Resumen:
El control y monitoreo de Anastrepha ludens Loew (Diptera: Tephritidae) implica un efecto negativo subyacente de la captura de organismos no objetivo.
Este estudio tuvo como objetivo analizar y comparar la captura de artrópodos no objetivo entre los métodos de control y las trampas de monitoreo. Se establecieron siete tratamientos en huertos de cítricos, con tres réplicas cada uno de mayo a agosto de 2022.
Cuatro tratamientos consistieron en trampeo masivo, dos con aspersiones terrestres y un tratamiento de control. Se utilizaron tres trampas de monitoreo por tratamiento cebadas con Cera Trap® y pellets de bórax de levadura torula para monitorear la captura de organismos no objetivo. Se utilizaron modelos lineales mixtos y tablas de contingencia para evaluar el nivel de captura entre los tratamientos de control y las trampas de monitoreo. El número total de especímenes no objetivo fue de 46.084, 41 familias de Insecta y dos de Arachnida. Diptera fue el grupo más diverso, con 17 familias. Catorce familias representaron organismos benéficos en cinco órdenes. No se observaron efectos entre los métodos de control en la captura promedio de artrópodos no objetivo y organismos benéficos. Las trampas de monitoreo mostraron diferencias en la captura promedio de artrópodos no objetivo y organismos benéficos. La levadura de torula fue el atrayente con la mayor cantidad de capturas de insectos.
El trampeo masivo con Cera Trap® o GF 120® con trampas cebadas con Cera Trap® mostró una disminución en la captura de organismos benéficos. Se necesita más investigación a nivel de campo para evaluar el impacto de los métodos de control de A. ludens en especies no objetivo para fines de conservación.
A comparison of the capture of non-target arthropods between control methods and monitoring traps of Anastrepha ludens in citrus agroecosystems
Venancio Vanoye-Eligio*, Edgar M. Cotoc-Roldan, María de la Luz Vázquez-Sauceda, Leroy Soria-Díaz and Griselda Gaona-García
Abstract: Control and monitoring of Anastrepha ludens Loew (Diptera: Tephritidae) involve an underlying negative effect of the capture of non-target organisms. This study aimed to analyze and compare the capture of non-target arthropods between control methods and monitoring traps. Seven treatments were established in citrus orchards, with three replicates each in May to August 2022. Four treatments consisted of mass trapping, two using bait sprays, and a control treatment. Three monitoring traps per treatment baited with Cera Trap® and torula yeast borax pellets were used to monitor the capture of non-target organisms. Linear mixed models and contingency tables were used to evaluate the capture level between control treatments and monitoring traps. The total number of non-target specimens was 46,084, 41 families of Insecta and two of Arachnida. Diptera was the most diverse group, with 17 families. Fourteen families represented beneficial organisms in five orders. No effects were noted between control methods on the average capture of non-target arthropods and beneficial organisms. Monitoring traps showed differences in the average capture of non-target arthropods and beneficial organisms. Torula yeast was the attractant with the most insect captures. Mass trapping with Cera Trap® or GF 120® with traps baited with Cera Trap® revealed a decreased capture of beneficial organisms. Further research is needed to assess the impact of A. ludens control methods on non-target species for conservation purposes at the field level.
1 Introduction
Control and suppression of adult populations of the Mexican fruit fly, Anastrepha ludens (Loew) (Diptera: Tephritidae), and other tephritid pests in diverse agroecosystems involve the strategical combination of several methods, chiefly toxic bait sprays, mass trapping, bait stations, biological control, and release of sterile flies (Hendrichs et al. 2005; Lasa et al. 2014a; Loera-Gallardo et al. 2012; Montoya et al. 2007; Navarro-Llopis et al. 2015). Moreover, monitoring of pest populations through trapping activities at a local scale and over agricultural landscapes encompasses the use of different attractants and devices aimed at improving the detection of fruit fly populations (FAO/IAEA 2003; Lasa et al. 2015; Martinez et al. 2007).
