Can stable isotope markers be used to distinguish wild and mass-reared Anastrepha fraterculus flies?

The availability of accurate techniques to discriminate between marked laboratory-reared flies and unmarked wild flies captured in monitoring traps is essential for programs that integrate the Sterile Insect Technique (SIT) to manage fruit flies. In this study, the feasibility of using a stable isotope marking technique for the South American fruit fly, Anastrepha fraterculus (Wiedemann), was assessed. Wild flies were collected from apple orchards, which are a target of a SIT project in southern Brazil. To verify if adult flies could be labelled by the stable isotopes from larval diets, larvae were reared on two different C4-based diets and fruits in laboratory. To evaluate the influence of the two most common attractants applied to capture A. fraterculus (grape juice and CeraTrapTM) and the most common preservation method in fruit fly collections (ethanol), laboratory-reared flies were immersed in McPhail traps containing the respective treatments for two periods of time. Samples were analyzed in an elemental analyzer coupled to a Continuous Flow Isotope Ratio Mass Spectrometer (CF-IRMS) at CENA/USP. The δ13C signatures of flies reared on artificial diets differed significantly from the δ13C of flies whose larvae were reared on fruits and from wild flies. In contrast, the δ15N values were less conclusive and the technique could not rely solely on them. In all cases considered, the δ13C and δ15N signatures from males did not differ from females. Despite the alterations caused by the attractants tested and ethanol, laboratory-flies could be distinguished from the wild ones based on δ13C signatures. This is the first comprehensive study to demonstrate that it is possible to distinguish wild A. fraterculus from flies reared on larval diets containing C4 sugar. The first experimentally derived trophic discrimination factors were also obtained for this species. Thus, intrinsic isotope labelling can serve as a backup to conventional dye marking.

Introduction

The Sterile Insect Technique (SIT), an autocidal method of control that relies on area-wide inundative releases of sterile insects, has been successfully applied by many operational programs to eradicate, suppress, and contain fruit fly pests [1]. An accurate discrimination between wild and released marked flies captured in monitoring traps is necessary to identify remaining pockets of wild populations, determine actual overflooding ratios, and evaluate progress of the SIT program [2, 3]. In SIT containment programs, it is crucial to avoid misidentification of the released insects, since costly quarantine restrictions could be triggered [2].

The fluorescent dye marking method is still the most used in SIT programs against fruit flies [4]. This method consists of impregnating sterilized pupae with fluorescent dye powder of plant origin (e.g. Day-Glo®). During the emergence from the puparium, the cephalic ptilinum sac of the fly touches the dye, taking some of it when the sac retrieves back inside the head. As fluorescent markings from wings, legs and other body parts of the sterile flies can be lost by weathering, grooming or other reasons, dye marks preferably in the ptilinial suture are sought under a UV lamp or epifluorescence microscope. Despite the high efficiency of the marking process, a very small proportion of the flies can be poorly marked. In the Mexico/United States Mediterranean fruit fly eradication program, for example, the percentage of unmarked captured flies varied between 0.05 and 1% [5]. For flies whose heads do not present internal markings, their reproductive organs must undergo a detailed cytohistological study [6].

Elemental markers can be an alternative to dye marking [7]. Naturally occurring stable isotopes have become a robust identification strategy and its application in SIT programs has been stimulated in recent years [8]. The ratios of carbon and nitrogen stable isotopes (i.e. 13C/12C and 15N/14N) have been extensively used by entomologists [9, 10, 11, 12]. Photosynthesis alter the ratio 13C/C in relation to atmospheric CO2 depending on the photosynthetic pathways, resulting in different isotopic composition of plants with C3 or C4 metabolism [13, 14]. On the other hand, the 15N/14N ratio is altered along food chain, with 15N becoming concentrated as the trophic level increases [10, 12].

As the isotopic signature of an animal reflects the isotopic composition of its food [15, 16], the carbon and nitrogen isotopic signature can be used to track feeding patterns of insects, for instance. Many studies have demonstrated that stable isotopes can be suitable intrinsic markers for mosquitoes [17, 18, 19], tsetse flies [20], moths [21, 22] and fruit flies [23].

The great majority of fruit fly larvae develop on C3 plants in the wild [24], but most of the commercially available sugars used in mass-rearing facilities have a C4 source, mainly sugarcane (Saccharum spp., Poales: Poaceae) [25]. Hood-Nowotny et al. [23] demonstrated that mass-reared Mediterranean fruit flies (medfly), Ceratitis capitata (Wiedemann), can be effectively labelled and distinguished from wild populations when C4 sugar is used in the larval-rearing diet.

