, standardized rearing method for P. variegata and P. oldenbergi flies, based on an artificial diet and characterized parameters of density and climatic conditions suitable to create a potentially stable colony. Results will represent an important starting point for controlled studies on Phortica spp. life cycle and vector role, toward in vitro testing of new insecticidal drugs and reducing the vector capacity of these drosophilids.

Sample collection
Phortica spp. gravid females were collected in Manziana (Lazio region, Italy, 42°07′09″N, 12°06′58″E; altitude 378 m a.s.l.) from May to September 2020. Wild females were collected with an entomological net around a bait of decaying fruits (i.e. apples, bananas, peaches, pears), which was placed into a cloth tied with a string around the bark of Turkey oak trees (Quercus cerris). Specimens were subsequently identified as P. oldenbergi, P. semivirgo and P. variegata using identification keys [18]. Gravid females were identified based on the enlarged abdomen and a yellowish colour on the abdominal sternites. Later on, the flies were individually transferred into plastic pots and immediately transported alive to the laboratory at the Department of Public Health and Infectious Diseases, Sapienza University of Rome (Italy). During the travel, pieces of fresh apple were added to the pots as food source and the containers were placed in a plastic box with a moist cloth pad to maintain high humidity and to protect flies from sunlight and excessive heat.

Rearing conditions
A maximum of ten female specimens was reared in 200ml (8 cm height, 5.5 cm diameter) cylindric transparent plastic containers (oviposition pots, hereafter called "ovipots") with a hole at the bottom filled with a 2-cm layer of plaster to maintain high humidity without water condensation. The container was closed by a lid with a net to prevent larvae from escaping and to allow adult feeding (and potentially ovipositing) on a slice of apple placed on the top. During ovipot servicing, flies were temporarily put in an empty cage of 30 cm 3 . Checks were performed every 2 days, when the piece of apple was changed and eggs/larvae potentially present on it were gently transferred using a dissection needle into a 10-ml container filled with a solid medium for hatching and larval development.
Additionally, adult feeding was enriched by a liquid dietary supplement (i.e. 77.7% distilled water, 19.4% snail extract-based syrup-Siromuicil, Herbit Italia Srl-and 2.9% sodium chloride) soaking a cotton wool inside the pot without using chemicals against moulds.
Ovipot moisture was regularly provided, wetting the plaster as needed, contemporarily avoiding excess of water and the consequent development of moulds. Sibling puparia were transferred and pooled in another plastic container in dry conditions during the entire pupation period in a plastic box and discarded if no adults emerged after 30 days. Adult progeny of a single wild female was kept in pool at the same parental conditions (maximum 10 flies/ovipot) maintaining both sexes in the same container to allow mating. All ovipots were stored in large plastic boxes (50 × 80 × 40 cm) placed  15:200 in a climatic chamber with a photoperiod of 14:10 h light:dark, 26 ± 2 °C and 80 ± 10% RH.

Statistical analysis
As only P. oldenbergi and P. variegata are known to be vectors of Thelazia callipaeda, focus was placed on these two species, as no production of progeny was obtained for P. semivirgo.
A negative binomial generalized linear model was performed to test the differences in oviposition rates between field-collected females of P. oldenbergi and P. variegata, as follows: where y i is the oviposition rate for the ith pot and β is the effect of the jth species, with j representing a factor with two levels (P. oldenbergi and P. variegata) [22].
Additionally, the same model structure using a linear model analysis was carried out to test puparia production of P. oldenbergi and P. variegata according to artificial diets. In this case, y i is the puparia rate (n. puparia/ females/pot) for the i th pot and β is the effect of the j th artificial diet, with j representing a factor with two levels (chestnut and standard media).
Kaplan-Meier analysis was carried out to determine the survival probability of Phortica puparia and adults. The survival probability at time t i , S (t i ) is calculated as follows: where S is survival, t i is time, d i is the number of events, and n i is the number of flies alive just before t i [23].
To test the robustness of the analysis, a log-rank test was performed approximately distributed as a chi-square function.