Citrus agroecosystems are shelters for arthropod communities that provide ecological services to pest control through beneficial organisms, such as predators and parasitoids (Ruíz et al. 2006; Sorribas et al. 2016). These environments are commonly treated with ground sprays based on organophosphates or spinosad protein bait to reduce
A. ludens infestation levels (Conway and Forrester 2011; SAGARPA/SENASICA 2012). Mass trapping based upon improved protein lures at an appropriate trap density, is an alternative control method against Anastrepha pest species that is more environmentally friendly (Lasa et al. 2014b; Navarro-Llopis and Vacas 2014; Villalobos et al. 2017). For example, the ready-to-use formulation Success® 0.02 CB (Spinosad; CTVA Proteção de Cultivos Ltda., Barueri, Säo Paulo, Brazil) showed low mortality in adults of Doryctobracon areolatus (Szépligeti) (Hymenoptera: Braconidae), a native parasitoid of Anastrepha species (Piovesan et al. 2023). Mass trapping based on Cera Trap® (Bioiberica, Barcelona, Spain), a food-based attractant mixed with a toxic active ingredient, could be an ecological approach to reducing fruit fly infestations (Villalobos et al. 2017).
Previous studies focused on attractants based on torula yeast borax pellets, and a synthetic 2-component lure (putrescine and ammonium acetate) reported a high percentage of captures, particularly of Diptera. However, traps baited with the 2-component lure reported lower captures of predator and parasitoid guilds than traps baited with torula yeast (Martinez et al. 2007; Thomas 2003a). Moreover, in citrus orchards, the three-component BioLure (ammonium acetate, trimethylamine hydrochloride, and putrescine) was more attractive to non-target dipterans than torula yeast, particularly for Drosophilidae, Neriidae, Phoridae, Calliphoridae, Sarcophagidae, and Muscidae (Leblanc et al. 2010). Likewise, grape juice lures had less selectivity against non-target insects and higher capture of beneficial lacewings (Herrera et al. 2015). The susceptibility of bene- ficial species to control treatments targeting fruit fly pop- ulations seems to vary depending on the type of bait, experimental conditions, and even the life stages of the organisms (Michaud 2003; Thomas and Mangan 2005; Yee and Phillips 2004).
Food attractants are essential to control methods to suppress or control A. ludens populations in citrus agro- ecosystems (Conway and Forrester 2011; Lasa et al. 2014a). The capture of non-target insects and beneficial organisms may vary based on the specificity of the lure and how it is applied in the field. We conducted an experimental field study that compared the number of captures of non-target arthropods in citrus orchards using different control methods and monitoring traps for A. ludens adult populations.
2 Materials and methods
2.1 Study area
The study was conducted in an agricultural landscape composed of citrus orchards at the locality of Marte R. Gómez (24.0886111°N; 99.0294444°W) in the Padilla municipality of Tamaulipas in northeastern Mexico (Figure 1) from May to August 2022. The field study covered the end of the harvest of the Valencia orange and most of the summer season. This period is characterized by fruit scarcity, limiting fly development conditions and the number of captured flies (Thomas 2003b). Based on the drought monitor of Mexico (SMN 2023), a moderate to severe drought affected this municipality in the first six months of 2022. Average temperatures ranged from 27 to 34 °C and maximum temperatures from 34 to 44 °C. The region registered an accumulated precipitation of 83 mm during the study period.
2.2 Experimental design and control methods
2.2.1 Mass trapping
Seven 1-ha plots 16 m apart were established in a block design with three blocks (replications) and seven treatments per block (Figure 1). There were four treatments involving mass trapping: (1) PET (polyethylene terephthalate) traps with 250 mL of hydrolyzed enzymatic protein Cera Trap® (Bioiberica, SA, Barcelona, Spain) at 25 traps/ha (CT25); (2) PET traps with 250 mL of Cera Trap® at 40 traps/ha (CT40); (3) PET traps with 250 mL of the mix malathion 50 % + Captor 300® (1:4) (Promotora Agropecuaria Universal, S.A. C.V., Lucava, Mexico) hydrolyzed protein at 25 traps/ha (MC25); and (4) PET traps with 250 mL of GF-120® bait (Naturalyte® Fruit Fly Bait, Dow AgriSciences LLC, Indianapolis, Indiana) at 25 traps/ha (GF25). All PET bottles were modified by drilling four 5 mm diameter holes, 5 cm apart and three fourths above the base. These bottles were evenly distributed in the treatment plot. In all cases, PET traps were rebaited every 15 days during the study.