The feasibility of using the stable isotope marking technique for the South American fruit fly, Anastrepha fraterculus (Wiedemann), has never been assessed. This is the most economically important fruit fly in commercial apple and peach orchards in southern Brazil [26]. Since 2015, A. fraterculus populations from this region have been the target of a pest management program called MOSCASUL, which integrates the use of sterile insects and parasitoids [27]. The availability of an alternative accurate labelling technology to identify trapped flies would be extremely valuable for the program, serving as a backup to conventional dye marking.

Aiming to determine whether stable isotopes of carbon and nitrogen would serve as reliable markers for A. fraterculus, based on natural signature differences between wild and flies reared on artificial diets, the isotopic compositions of larval diets (artificial and natural), adult flies reared under semi-mass rearing conditions, and adults captured from wild populations were assessed. As wild females are usually captured in higher numbers in traps placed in temperate fruit orchards [28], possible differences in carbon and nitrogen isotopic signatures of males and females from both wild and laboratory-reared flies were assessed. The influence of attractive and preservative substances commonly used on traps or for sample preservation over the isotopic signature of the flies was also evaluated.

Materials and methods

Species identification was performed following Hernández-Ortiz et al. [29]. No endangered or protected species were involved in the study and no specific permits were required for the described field studies.

Collection of wild Anastrepha fraterculus flies

The capture of wild A. fraterculus flies was carried out in January-February 2017 with McPhail traps containing CeratrapTM (Bioibérica, Barcelona, Spain). All collections were performed in apple orchards located in the municipality of Vacaria, State of Rio Grande do Sul, Brazil. The flies were captured in three different locations: (a) apple orchard on private property (28°30’59.26” S; 50°52’16.80” W); (b) experimental field area of Embrapa Uva & Vinho (28°30’58.98 S; 50°53’0.97” W); and (c) commercial area of Schio Ltda. (28°32’9.30” S; 50°49’18.30” W).

The traps were inspected at 12 h intervals. For the stable isotope analysis (SIA), 20 flies (10 males and 10 females) from each sampling site were used. All manipulations of flies in the field were carried out using gloves to avoid contaminations. After collection, the flies were gently washed with distilled water and dried for 6 h on paper tissue. The individual samples were conditioned in Eppendorf tubes and stored until the analyses.

Labelling Anastrepha fraterculus flies with different larval diets

To verify if adult flies could be labelled by the stable isotopes from larval diets, an experiment with different artificial diets and fruit hosts was conducted. A bisexual strain of A. fraterculus, originally established with wild populations from Vacaria [30], has been maintained under semi-mass rearing conditions at the Center for Nuclear Energy in Agriculture of the University of São Paulo (CENA/USP) [31], and it served to provide eggs or adult flies for the test.

In order to obtain the laboratory-reared flies, eggs from the mother colony with the aforementioned bisexual strain were collected and bubbled in water bath at 25°C for 72 h. Aliquots of 2 mL of eggs were seeded in trays containing 1 L of artificial diet, with the larvae remaining there for 10–12 days at 24°C. The larvae were reared on two distinct artificial diets (C4-based diets): (a) Diet I (formulated with 50 g of brewer's yeast BrewcellTM (Biorigin, Brazil), 30 g of sugar (sugarcane source) CaravelasTM (Usina Colombo, Brazil), 300 g of corn bran YokiTM (General Mill Alimentos, Brazil), 2 mL of Nipagin, 2 g of sodium benzoate, 6 g of citric acid, and 1,000 mL of distilled water); and (b) Diet II (a gelled diet adapted from Salles [32], formulated with 3 g of agar, 60 g of corn bran YokiTM (General Mill Alimentos, Brazil), 60 g of sugar CaravelasTM (Usina Colombo SA, Brazil), 60 g of brewer's yeast BrewcellTM (Biorigin, Brazil), 1 g of sodium benzoate, 6 mL of hydrochloric acid, 8 mL of Nipagin and 900 mL of distilled water). After the larval period, third instar larvae were separated from the artificial diets by washing and then transferred to 500 mL plastic cups with moist vermiculite for pupation in a dark room.

To obtain adult flies from a natural and an alternative host (C3-based diets), larvae were reared on apples (Malus domestica Borkh. cv. ‘Gala’) [33] and papaya fruits (Carica papaya L. cv 'Golden') [34], respectively. The fruits were purchased at the market and cleaned four times with detergent and water. Newly-emerged adults from the mother-colony of the A. fraterculus bisexual strain were kept in screened cages (30 x 30 x 30 cm; ≈ 600 couples/cage) with water and a mixture of sugar and hydrolyzed brewer's yeast Bionis YE MFTM (Biorigin, Lençois Paulista, Brazil) at 3:1 rate and water ad libitum [35] under laboratory conditions (25±1°C and 65±10% RH) for the infestation of the fruits. When the flies were 7 days old, the fruits were inserted in each cage separately. After 24 h, the fruits were removed and conditioned in plastic boxes (40 x 20 x 15 cm). Ten days later, the larvae started crawling from the fruits. They were collected twice a day and left to pupate in 500 mL plastic cups with vermiculite.