Phortica spp. development and reproduction parameters
Field-collected Phortica spp. females (n = 130; P. oldenbergi = 71; P. semivirgo = 3; P. variegata = 56) left to singularly oviposit in the plastic containers led to the production of three generations (F1 = 783; F2 = 109; F3 = 6) ( Table 1). The oviposition rates for field-collected females (based on mean number of eggs oviposited per female) were significantly higher for P. oldenbergi compared to P. variegata (negative binomial GLM; Z = − 2.637, P = 0.008; Table 1). Among F2 specimens, only those belonging to P. oldenbergi were able to produce F3 eggs. When comparing the mean number of puparia and adults per female in F1 and F2 generations, the highest value was reached by P. oldenbergi followed by P. variegata Table 1). The mean number of development days from puparia to adults (pupation) varied from a minimum of 12 ± 1 (range 4-49 ± 1) in P. oldenbergi F1 to a maximum of 20 ± 1 (range 11-20 ± 1) in F3 of the same species. Sex ratio of F1 was slightly unbalanced in favour of females along the sampling season for both P. variegata and P. oldenbergi with an average value of female proportions of 56% ( Table 2).

Influence of artificial diets on Phortica spp. breeding
The chestnut rearing medium led to a higher production of P. oldenbergi F1 adults compared to standard medium (4.6:1; negative binomial GLM: Z = − 2.940, P = 0.003). Conversely, no significant difference for P. variegata F1 adults was obtained between the two rearing media (2.5:1; negative binomial GLM: Z = − 1.19, P = 0.23; Table 1). Comparing the two feeding media, the average development time of P. oldenbergi progeny was lower for larvae fed with the standard medium (F1: 12 ± 1 days; F2: 15 ± 1 days) compared with chestnut medium (F1: 14 ± 1 days; F2: 23 ± 1 days). Conversely, F1 development of P. variegata showed no difference between the two media (13 ± 1 days with both feeding conditions; Table 3). The survival probability of chestnut-reared puparia of P. oldenbergi was higher compared to the standard-reared ones (log-rank test: χ 2 = 39.1; P < 0.0001; Fig. 1a). These results indicate that chestnut medium favoured pupal development of this species (median survival time for chestnut and standard: 22 and 18 days, respectively). Accordingly, also survival curves for P. variegata puparia showed significant difference in the use of the two media (log-rank test: χ 2 = 7.4, P = 0.007; median survival time of 21 days in chestnut-reared puparia and 16 days in standard-reared puparia, Fig. 1b). Likewise, the survival probability curves of P. oldenbergi adults showed a significant difference between the two media, with the median survival time of the chestnut-reared flies that is almost doubled compared to the standard-reared ones (34 and 19 days, respectively; log-rank test: χ 2 = 14.9; P = 0.0001; Fig. 2a). Similarly, survival probability is also higher for P. variegata chestnut-reared adults (chestnut: 28 days, standard: 15 days; log-rank test: χ 2 = 9.3; P = 0.002; Fig. 2b).

