Key message

  • The control of the Mediterranean fruit fly, Ceratitis capitata (Wiedemann), is particularly difficult in organic orchards, where synthetic insecticides are not permitted. Natural products may represent a potential tool in managing this insect pest.

  • We tested alcoholic extracts from leaves of four Mediterranean plant species: Ruta graveolens, Eriobotrya japonica, Rubus ulmifolius, and Ficus carica, in order to evaluate their attraction/repellence activity toward ovipositing females and toxicity on adults.

  • Ruta graveolens extract elicited significant results both in terms of behavioral response in EAG and oviposition assays and in terms of induced mortality in toxicity bioassay. The incorporation of such extract or its components into food bait could be useful in medfly “attract and kill” control strategies.

Introduction

The Mediterranean fruit fly, Ceratitis capitata (Wiedemann) (Diptera: Tephritidae), commonly named medfly, is one of the world’s most destructive pests attaching both wild and cultivated plants, and it is considered to be the most invasive of all the Tephritidae (Zucchi 2001). This species is multivoltine and highly polyphagous, being able to attack more than 350 different species of fruits (McQuate and Liquido 2017), thus causing quantitative and qualitative losses to several crops. The high reproductive potential and adaptability of C. capitata combined with the low effectiveness of natural enemies, and its wide host range causes great concern to growers (Castillo et al. 2000).

In order to reduce the medfly populations below injury threshold, chemical control using insecticide baited sprays is a common method used in most countries (Urbaneja et al. 2009). However, the widespread use of chemical products has led to insect resistance, and control failures have been reported in field conditions (Magana et al. 2007; Couso-Ferrer et al. 2011). In organic orchards, the control of C. capitata is particularly difficult as synthetic insecticides are not permitted. The main natural product used in baited sprays is Spinosad, which is expensive and already demonstrated to induce resistance in Bactrocera dorsalis (Hendel) (Diptera Tephritidae) (Hsu and Feng 2006). In organic orange groves, kaolin sprays showed to be effective due to a repellent and antiovipositional action (Caleca et al. 2010; Lo Verde et al. 2011).

The increasing knowledge about the harmful consequences derived from the indiscriminate use of synthetic insecticides has encouraged studies related to novel tactics in pest control, among them the use of natural products, for their efficacy and degradability. More than 2000 plant species are known to possess potential activity against insects (Klocke 1989; Souza et al. 2017), through several mechanisms including contact toxicity, antifeedant/growth inhibitor, suppression of reproductive behavior, and reduction of fecundity and fertility (Carlini and Grossi-de Sá 2002; Isman 2006; Petroski and Stanley 2009; Regnault-Roger et al. 2012; Tak et al. 2017).

With regard to C. capitata, the essential oils from several plants or their components have been investigated for their insecticidal activity, among them Rosmarinus officinalis L., Salvia officinalis L., Lavandula angustifolia Miller, Hyptis suaveolens L., Ocimum basilicum L., Mentha pulegium L., Thymbra capitata L., Thymus spp. (Lamiales Lamiaceae), Thuja occidentalis L. (Pinales Cupressaceae), Melaleuca alternifolia Cheel (Myrtales Myrtaceae), Melia azedarach L. (Sapindales Meliaceae), Ruta graveolens L. (Sapindales Rutaceae), Zingiber officinalis Roscoe (Zingiberales Zingiberaceae), Allium sativum L. ( Liliales Amaryllidaceae), Citrus spp. (Sapindales Rutaceae), Tagetes spp. (Asterales Asteraceae) (Bazzoni et al. 1997; Hamraoui and Regnault-Roger 1997; Sanna-Passino et al. 1999; Chang et al. 2009; Papachristos et al. 2009; Siskos et al. 2009; Miguel et al. 2010; López et al. 2011; Faraone et al. 2012a, b; Benelli et al. 2012, 2013; Rohde et al. 2013; Ruiz et al. 2014).