Treatments were replicated five times with one larval tray or fruit making up each replication. One day before emergence, samples with up to 100 mL of pupae from each replicate were placed in screened cages (30 x 30 x 30 cm) for adult emergence, with only water provided by vials with cotton wicks. After emergence, the 24-h-old flies were frozen to death and the individuals were conditioned in Eppendorf microtubes. Two couples were collected from each of the five replicates of the four treatments. Half of the couples were used for the sex comparison test, while the other half was used to compare the isotopic signals of flies from the different artificial diets and fruits.

Evaluation of the influence of attractive and preservative substances on the isotopic composition of A. fraterculus flies

This experiment was carried out to evaluate the influence of the two most commonly attractants applied to capture fruit flies [36] and the most common preservation method in fruit fly taxonomic collections [37] on 13C and 15N signatures of A. fraterculus flies. The three substances were: CeraTrapTM (Bioibérica, Barcelona, Spain), concentrated grape juice (Cooperativa Vinícola Garibaldi, Garibaldi, RS, Brazil) and absolute ethanol (Química Moderna Indústria & Comércio, Barueri, SP, Brazil). They were placed separately in McPhail traps (200 mL of each substance per trap), which were distributed randomly in an acclimatized room (25 ± 1°C and 65 ± 10% RH). For the control group, the McPhail traps were filled with distilled water only.

To verify the combined effect of an attractant (CeraTrapTM) and ethanol on the isotopic composition of the flies, a group of flies from Diets I and II that had been immersed in CeraTrapTM for seven days was washed and then immersed in absolute ethanol for one week. The treatments and their respective acronyms are listed in S1 Table.

About 100 flies (24-h-old) whose larvae were reared on Diet I and Diet II were separated to be anesthetized by cold (10 min. at -20°C) and then immersed in McPhail traps containing each of the treatments. All the four treatments and control were replicated by three separate traps. Two flies (1 male and 1 female) were randomly collected from each trap on the first and seventh day after immersion, washed with distilled water and dried for 6 h on paper tissue. After that, the samples (i.e. 6 flies/treatment/day) were individualized as previously described.

Stable isotope analyses

Analyses of 13C/12C and 15N/14N ratios were performed to determine the isotopic compositions of wild flies collected in Vacaria, the artificial diets and fruits, newly-emerged laboratory-reared flies, attractive and preservative substances, and fly samples immersed in different substances.

The δ was calculated by δX = [(Rsample/Rstandard) −1], where δX refers to δ13C or δ15N, and R is molar ratio of rare to abundant isotopes (13C/12C or 15N/14N) of the sample (Rsample) and standard (Rstandard). Isotope ratios are expressed in delta units (parts per thousand; δ ‰) relative to the international reference standards (Rstandard) which are Vienna PeeDee Belemnite (VPDB) and atmospheric nitrogen (N2) for C and N, respectively.

The individualized samples were placed to dry in a ventilated laboratory stove at 50°C for 72 h, and then macerated (whole body) until reaching constant masses. All dried materials were weighed (0.3–0.5 mg for artificial diets, 0.8–1.0 mg for fly samples, and 2.5–2.8 mg for fruit samples) in a precision analytical balance ME 36S (Sartorius, Göttingen, Germany) and enclosed in tin capsules (Elemental Analysis 5 x 3.5 mm). For the ethanol and attractive substances, about 1 mL of the stock substance was used for the analyses.

The enclosed samples were analyzed in an elemental analyzer (CHN 1110 –CE Instruments, Rodano, Italy), coupled to a Continuous Flow Isotope Ratio Mass Spectrometer (CF-IRMS) (Delta Plus–Thermo Scientific, Bremen, Germany). Samples were fired in an oxygen atmosphere at approximately 1700°C, and the resulting N2 and CO2 passed through a series of scrubbers to remove the impurities and waste water in the elemental analyzer through a chromatographic separation column in ultrapure helium carrier. The CO2 and the N2 peaks were evaluated in the CF-IRMS to determine the isotopic ratios.

The results were normalized to the international standards using secondary reference materials (NBS-19, NBS-22, IAEA-N1, IAEA-N2) [38]. An internal standard (i.e. sugarcane leaf) was used for quality control every 11 samples in each run. Each sample was analyzed twice and the analytical error was estimated at ±0.15‰ for both δ13C and δ15N measurements.

Trophic discrimination factors (TDFs) of carbon (Δ13C) and nitrogen (Δ15N) were calculated by subtracting the mean δ13C or δ15N values of the whole-body of the 24-h-old flies from the mean δ13C or δ15N values of the diet: ΔX = δXfly’ whole-body− δXdiet, where Δ is the trophic discrimination factor and X represents the chemical element being used (13C or 15N) [39].