Figure 1: Geographic location of the citrus area in Tamaulipas, Mexico. Additionally, the location of the three blocks contains an example of the arrangement of the six control methods and the control treatment. Black dots represent the three monitoring traps placed at the center of each treatment. CT25 = 25 PET traps with Cera Trap®; CT40 = 40 PET traps with Cera Trap®; MC25 = 25 PET traps with the mix malathion 50 % + Captor 300®; GF25 = 25 PET traps with GF-120®; BSGF = bait sprays with GF-120®; and BSMC = bait sprays with malathion and captor; and CT = control treatment.
2.2.2 Ground bait sprays
Two treatments based on ground bait sprays were evaluated: (5) GF-120® bait mixed with water (1:1.5 GF-120:water) (BSGF) applying 4 L of the mix per hectare in alternate tree rows; (6) a mix of malathion 50 % + Captor 300® + water (BSMC) (1:4:95) applying 325 mL per tree in alternate rows. This procedure was repeated every seven days during the study. The amount needed on each tree was previously calibrated in a graduated container before each application. Bait sprays were applied through a 20 L capacity hand-operated backpack Swissmex® sprayer. Equipment nozzles were calibrated to obtain size drops between 3 and 6 mm. Ground bait sprays followed the manual’s guidelines provided by SAGARPA/SENASICA (2012).
2.2.3 Entomofauna monitoring
Three monitoring traps placed at the center of each treat- ment plot evaluated the effect of the control method on the insect community. A Multilure® trap (Better World Manufacturing Inc., Fresno, California) with three pellets of torula (MT), a Multilure® trap with Cera Trap® (MC), and a PET trap with Cera Trap® (BC). The control treatment did not employ any control method and involved establishing the three monitoring traps. The traps were hung in a citrus tree 3–4 m above the ground, serviced weekly, and rotated clockwise (Figure 1). Specimens captured were separated according to the monitoring trap and identified at a family level in the taxonomy laboratory of the National Campaign against Fruit Flies in Tamaulipas. Family-level identification provides sufficient taxonomic information for interpreting biodiversity patterns and the study of insect communities (Bevilacqua et al. 2021; Timms et al. 2013).
2.2.4 Statistical analyses
All statistical analyses were performed with R (R Development Core Team 2018). A constant of 1 was added (Y + 1) to the total sum of the captured specimens by treatment and traps to avoid zeros. We obtained average values of captured specimens from the monitoring traps of each treatment, which were log-transformed to stabilize the variance. A Levene test was performed to evaluate the homoscedasticity of the data. To analyze the effects of control treatments and monitoring traps (fixed effects) on the capture of non-target and beneficial organisms, a linear mixed model (LMM) fitted by the maximum likelihood method using the lme4 package (Bates et al. 2015) was performed using the log-transformed data of captures as a response variable. The number of times that traps were inspected, treatments, monitoring traps, and blocks were introduced into the model as random effects to account for the field effect and correlated data on the capture of specimens. Such effects were introduced as a nested structure based on the hierarchical arrangement: blocks, treatments, and type of trap. A signif- icance level of 0.05 was considered. The emmeans package (Lenth 2023) was used to determine differences between the least-square means (LS-means) of control treatments and the type of monitoring trap regarding the capture of arthropod and beneficial organisms. If applicable, the Tukey method adjusted p-values for multiple comparisons. Contingency tables were performed to evaluate the independence between the proportion of captured specimens and beneficial organisms concerning the combination of the type of trap and control treatments. Mosaic plots were displayed using the vcd package (Zeileis et al. 2007) to show significant relationships between the frequency of specimen captures and control treatments.