In order to develop control methods based on natural products, particularly needed in organic orchards, a laboratory study was carried out to evaluate the medfly adult response to alcoholic extracts from four plant species, Eriobotrya japonica (Thunb.) Lindl. (Rosales Rosaceae), Ficus carica L. (Urticales Moraceae), Rubus ulmifolius Schott (Rosales Rosaceae), and R. graveolens, collected in Tunisia. These species were chosen as are known for their richness of bioactive secondary metabolites (Milesi et al. 2001; Vaya and Mahmood 2006; Sultana et al. 2014; Chaftar et al. 2015), but their alcoholic extracts have not been investigated for their insecticidal or behavior-modifying on adult tephritid flies. The effectiveness of extracts from these plants on arthropod pests has been investigated for R. graveolens versus Pediculus humanus capitis De Geer (Anoplura Pediculidae, Jorge et al. 2009) and Aedes aegypti L. (Diptera Culicidae, Tabanca et al. 2012). Besides, aqueous extract from R. graveolens was tested on C. capitata, but the list of the compounds present was not reported (Rohde et al. 2013). Moreover, F. carica extracts were tested against Panonychus citri (McGregor), Tetranychus urticae Koch (Acarina Tetranychidae), Aphis gossypii Glover, Myzus persicae Sulzer (Hemiptera Aphididae), and Trialeurodes vaporariorum (Westwood) (Hemiptera Aleyrodidae) (Kim et al. 2005; Chon et al. 2008).

The use of these four species is encouraged by their inexpensiveness, due to their wide availability in many countries of Mediterranean Basin. Therefore, they can be considered an economically sustainable source of plant extracts as useful tools in pest management programs (Ghosh et al. 2012).

Ethanolic extracts from these four plant species were analyzed and tested in laboratory bioassays on C. capitata adults to assess the electroantennographic response, attraction/repellence toward ovipositing females, and toxicity on adults.

Materials and methods

Insects rearing

Experiments were carried out at the Department of Agricultural, Food and Forest Science of the University of Palermo (Italy), using C. capitata adults obtained from artificially reared colonies maintained, for several generations, in the insectarium facility of the CIHEAM–IAMB, Mediterranean Agronomic Institute of Bari (Italy). A colony of C. capitata has been established in 2000 from individuals obtained from infested fruits growing in organic orchards from Apulia, Italy. The colony of C. capitata was regularly renewed or supplemented with field-collected adults. Ceratitis capitata was reared according to the commonly adopted procedures, as described in Raspi and Loni (1994), Loni (1997), Canale and Benelli (2012), and Oreste et al. (2015). The insect rearing was carried out at 24–26 °C temperature (Loni 1997), 64% relative humidity, and 12/12 D/L photoperiod (Carey 1984). About 8000 C. capitata adults (sex ratio 1/1) were kept in Plexiglas cages (40 × 50 × 45 cm) and provided with a medium consisting of protein bait (30 g), dry yeast (8.4 g), sugar (40 g), and water (40 mL); wet paper was used to supply water. One side of the cage consisted of a net providing a surface for females to lay their eggs, which dropped into a water reservoir located at the base of the cage. Eggs were collected daily and distributed in cups filled with artificial food consisting of bran (96.8 g), sugar (64.8 g), dry yeast (9.07 g), citric acid monohydrate (2.4 g), and sodium benzoate (2 g), suspended in 200 mL of water (Cavalloro and Girolami 1969). The cups, each containing 375 g of food mixture and 0.25 mL of eggs, were placed into the insectarium growth chambers. Mature larvae were collected in water-filled trays, then removed and placed in growing room conditions to obtain pupae and, after about 10 days, adults.

Plant material

Leaves from E. japonica were collected from a commercial orchard in the Rafraf area (27°31′0″N, 41°41′0″E—Bizerte, north-east Tunisia). F. carica and R. ulmifolius leaves have been collected from cultivated plants located in the botanical garden of INRAT; National Agricultural Research Institute of Tunisia, (36°47′51″N, 10°09′57″ E, Ariana, north Tunisia). Leaves of R. graveolens were collected from a wild population in the region of El Kef (36°11′10″N, 8°42′00″E, north-west Tunisia). The leaves of each plant were air dried for 2 weeks at 26 ± 2 °C. Dried leaves were powdered and stored in glass containers, protected from humidity, light, and extreme temperatures (above 30º C), for later extractions and analysis.