Statistical analyses

The mean values of isotopic compositions (δ13C and δ15N) from males and females of the three apple orchards, from each larval diet and fruit were compared by the Student’s t-test (α = 0.05). For the δ13C and δ15N values from the wild flies of the three collection sites, the one-way analysis of variance F-test was applied at the 5% of significance (ANOVA) and, when significant differences were detected, the Tukey’s honestly significance difference (HSD) test (α = 0.05) was applied to compare the means. The mean values of δ13C and δ15N from the larval diets, fruits, adults obtained from the diets and wild flies were compared by the Tukey’s test (α = 0.05). For the isotopic compositions (δ13C and δ15N) of wild flies and from the different treatments and times (immersion for 1 or 7 days), a two-way analysis of variance F-test was applied at a 5% level of significance, considering time and treatment (factorial scheme 2 by 11) as factors. Homogeneity of variances and normality of model residuals were checked in all instances [40, 41]. The analyses were performed by the statistical program SAS 9.4 [42]. Raw isotope data included in S1 Dataset.

Results

Isotopic compositions of wild A. fraterculus flies

A total of 59 wild flies collected in three different apple orchards from Vacaria were analyzed for δ13C and δ15N. The isotopic compositions of wild males did not differ from the compositions of females within each of the apple orchards (sites A, B, and C) (P > 0.05) (Table 1).

Table 1

Isotopic composition of males and females (means ± SE) of wild Anastrepha fraterculus flies from three apple orchards.

Orchard	Sex	δ13C ‰	ANOVA	δ15N ‰	ANOVA
A	Male (n = 10)	-25.5 ± 0.2†	F1,18 = 3.5; P = 0.079	5.9 ± 0.6	F1,18 = 0.81; P = 0.38
	Female (n = 9)	-25.1 ±0.1		5.4 ± 0.3
B	Male (n = 10)	-25.8 ± 0.2	F1,19 = 0.47; P = 0.503	4.3 ± 0.6	F1,19 = 0.04; P = 0.84
	Female (n = 10)	-25.6 ± 0.3		4.24 ± 0.7
C	Male (n = 10)	-26 ± 0.2	F1,19 = 1.0; P = 0.33	5.8 ± 0.7	F1,19 = 1.72; P = 0.21
	Female (n = 10)	-26.2 ± 0.2		6.6 ± 0.6

† Means (± SE) within columns did not differ significantly at the 5% probability level by the Student’s t-test.

As the isotopic compositions of the sexes within each collection site did not differ significantly, the δ13C and δ15N values from the wild flies of each site were compared among them (Table 2). The δ13C values of flies from site A differed significantly from the flies from site C, and the δ15N values of the flies from site B differed from both sites A and C (Table 2). Although there were such differences, the overall δ13C mean values of all flies (n = 59) was -25.7 ± 0.2‰, clearly indicating a predominance of C3 carbon from apples or natural hosts in the flies. The overall δ15N of flies was 5.4 ± 0.6‰.

Table 2

Isotopic composition of wild Anastrepha fraterculus flies from three different apple orchards.

Orchard	δ13C ‰	ANOVA	δ15N ‰	ANOVA
A (n = 19)	-25.3 ± 0.2 a†	F2,58 = 9.45; P< 10−3	5.6 ± 0.3 a	F2,58 = 10.28;P< 10−3
B (n = 20)	-25.7 ± 0.1 ab		4.3 ± 0.3 b
C (n = 20)	-26.1 ± 0.1 b		6.2 ± 0.3 a

† Means (± SE) within columns followed by the same letter do not differ significantly at the 5% probability level by the Tukey’s test.

Influence of larval diets on the isotopic compositions of A.fraterculus flies

The isotopic compositions of males and females of flies whose larvae were reared on larval diets and fruits did not differ significantly (P> 0.05) (Table 3).

Table 3

Mean isotopic composition of males and females of Anastrepha fraterculus flies reared on different larval diets and fruits.

Treatment	Sex	δ13C ‰	ANOVA	δ15N ‰	ANOVA
Flies reared on Diet I	Male (n = 4)	-15.4 ± 0.1†	F1,7 = 1.63; P = 0.25	3.95 ± 0.2	F1,7 = 0.16; P = 0.71
	Female (n = 4)	-15.2 ± 0.1		4.01 ± 0.06
Flies reared on Diet II	Male (n = 5)	-16 ± 0.2	F1,9 = 0.07;P = 0.80	4.3 ± 0.1	F1,9 = 0.58; P = 0.467
	Female (n = 5)	-16.1 ± 0.1		4.0 ± 0.3
Flies reared on papaya	Male (n = 5)	-27.1 ± 0.2	F1,9 = 0.28; P = 0.615	6.9 ± 0.4	F1,9 = 0.04; P = 0.838
	Female (n = 5)	-26.9 ± 0.4		6.8 ± 0.4
Flies reared on apple	Male (n = 4)	-25.8 ± 0.03	F1,8 = 0.01; P = 0.93	2.96 ± 0.1	F1,8 = 1.48; P = 0.27
	Female (n = 5)	-25.9 ± 0.1		2.9 ± 0.4

† Means (± SE) within columns did not differ significantly at the 5% probability level by the Student’s t-test.