3 Results
Monitoring traps registered 46,084 specimens of eight orders representing 41 families of Insecta and one order with two families of Arachnida (Table 1). The largest group was Diptera with 17 families, followed by seven of Coleoptera, six of Hymenoptera, three of Blattodea, three of Lepidoptera, three of Hemiptera, one each of Orthoptera and Neuroptera, and two of Araneae (Tables 2 and 3). Flies represented the highest proportion of captured insects, followed by ants (Hymenoptera: Formicidae). Ants were the predominant group in the capture of beneficial organisms, followed by the families Anthicidae (Coleoptera), Clubionidae (Arachnida: Araneae), and Chrysopidae (Neuroptera). Ant specimens represented 98 % of 7,779 captured beneficial organisms (Table 4).
The order Diptera represented about 82 % of captured insects, belonging mainly to the family Muscidae, which represented 78 % of the dipterans and 64 % of the total of non-target arthropods, followed by Drosophilidae and Phoridae. Rare specimens of Diptera were represented by the families Neriidae, Rhinophoridae, Syrphidae, and Chloropidae. Also, beneficial insects, although to a much lower proportion, were included in this group, such as the family Cecidomyiidae, Tachinidae recognized as parasitoids, as well as Asilidae and Syrphidae as predators of agricultural pests.
No effects on the capture of non-target organisms were shown among the three blocks or control treatments ( χ2 = 0.346, df = 2, p = 0.8408; χ2 = 2.02, df = 6, p = 0.917, respectively). However, the type of monitoring trap and its attractant showed an effect on the average capture of non-target arthropods ( χ2 = 82.3, df = 2, p < 0.001) (Figure 2). Moreover, no effects were noted in the average capture of beneficial organisms among the three blocks or control treatments ( χ2 = 8.582, df = 6, p = 0.1985), but an effect was observed among the type of monitoring trap ( χ2 = 16.123, df = 2, p < 0.001).
Contingency tables revealed a significant relationship between treatments and the type of traps regarding the number of captured insects ( χ2 = 1,503.3, df = 12, p < 0.001) and a significant association in the number of captures of beneficial organisms between the type of trap and treat- ments ( χ2 = 207.4, df = 12, p < 0.001) (Figure 3). Multilure traps baited with torula yeast pellets accounted for the most sig- nificant proportion of captures of non-target insects in all treatments. By contrast, monitoring traps with Cera Trap® (BC and MC) accounted for the lowest proportions of captured non-target organisms.
In the case of beneficial organisms, the mosaic plot depicted fewer significant cells and capture proportions seemed to be more homogeneous. However, the MC trap revealed the lowest capture of beneficial organisms in almost all treatments. The CT40 treatment (mass trapping) exhibited a significant relationship with the three monitoring traps. Mass trapping with Cera Trap® and GF 120® associated with the BC trap revealed a negative trend in the counts of beneficial organisms (Figure 3).
Table 1: Number of captured specimens by class and order and distributed by block and type of monitoring trap in citrus agroecosystems in Mexico. BC = plastic bottle with Cera Trap®; MC = multilure trap with Cera Trap®; and MT = multilure trap with torula yeast borax.
Table 2: Counts of captured specimens in citrus agroecosystems in Mexico, classified by family, order, and class, except Diptera.
Table 3: Counts of captured specimens of Diptera in citrus agro- ecosystems in Mexico, structured by families depending on the type of monitoring trap. BC = plastic bottle with Cera Trap®; MC = multilure trap with Cera Trap®; and MT = multilure trap with torula yeast borax.
4 Discussion
This study showed no effects of control treatments on the average capture of non-target arthropods, including beneficial insects. However, monitoring traps baited with torula yeast borax pellets provided evidence of a higher capture level of non-target organisms and beneficial insects. Fortyone families represented the diversity of non-target organisms, highlighting the orders Diptera and Hymenoptera (ants) as taxa with more captured specimens in the traps. Dipterans were distributed in 17 families and Diptera was the taxa with the highest number of captures. Several groups of beneficial insects belonging to the orders Hymenoptera, Coleoptera, Diptera, Araneae, and Neuroptera were observed, although their captures were low.
Table 4: Number of captures of beneficial arthropods by block and type of monitoring trap and classified by class and order in citrus agroecosystems in Mexico. BC = plastic bottle with Cera Trap®; MC = multilure trap with Cera Trap®; and MT = multilure trap with torula yeast borax.