Extraction procedure and chemical analysis

Air-dried leaf powders from each plant were subjected to a cold extraction by ethanol. Cold extraction method was chosen to avoid damaging the compounds obtained. Moreover, the use of pure extract or ethanolic solutions with water is duly authorized in organic farming. In each experiment, 30 g of powder was shaken at 300 rpm with 300 mL of ethanol overnight. Afterward, the resulting extracts were filtered, and solvent was evaporated under reduced pressure at 50 °C, using rotary vacuum evaporator, to the minimum ethanol volume to avoid any precipitation. The resulting residues were sealed and kept under refrigerated conditions (4–6 °C) until used. An Agilent 6890 gas chromatograph instrument, equipped with the mass spectrometer detector Agilent 5975 B, was used for the chromatographic analyses. A fused silica capillary column SLB-5MS (length 30 m, internal diameter 0.25 mm, 0.25 μm film thickness of silphenylene polymer equivalent in polarity to poly-5% diphenyl/95% dimethyl siloxane phase) from Supelco, Italy, was the stationary phase. The injector, in splitless mode, had a temperature of 250 °C. Experimental chromatographic conditions were as follows: helium carrier gas at 1 mL min−1; oven temperature program: 5 min isotherm at 40 °C followed by a linear temperature increase of 4 °C min−1 up to 200 °C held for 2 min. MS scan conditions were ionization technique, electronic impact (EI) at 70 eV, source temperature 230 °C, interface temperature 280 °C, and mass scan range 33–350 m/z. The injected sample (1:100 v/v diluted in ethanol) was 1 μL. For quantitative results, each sample was analyzed in GC-FID. The instrumental conditions for the gas chromatograph were the same as above reported. The FID detector was set at 250 °C, and 1 μL was injected. The quantitative composition was obtained by peak area normalization, the response factor for each component was considered equal to one, and three replicates of each sample were made. Internal standard was undecane. Identification of the individual components was based (1) on comparison with retention time and spectra of authentic standards (Sigma Aldrich, Milan, Italy), (2) on comparison of their GC retention indices (RI), determined relatively to the retention time of a series of n-alkanes, and (3) on computer matching with mass spectral libraries (NIST 07) or literature data.

EAG bioassays

Electroantennogram recordings on C. capitata females (8–10 days old) were carried out by testing the plant extracts of R. graveolens, F. carica, R. ulmifolius, and E. japonica. Medfly females used for this experiment were anesthetized with CO2, and their head was cut at the base. For EAG preparation, glass capillary tubes filled with 0.1 M KCl solution were connected to the silver wire recording and reference electrodes. The first was placed gently touching the tip of one randomly selected antenna, while the second was placed at the base of the head. A standard 2 μL aliquot of each plant extract was pipetted onto a piece of filter paper (Whatman No. 1), exposed to air for 5 min allowing the solvent to evaporate, and then inserted into a glass Pasteur pipette. A stimulus-flow controller (model CS-05; Synthech, the Netherlands) was used to generate a 3.0 s stimulus at 1 min interval, with a flow rate of 1.5 L min−1. The signals generated by the antennae were passed through a high impedance amplifier (model IDAC-4, Synthech, Hilversum, the Netherlands) and recorded with specialized software (Synthech). At the beginning and the end of the stimulation of the antenna with each plant extract, 2 μL of pure ethyl alcohol was puffed as control stimulus. The same antenna was used to test all of the plant extracts. Each extract was tested on ten females (N = 10) antennae using one antenna per fly. The sequence of the tested extracts was randomized.

Behavioral bioassay

Experiments were performed using sexually mature adults (at least 8–10 days old) fed with a mixture of sucrose, casein, and yeast (ratio 4:3:3), supplied on a wet cotton ball. Medflies had never been exposed to oviposition before being used in the bioassays.

The insects (10 females and 10 males per each test) were moved into 6 cylindrical cages (33 cm length, 30 cm diameter), in which low-density-polyethylene yellow plastic spheres (Ø = 50 mm) were placed. The size and color of spheres were chosen for their similarity with fruit species that are suitable for medfly oviposition (Katsoyannos et al. 1986; McInnis 1989; Faraone et al. 2012a).