The δ13C and δ15N of larval diets, fruits, laboratory and wild flies are presented in Table 4 and the distribution of their mean values is shown in Fig 1.

Table 4

Isotopic composition of larval diets, fruits, laboratory and wild Anastrepha fraterculus flies.

SAMPLE	δ13C ‰	ANOVA	δ15N ‰	ANOVA
Artificial Diet I (n = 5)	-12.6 ± 0.1 a†	F3,19 = 14269.6; P< 10−3	2.9 ± 0.1 b	F3,19 = 96.05; P< 10−3
Artificial Diet II (n = 5)	-14.3 ± 0.1 b		2.4 ± 0.1 bc
Papaya (n = 5)	-26.0 ± 0.1 c		5.5 ± 0.2 a
Apple (n = 5)	-28.4 ± 0.1 d		2.0 ± 0.2 c
Flies reared on Diet I (n = 10)	-15.3 ± 0.1 a	F4,42 = 2063.6; P< 10−3	4.0 ± 0.1 c	F4,42 = 47.10; P< 10−3
Flies reared on Diet II (n = 10)	-16.0 ± 0.1 b		4.1 ± 0.2 c
Flies reared on papaya (n = 10)	-27.0 ± 0.2 d		6.8 ± 0.3 a
Flies reared on apple (n = 10)	-25.9 ± 0.1 c		2.9 ± 0.2 d
Wild flies (n = 3) ‡	-25.7 ± 0.2 c		5.4 ± 0.6 b

† Means (± SE) within columns followed by the same letter do not differ significantly at the 5% probability level by the Tukey’s test.

‡ The mean values of wild flies from the three different apple orchards were considered.

Fig 1

Distribution of the mean values of δ13C and δ15N of larval diets, fruits, and Anastrepha fraterculus flies reared in laboratory and wild flies (original means±standard errors in Table 4).

All diets differed from each other in δ13C. Considering the δ15N values, papaya presented the highest δ15N value (5.5‰), differing significantly from the other diets.

The results also indicated that the δ13C values of the A. fraterculus adults reared in laboratory were significantly influenced by the medium in which they developed. The δ13C mean values of A. fraterculus reared on artificial Diets I and II reflected that of C4-carbon sources, differing from the δ13C of flies reared on fruits and wild flies. The flies from Diet I presented the highest δ13C mean value (-15.3‰), while flies from papaya presented the lowest value (-27‰). The δ15N of A. fraterculus reared in either fruits or artificial diets differed from the values of wild flies (P < 0.05) (Table 4).

The estimated TDFs showed both enrichment and depletion of δ13C in adult flies depending on the diet (S2 Table). In comparison to the respective diets, the flies reared on larval Diets I and II had a decrease in δ13C of 2.7 and 1.7‰, respectively. A decrease of 1‰ in δ13C value of the flies whose larvae were reared on papaya and an increase of 2.5‰ in δ13C of the flies whose larvae were reared on apple were also observed.

Considering the Δ 15N values, only δ15N enrichment was observed in all treatments, which is expected with the increase of trophic level [10, 43] (S2 Table).

Influence of capture and preservative substances on the isotopic composition of A. fraterculus flies

The mean δ13C and δ15N values of CeraTrapTM were -25 ± 0.1‰ and 3.4 ± 0.2‰, respectively. For grape juice, the means were -26.3 ± 0.1‰ and 2.8 ± 0.1‰, respectively; while the values from absolute ethanol were -12.1 ± 0.5‰ for δ13C and 1.7‰ for δ15N. For wild flies and those from control treatments (Water I and II; acronyms for the treatments are given in Tables 5 and 6), the isotopic composition values were kept without noticeable variation until the seventh day of evaluation.

Table 5

Mean δ13C values of wild and laboratory Anastrepha fraterculus flies that were immersed in attractive and preservative substances for one and 7 days.