Discussion
The breeding protocol presented here indicates the importance of optimizing parameters such as adult density and diet in successfully breeding Phortica species. In fact, a F3 has been obtained for the first time, with a substantial improvement compared to previous attempts [19]. To date, a single laboratory breeding protocol for P. variegata has been described [19], in which the authors successfully bred this species up to the second generation adopting a simple approach based on feeding ad libitum with fresh apple at all development stages of the flies in a 30-cm 3 cage with high relative humidity. Compared to previous protocols [19], in which 0.5 F1 adults/females were obtained, this study shows that limiting the fly numbers per pot, the modulation of relative humidity during the life cycle, with drier conditions for pupal phase, and optimal feeding based on chestnut flour and liquid dietary supplement provide higher performance in terms of adult progeny for both F1 and F2 (P. variegata: 4.5 F1/f, 1.8 F2/f; P. oldenbergi: 13.5 F1/f, 1.4 F2/f; Table 1). The unequal number of females tested with chestnut medium compared with the standard one was a consequence of its late introduction in the rearing protocol. It was impossible to reach equal numbers of flies for both media as the season ended and field flies were no longer available.
Moreover, this rearing protocol allowed breeding P. oldenbergi for the first time to our knowledge, providing first data on its life history traits. Also, this species has been demonstrated to be more adapted to insectary conditions compared to P. variegata, encouraging its employment as a potential model for challenge studies and trials of veterinary products.
The assumption that P. variegata is closely associated with oak forests [24], where acorns or other nuts can be one of the potential feeding sources for larvae (J. Jaenike, personal communication), might not be acceptable for this species over its whole areal, as oak species are rare to absent in central and northern Europe. However, several nut tree species are widely distributed in the southern part of the European continent, particularly in Italy, where P. variegata is diffused and P. oldenbergi was detected (Bernardini et al., unpublished data). Therefore, a chestnut-based medium could approximate natural feeding conditions, with better performance than a standard Drosophila medium based on corn flour, but also compared to fresh fruit. In addition, the liquid dietary supplement containing sodium chloride and mucin proteins (snail extract) might have partially compensated the deficiency of salts and proteins consequent to the lack of lachryphagy under laboratory breeding conditions. Lachryphagous behaviour is described in several Steganinae species other than those belonging to Phortica genus, in particular Amiota, Gitona, Paraleucophenga and Apenthecia [1]. However, this is not a peculiar behaviour of drosophilid flies, as other Diptera (e.g. Muscidae, Fanniidae, Chloropidae and Paraleucopidae) show some degree of lachryphagy in both sexes, with a prevalence of females as opposed to males of Steganinae [1]. Also several Lepidoptera show a lachryphagous behaviour in males and occasionally also in females [1]. Finally, this behaviour was observed in cockroaches exploiting lacrimal secretions from reptiles [25]. This supplementary feeding seems to be useful to obtain essential compounds, such as minerals, urea, glucose and proteins, possibly useful to increase fitness of lachryphagous species [25].   Despite these encouraging results, the low initial number of wild females used in this study, especially the low number reared with the chestnut medium, might have affected the possibility to obtain a stable colony. This may be a consequence of the low initial genetic variability of the laboratory population [26] as well as an intrinsic low oviposition rate of the field-collected Phortica spp. (P. oldenbergi: 33.9 eggs/female, P. variegata: 22.6 eggs/ female). Comparing data herein obtained about oviposition rates with those of species belonging to the Drosophila genus (> 2500 eggs/female; [27]), the low progeny numbers per generation may account for the biological limitation of Phortica spp. in obtaining a stable colony. New attempts will be conducted with a higher starting number of field-collected females to overcome these limitations and try to obtain a stable colony.
Data also allowed clarifying an open question about the population dynamics of P. variegata in the field. In fact, it is known that this species shows a switch of sex ratio along the breeding season, with an increase in lachryphagous males during late summer [28]. The F1 obtained by field-collected females from May to October did not indicate any shift in progeny sex ratio along the season ( Table 2). This led to the conclusion that the switch of relative proportion of males during the season is a consequence of a sampling bias due to their feeding behaviour instead of a physiological change of sex ratio in the population.

Conclusions
This novel breeding protocol of Phortica spp. allowed to investigate aspects of life history traits of these drosophilids, which are involved in the transmission of the zoonotic eyeworm T. callipaeda. Controlled climatic parameters and fly densities, together with the introduction of a more proper feeding medium (i.e. considering the needs of Phortica spp. associated to oak forests) significantly improved the survival and fecundity of both P. variegata and P. oldenbergi. This standardized approach allowed to reach F3 generation for the first time, representing the basis for the establishment of stable colonies, which are an essential requirement for future behavioural/physiological studies on these vectors as well as pharmaceutical trials of veterinary and medical products.