To encourage oviposition, twenty holes (Ø = 0.5–1.0 mm) were drilled on each sphere around the middle, and 2 mL of attractant liquid was pumped inside by means of a syringe. The attractant was composed of 40 g sucrose, 8.3 g dry yeast, 21.7 mL water, and 30 g NuBait® (Biogard, CBC Europe, Italy), a protein bait used in the control of tephritid flies (Mattedi et al. 2016; Sacchetti et al. 2017).

Each sphere was suspended to a cotton yarn of c.a. 30 cm within the Potter Tower and sprayed (pressure 9–10 psi) with 8 mL of hydroalcoholic solution (10% v/v ethanol) containing 800 μL of extract for each plant (treatment) or with 8 mL of hydroalcoholic solution (10% v/v ethanol) in the control. After 30 min (needed for solution drying), one treated and one untreated sphere were suspended in each cage, letting the flies lay eggs inside the spheres through the holes. After 3 days, spheres were cut and the number of eggs laid inside them was recorded. For each tested stimulus, six replicates were carried out.

Toxicity bioassays

The toxicity bioassays were conducted for all plant extracts, using Petri dishes (90 mm diameter, 15 mm height) as arenas. Both sides of Petri dishes were lined with a filter paper, each one treated by 1 mL of a water solution containing 100 μL of each plant extract, while the control group was treated by an equal amount of water solution containing 100 μL of ethanol. Four replicates were conducted for each test, placing ten adults inside each Petri dish. Both treated and untreated insects were maintained under laboratory conditions (25 ± 1 °C, 75% R.H., 16:8 h L:D photoperiod) and fed with a water solution containing 20% of organic honey, placed on a cotton wick. Mortality was evaluated 24 h after the treatment. After this first screening, the plant extracts that elicited significant mortality were tested in a second experiment, carried out using the following plant extract concentrations: 12.5, 25.0, 37.5, 50.0, and 100.0 μL diluted up to a final volume of 1 mL.

Statistical analysis

The EAG responses to each plant extract, given as means of antennal responses in mV, after a Box-Cox data transformation, were analyzed by one-way ANOVA followed by Tukey test. Data obtained from behavioral bioassays expressed as the mean percentage of eggs laid on control and treated spheres were compared for each plant extract, using a t-test. Mortality data obtained in toxicity bioassays using the same concentration of the four plant extracts have been preliminarily analyzed using binary logistic regression, to assess the differences among treatments that showed percentage mortality > 0% or < 100%. Afterward, a nonparametric Kruskal–Wallis analysis of variance was applied, which is independent of the distribution of the data. For tests carried out for toxicity assessment using different extracts concentrations, a probit analysis by means of maximum likelihood estimation method was performed, calculating also the lethal concentration 50% (LC50) and the lethal concentration 90% (LC90) as the antilogarithm of the predicted dose values corresponding to a mortality rate of 50 and 90%, respectively. Chi-square test was applied to evaluate the goodness of fit (Finney 1971). All data were analyzed using Minitab® (Minitab Inc., State College, Pennsylvania, USA).

Results

Plant extracts

The constituents and their percentage recorded in plant extracts are shown in Table 1. The principal constituents (> 5% of the total of the extracts compounds) of the leaves of E. japonica are α-terpinene (a hydrocarbon monoterpene, 20.4%), terpinen-4-ol (a monoterpene alcohol, 15.7%), trans-dihydronerolidol (a sesquiterpene alcohol 14.5%), and palmitic acid (a fatty acid, 8.8%). The two furanocoumarins angelicin (72.1%) and bergapten (8.9%) were the main components of F. carica extract. The extract of R. ulmifolius contained palmitic acid (a fatty acid, 23.4%), phytol (a linear diterpene alcohol, 23.1%), and linolenyl alcohol (an unsaturated alcohol, 20.4%). Finally, in R. graveolens, main compounds were 2-undecanone and 2-nonanone (both ketones, 34.5 and 10.1%, respectively) and chalepensin (a furanocoumarin, 8.7%).