TREATMENT	δ13C ‰ (1 day)	δ13C ‰ (7 days)	ANOVA
Wild flies (n = 3) †	-25.7 ± 0.2 A	-25.7 ± 0.2 A	-
Water I (n = 6)	-15.4 ± 0.1 DE	-15.4 ± 0.1 DE	-
Water II (n = 6)	-15.9 ± 0.1 CD	-15.9 ± 0.1 CDE	-
ET I (n = 6)	-15 ± 0.2 E a ‡ §	-15 ± 0.2 E a	F1,11 = 0.03; P = 0.856
ET II (n = 6)	-16.8 ± 0.1 B a	-17.1 ± 0.1 BCD b	F1,11 = 6.53; P = 0.028
CT I (n = 6)	-15.2 ± 0.2 DE a	-16.3 ± 0.3 CDE b	F1,11 = 11.14; P = 0.007
CT II (n = 6)	-15.6 ± 0.2 CDE a	-16.8 ± 0.2 BCD b	F1,11 = 13.82; P = 0.004
GJ I (n = 6)	-16.4 ± 0.1 BC a	-17.3 ± 0.8 BC a	F1,11 = 1.39; P = 0.266
GJ II (n = 6)	-16.8 ± 0.1 B a	-18.1 ± 0.9 B a	F1,11 = 2.47; P = 0.147
CTET I (n = 6)	-16.3 ± 0.3 BC a	-16.3 ± 0.1 CDE a	F1,11 = 0,01; P = 0.963
CTET II (n = 6)	-16.8 ± 0.2 B a	-17.6 ± 0.2 BC b	F1,11 = 6.2; P = 0.03
ANOVA	F10,62 = 399.13; P<10−3	F10,62 = 80.03; P< 10−3

† The mean values of wild flies from the three different apple orchards were considered all together. Acronyms: ET = ethanol; CT = CeraTrapTM; GJ = grape juice; CTET = CeraTrapTM for 7 days and then flies immersed in absolute ethanol for 7 days; I = performed with flies from Diet I; II = performed with flies from Diet II.

‡ Means (± SE) followed by the same uppercase letter in the columns do not differ significantly at the 5% probability level by the Tukey’s test.

§ Means (± SE) followed by the same lowercase letter in the lines do not differ significantly at the 5% probability level by the Student’s t-test.

Table 6

Mean δ15N values of wild and laboratory Anastrepha fraterculus flies that were immersed in attractive and preservative substances for one and 7 days.

TREATMENT	δ15N ‰ (1 day)	δ15N ‰ (7 days)	ANOVA
Wild flies (n = 3) †	5.4 ± 0.6 AB	5.4 ± 0.6 AB	-
Water I (n = 6)	4 ± 0.1 C	4 ± 0.1 C	-
Water II (n = 6)	4.1 ± 0.3 C	4.1 ± 0.3 C	-
ET I (n = 6)	4.4 ± 0.1 BC a ‡ §	4.5 ± 0.1 BC a	F1,11 = 0.28; P = 0.608
ET II (n = 6)	4.6 ± 0.1 BC a	4.9 ± 0.1 BC b	F1,11 = 15.25; P = 0.002
CT I (n = 6)	4.5 ± 0.2 BC a	6.3 ± 0.4 A b	F1,11 = 18.35; P = 0.001
CT II (n = 6)	4.4 ± 0.2 BC a	5.5 ± 0.2 AB b	F1,11 = 13.59; P = 0.004
GJ I (n = 6)	4.8 ± 0.3 BC a	4.2 ± 0.4 C a	F1,11 = 1.71; P = 0.22
GJ II (n = 6)	4.6 ± 0.1 BC a	4.8 ± 0.1 BC a	F1,11 = 0.99; P = 0.343
CTET I (n = 6)	6.3 ± 0.4 A a	5.4 ± 0.1 AB b	F1,11 = 5.86; P = 0.036
CTET II (n = 6)	5.5 ± 0.2 AB a	5 ± 0.2 BC a	F1,11 = 2.01; P = 0.187
ANOVA	F10,62 = 8.54; P< 10−3	F10,62 = 8.35; P< 10−3

† The mean values of wild flies from the three different apple orchards were considered all together. Acronyms: ET = ethanol; CT = CeraTrapTM; GJ = grape juice; CTET = CeraTrapTM for 7 days and then flies immersed in absolute ethanol for 7 days; I = performed with flies from Diet I; II = performed with flies from Diet II.

‡ Means (± SE) followed by the same uppercase letter in the columns do not differ significantly at the 5% probability level by the Tukey’s test.

§ Means (± SE) followed by the same lowercase letter in the lines do not differ significantly at the 5% probability level by the Student’s t-test.

The δ13C and δ15N values of the flies were significantly affected after immersion in different substances (F = 271.83, P< 10−3; F = 11.93, P < 10−3, respectively), but the exposure time (1 or 7 days) in traps was significant only for δ13C (F = 17.49, P< 10−3, F = 1.81, P = 0.18, respectively). The interaction of factors time and treatments was not significant on the δ13C values of the flies (F = 1.90, P = 0.051; F = 4.96, P< 10−3, respectively).

Grape juice caused the most of δ13C depletion (up to 0.6‰ and 2.1‰ after one and 7 days of immersion, respectively). Consequently, the flies immersed in grape juice for 7 days differed significantly from both water controls (P < 0.05) (Table 5). After 7 days of immersion, most of the treatments resulted in depletion of δ13C compared to the Water I control (CT I = -0.9‰; GJ I = -1.9‰), with enrichment only in ET I (+0.4‰). Compared to the Water II control, depletion was observed in all treatments (ET II = -1.2‰; CT II = -0.9‰; GJ II = -1.9‰).

In its turn, the flies immersed in CeraTrapTM for 7 days had their δ13C values not significantly altered compared to controls, but CeraTrapTM was the substance that most impacted the δ15N values (Table 6). Actually, all the substances tested enriched δ15N values after 7 days in comparison to Water I (ET I = +0.5‰; CT I = +2.3‰; GJ I = +0.1‰) and Water II controls (ET II = +0.8‰; CT II = +1.4‰; GJ II = +0.7‰). When CeraTrapTM and ethanol were combined, differences by up to -1.7 and +1.4‰ in δ13C and δ15N signatures were respectively observed (Table 6).

Considering factor time within each of the treatments, significant differences in either δ13C and δ15N values from the ET II, CT I and CT II treatments were observed (P<0.05) (Tables 5 and 6). On the seventh day of evaluation, the δ13C values of the flies from these three treatments suffered depletions of 0.3‰, 1.1‰ and 1.2‰, while δ15N was enriched in 0.3‰, 1.8‰ and 1.1‰, respectively.

Independent of treatment and immersion time, the differences in δ13C values between wild flies and those reared on Diets I and II were still significant (P < 0.05) (Table 5). Considering the δ15N signals, the flies from the different treatments (except the controls and GJ I after 7 days) did not differ significantly from the wild flies (Table 6).

Discussion

Significant differences in δ13C values were observed when larvae fed on diets with distinct isotope signatures (Table 4). Due to the amounts of cane sugar and corn used in Diets I and II, they can be considered C4-based diets. The δ13C signals of these diets ranged from -12.5 to -14.5‰, but some C3 ingredients such as brewer’s yeast might have helped to slightly decrease the signatures from both diets [44]. The δ13C signatures of A. fraterculus reared on artificial diets tended to reflect those of the larval diets (-15.3‰ and -16‰, respectively), differing from the mean δ13C values from flies whose larvae were reared on fruits (apple and papaya) and wild flies (Table 4).

In Brazil, 116 hosts of A. fraterculus are known [45, 46, 47, 48], with most of the species belonging to the families Myrtaceae and Rosaceae [46]. In the State of Rio Grande do Sul, besides apple orchards, sources of infestation are observed in commercial peach orchards (Prunus persica), citrus (Citrus spp.) and grapevine (Vitis vinifera) [45, 46, 48, 49, 50]. Cultivation of new fruits, such as blackberry (Rubus spp.), blueberry (Vaccinium ashei) and Brazilian cherry (Eugenia uniflora), may also contribute with new breeding sites for A. fraterculus [51]. In a recent study performed by Marsaro Junior [52], infestations by A. fraterculus were observed in fruits of Annona rugulosa, Acca sellowiana, Campomanesia guazumifolia, Campomanesia xanthocarpa, Diospyros kaki, Eugenia involucrata, Eugenia pyriformis, Eugenia uniflora, Psidium cattleianum, Psidium guajava and Prunus persica. In the municipality of Vacaria, large infestations of this pest can be observed in Campomanesia xanthocarpa, Eugenia involucrata and Feijoa sellowiana [53]. All these fruits follow the C3 photosynthetic pathway [13] and future determinations of their isotopic signatures could help to elucidate the differences observed among wild flies from different sites. Although some variability in their carbon isotopic composition will exist, their δ13C values will be consistently lower than flies reared on C4-based diets (Fig 1). This fact assures the usefulness of stable isotopes to label mass-reared A. fraterculus.

This usefulness is reinforced by the fact that the δ13C values of males did not differ from females, neither in flies reared on different larval diets nor in wild flies (Tables 1 and 3). This finding is in line with Hood-Nowotny et al. [22] that also found no significant differences in isotope signatures between males and females of the cactus moth. However, it seems that sex differences depend on the species. For instance, Hood-Nowotny et al. [23] observed differences in the δ13C signals between males and females of C. capitata reared on artificial larval diets, and Sato and Azuma [43] detected significant differences in δ13C of males and females of Euthrix potatoria L. However, such differences were not large enough to preclude the use of stable isotope as markers of flies reared in laboratory [23]. Future analyses of specific body structures instead of whole body could help to elucidate the existence of any differences in isotopic signatures between sexes of A. fraterculus.