Table 1 Chemical composition (%) of the four plant extracts and Retention Index (R.I) of identified compounds
Full size table

EAG bioassays

Results of EAG bioassays are reported in Fig. 1. The antennae of C. capitata females showed difference in the magnitude of the EAG responses elicited by the different plant extracts tested (F4.45 = 9.372, P < 0.001; one-way ANOVA). Specifically, R. graveolens extract produced a greater EAG response in comparison with the extract of E. japonica, R. ulmifolius and the solvent (P < 0.001). The F. carica extract showed a greater EAG response compared to the solvent (P = 0.011) but not in comparison with the other extracts. Finally, the EAG responses elicited by E. japonica or R. ulmifolius extracts did not differ from the solvent response (P > 0.81).

Fig. 1
figure 1

Mean (+ SE) EAG responses of female C. capitata to test performed using 2 μL of the leaf extracts. Columns labelled with a common letter do not differ significantly at P < 0.05 (ANOVA followed by Tukey test)

Full size image

Behavioral bioassay

Results of oviposition tests are reported in Fig. 2. No significant differences in the number of eggs laid in control and treated spheres were recorded for E. japonica, F. carica, and R. ulmifolius (t-test, P = 0.67, 0.57 and 0.71, respectively, df = 10 for all tests). Only for R. graveolens the number of eggs laid by C. capitata females was significantly greater in treated spheres than in untreated spheres (t-test, P = 0.03; df = 10).

Fig. 2
figure 2

Mean (+ SE) number of eggs laid by C. capitata females. Asterisk indicates significant differences between control and treatment (t-test, P < 0.05; df = 10)

Full size image

Toxicity assessment

The number of dead flies ranged from a minimum of 0 in the control to a maximum of 40 (100%) in bioassay carried out using R. graveolens extract. In bioassays carried out using extracts of F. carica, R. ulmifolius, and E. japonica, the number of dead flies recorded was, respectively, 5 in the first and 1 in the other two plant species (Fig. 3). As binary logistic regression showed that no significant differences were present among E. japonica, F. carica, and R. ulmifolius (χ2 = 4.52; P = 0.10; df = 2), a Kruskal–Wallis statistical test was applied to compare control, R. graveolens and merged data from the other three extracts, showing that differences among the treatment groups were present (H = 87.65; df = 2; P = 0.000). Results of the two statistical analyses indicate that R. graveolens was the only plant extract different from the other treatments. Therefore, further bioassays were conducted only with R. graveolens extracts, applying the plant extract at different concentrations. As a result of the probit analysis, the likelihood ratio chi-square test showed a good fit of the probit model (χ2 = 3.29; P = 0.35), indicating a significant dose-dependent action for the extract obtained from R. graveolens (Slope 5.89 ± 0.85) and leading to assess that LD50 and LD90 were 36.4 μL (95% FL: 3.30–4.02) and 60.1 μL (95% FL: 5.24–7.53), respectively.

Fig. 3
figure 3

Mean (+ SE) mortality percentage recorded in the toxicity tests carried out with the four plant extracts. Bars labelled with a common letter do not differ significantly at P < 0.05 (Kruskal–Wallis non-parametric test)

Full size image

Discussion

The results of this study showed that R. graveolens extracts are strongly perceived physiologically and induced oviposition activity in C. capitata females. The strong EAG response elicited by R. graveolens extract suggests that the compounds contained in this blend may have influence on the insect behavior. As the EAG response is generally linked with insect olfaction, it is reasonable that this response was primarily triggered by the more volatile fraction of the extracts. In general, EAG results are strictly related to the chemical composition of tested compounds or blends. The leaf extract of R. graveolens is characterized by the high presence of ketones as 2-nonanone and 2-undecanone, which compose about 45% of the total extract. These two compounds are reported to elicit strong EAG response to some related species as the Diptera Tephritidae Rhagoletis cerasi L. (Raptopoulos et al. 1998) and Rhagoletis mendax Curran (Lugemwa et al. 1989). The screening of extracts or synthetic chemicals for their antennal activity using EAG recordings is an important first step to select candidates for behavioral tests (Cossé et al. 1995; Guarino et al. 2013; Ruschioni et al. 2015). In general, attraction behavior shown by dipteran species to a source is often mediated by the main volatiles of the chemical blend (Zito et al. 2013). In particular, 2-nonanone is reported as component of commercial tephritid attractants (Yufera et al. 2006). Among the less abundant compounds presents in the R. graveolens extracts, 2-nonanol (1.69%) and 2-undecanol (1.56%) have been reported among the volatiles of Citrus spp. fruits, and to be perceived by the other fruit fly species such as Bactrocera dorsalis Hendel (Diptera Tephritidae) in EAG experiments (Light and Jang 1987). Among the other plant leaf extracts tested, F. carica elicited EAG significant response compared with the solvent, despite the small fraction composed of volatile compounds. However, among the volatiles of F. carica leaf extract, β-caryophyllene (0.35% of the total extract) is known to stimulate strong antennal response on C. capitata (Cossé et al. 1995).