Since aggregation of isotopes from other sources or loss of molecules carrying lighter isotopes can result in improper interpretation of isotopic signatures [54], knowledge of the impact of attractant and preservation methodologies on isotopic values of insect samples is essential. For example, fuel and commercial ethanol caused enrichments in δ 13C and δ 15N of cricket samples as great as 1.5 and 0.9‰, respectively [55]. Hogsden and McHugh [56] compared that ethanol-preserved and lipid-extracted samples of 10 common New Zealand stream invertebrates enriched by 1–2‰ in δ 13C to frozen samples and found that short-term storage in 70% ethanol caused effects comparable to lipid removal. According to these studies, samples subjected to organic solvents can lose dissolved lipids or gain solvent constituent carbon, which can enhance the 13C signal.

Our results corroborate the findings of previous studies, demonstrating that 13C and 15N signatures of fly specimens can be affected by the attractant or preservative substance used and time that flies remain in a trap (Tables 5 and 6). Despite the alterations produced by the attractants tested and ethanol, it was still possible to distinguish laboratory-flies from the wild ones based on the δ13C values at the 5% probability level of significance (Table 5). As field traps are surveyed on a weekly basis [26], the use of CeraTrapTM and grape juice would not compromise the isotope-based method developed in this study. However, the effects of longer storage periods (> 30–60 days) in ethanol should be further investigated to confirm if the fly samples would still show reliable SIA results.

In contrast, the δ15N values were less conclusive compared to δ13C, with no significant differences among some treatments, making it impossible to distinguish the flies reared on artificial diets from wild ones. Compared to the respective diets, the flies reared on larval Diets I and II had an increase in δ15N values of 2.1‰ and 1.8‰, respectively (S2 Table). Flies can produce proteins that are enriched in 15N relative to the food source and periods of starvation can also induce enrichment in some tissues [57, 58]. The fractionation of stable isotopes of nitrogen is more unpredictable than for stable isotopes of carbon [9, 10], and, therefore, traceability should not rely solely on δ15N signals, unless the artificial diet is enriched with N-sources [19].

This study provided the first experimentally derived TDFs of stable carbon and nitrogen isotopes for A. fraterculus. Overall, the Δ13C values ranged from -2.7 to 2.5‰, while Δ15N varied from 0.9 to 2.1‰ (S2 Table). The isotopic enrichments or depletions observed are consistent with published data for other species. In a study involving 22 terrestrial herbivorous arthropods feeding on 18 different host plants, Spence and Rosenheim [59] verified TDFs from plant to herbivores in the range of -3.5‰ to 1.9‰ in δ13C (-0.5±0.3‰ of average) and δ15N enrichments of -0.2‰ to 6.6‰ (1.9±0.8‰ of average). Markow et al. [60] reported differences in δ13C and δ15N between wild-caught Drosophila spp. and their respective natural hosts between -1.8 and 5.2‰ and -0.4 and 5.0‰, respectively. In mass-reared medflies from Madeira Archipelago and Guatemala, the Δ13C values ranged from -1.7 to 2.1‰ and Δ15N ranged from 2.9 to 3.7‰ [23]. Intraspecific variations in TDFs are common and can be expected depending on diet quality, developmental stage and nutritional status of the consumer, different metabolic rates in animal tissues, lipid extraction and other factors [57, 61, 62, 63].

In conclusion, this is the first comprehensive study to demonstrate that it is possible to distinguish wild A. fraterculus from flies reared on larval diets containing C4 sugars. We provide a simple but effective marking method, with the advantage of not requiring changes in the larval rearing protocols. With recent technical advances in mass spectroscopy, the costs of isotopic analyses are now comparable to the most common molecular and chemical analyses (even < US$ 5–15.00/sample).

Our isotope marking system would have a wide application in SIT projects against A. fraterculus. For example, sterile flies are released in forested areas surrounding apple and peach orchards at southern Brazil, but occasionally some of them get caught in traps inside commercial orchards. If the flies are unmarked and a cytohistological analysis of their testis or ovaries is not possible to be performed (e.g. females cannot be easily distinguished on the basis of ovary development when they are younger than 7 days) [64], the SIA could be used to determine if the flies are wild, resolving the doubts of farmers and SIT managers. In addition to the already raised issues that should be further investigated, the next step of our research will be to determine isotopic turnover rates and the persistence of the stable isotopes of C and N during the life of A. fraterculus when the adult diet changes from C4 to C3 food sources in nature.

Dataset containing the raw isotope data collected for Tables 1 to 6.

(XLSX)

Click here for additional data file.

Treatments used for the evaluation of the influence of attractive and preservative substances on the isotopic composition of Anastrepha fraterculus flies (acronyms in parentheses).

(DOCX)

Click here for additional data file.

Mean trophic discrimination factors (Δ) between Anastrepha fraterculus flies and their respective larval diets.

(DOCX)

Click here for additional data file.