In accordance with the EAG tests, oviposition behavioral bioassays conducted in this study showed increased egg-laying activity by C. capitata females on spheres treated with the R. graveolens extracts. The extract from R. graveolens leaves acted in this case as a kairomone, presumably resulting in attraction (not quantified) followed by increased oviposition activity. Interestingly, this positive oviposition response was recorded for an extract coming from a non-host plant as R. graveolens. In another study conducted on Mansonia spp. (Diptera Culicidae), R. graveolens extract is reported to be repellent by applying it on human legs (Hadis et al. 2003). Moreover, Aivazi and Vijayan (2010) observed that R. graveolens extract can synergize the insecticide action against Anopheles stephensi Liston (Diptera Culicidae). In the toxicity bioassays, the comparison among the four plant extracts showed that exposure of adult C. capitata to R. graveolens extract resulted in 100% mortality. The insecticidal activity of R. graveolens extract, whether absorbed by tarsal gustatory sensilla of adults or inhaled after their vaporization inside Petri dishes, might be elicited by both the 2-undecanone and by the “heavy” fraction, containing the furocoumarins chalepensin (8.30%) and isopimpinellin (0.19%). As regards 2-undecanone, it has been already reported as toxic to arthropods (Akkari et al. 2015), whereas furocoumarins are known to have insecticidal effect determined by acetylcholinesterase inhibitory activity (Kang et al. 2001). On the other hand, in this study, F. carica extracts containing other furocoumarins like angelicin and bergapten, in relatively high amounts, did not exert toxic effects on C. capitata adults.

With regard to C. capitata, few literature data are available on the toxicity of plant extracts toward adults. López et al. (2011) found a marked toxicity of Tagete spp., whereas Benelli et al. (2012, 2013) tested essential oils extracted from fresh leaves of R. officinalis, L. angustifolia, H. suaveolens, T. occidentalis, and M. alternifolia, obtaining for all plant species remarkable toxic effects with regardless of the bioassays used (ingestion, contact, fumigation).

To our knowledge, this is the first study to report a plant extract that works as kairomone and at the same time have insecticidal properties toward C. capitata adults. The only similar study of plant extract reporting similar effect toward arthropods was carried out by Zorloni et al. (2010) using extracts of Calpurnia aurea (Ait.) Benth. (Fabales Fabaceae) leaves, that attracted, immobilized or killed the tick Rhipicephalus pulchellus Gerstäcker (Ixodida Ixodidae).

Further studies could investigate the role of the single components of R. graveolens extracts in order to better evaluate attractant/repellent and/or insecticidal properties. In addition, the efficacy and chemical stability of R. graveolens extract should be tested in field trials to assess the effect of environmental conditions on degradation and/or evaporation of some or all active components. The incorporation of such extracts into food bait could represent a useful tool in medfly “attract and kill” control strategies. Laboratory research carried out bioassaying plant extracts as candidate tools for pest management is an essential step for the future authorization of their use in agriculture, particularly when strict regulatory regimes such as in the EU are in force. Moreover, the use of plant species that are widely spread or cultivated can provide an economically sustainable source of extracts. In our case, the possible use of R. graveolens extracts into food baits could result in a low environmental impact and in an increased feasibility of using them, especially in organic farming, where a low number of products are currently authorized for insect pest management.

Author contributions

All of the authors conceived and designed the experiments; NB worked for insects rearing; FS performed the extraction from plant material and chemical analyses; MG, GLV, MS, and SG performed the toxicity and behavior experiments and analyzed the data; all of the authors interpreted the results and drafted and revised the manuscript.