The prevalence of dipterans and ants in the traps characterized the capture of non-target insects. An ecological indicator may be the predominance of saprophagous species, particularly the Muscidae family. This group represented the highest percentage of non-target insects associated with decaying fruit or dead organic matter. In contrast, the number and diversity of beneficial organisms were quite low compared to non-target organisms. Generalist predators, such as ant species, were the most abundant in the traps. Ants (Formicidae) and spiders (Salticidae) are identified as predators of larvae and adults of Tephritidae species, respectively (Aluja et al. 2005; Rao and Díaz-Fleischer 2012). Several other families of beneficial insects included Anthicidae, Chrysopidae, and Tachinidae. The Anthicida, known as ant-like flower beetles, involves some species considered predators of mites, small arthropods, and larvae of several insect pests (Werner and Chandler 1995). Chrysopids are common predators of several pest species in citrus groves. A preceding study indicated that lacewings may be susceptible to malathion bait sprays against tephritid pests (Michaud 2003). However, it seems to be that this group was not significantly affected by the formulation GF 120® (Thomas and Mangan 2005). The Tachinidae representsa parasitoid group linked to larval Lepidoptera. Tachinids are present in almost all terrestrial ecosystems and may constitute a large proportion of dipterans observed in diverse habitats (Stireman et al. 2006). The observed proportion of captured beneficial organisms in the current study resembles that reported by Thomas (2003a) in citrus groves of northeastern Mexico.
The attractant and type of monitoring trap significantly affected the capture of non-target arthropods. In this case, Cera Trap® minimized the capture of non-target organisms, as observed by Delgado et al. (2022). This lure used for mass or monitoring trapping of Anastrepha pest populations is an effective product based on enzymatically-hydrolyzed animal proteins (Lasa and Williams 2022; Lasa et al. 2014b). Also, it is attractive to arboreal ants (Hymenoptera: Formicidae) and is recognized in the trapping activities of fruit flies (García-Martínez et al. 2018; Vanoye-Eligio et al. 2020). According to our results, monitoring traps baited with Cera Trap® caught 80 % less non-target organisms compared to torula yeast. In this sense, the torula yeast baited traps caught the most dipterans, consistent with prior research in citrus orchards (Thomas 2003a). In addition, it is recognized that Cera Trap® is as effective as other attractants but also represents a stable alternative for monitoring Anastrepha species for several weeks (Lasa and Williams 2022; Perea- Castellanos et al. 2015).
No differences in the capture of non-target arthropods and beneficial insects were noted among the seven treatments in the current study. However, the nature of the dependence between control treatments and the monitoring traps highlighted the relationship between mass trapping and monitoring traps baited with Cera Trap®. This combination caught fewer beneficial organisms than torula yeast. Also, mass trapping based on GF 120® combined with plastic bottles with Cera Trap® led to the same outcome. By contrast, bait sprays of malathion, a common control measure in Mexico (SAGARPA/SENASICA 2012), did not reveal an association with a monitoring trap, suggesting a homogeneity in the capture level in the monitoring traps. However, an effective suppression or control of tephritid pests may require the mix of several control methods at a local and regional scale or reinforcement in the case of a mass trapping strategy (Flores et al. 2017; Leza et al. 2008; Navarro-Llopis et al. 2015). Our results indicate that control treatments and monitoring fruit fly pests using Cera Trap® or GF-120® reduce the capture of non-target arthropods and beneficial organisms.
This study does not provide evidence of the effects of capturing non-target arthropods between control methods of A. ludens. However, it revealed a relationship between trapping and control methods in capturing non-target organisms. We found that the post-harvest citrus season in northeastern Mexico harbors an insect community composed of diverse taxonomic families that include beneficial organisms, which showed low capture (except ants). As such, a more detailed analysis is needed to evaluate the effects of control methods of A. ludens on predators and parasitoid groups.
Overall, capturing non-target arthropods highlights the attraction of torula yeast, although ant communities preferred Cera Trap®. Combining phytosanitary control measures and a particular trapping method may play a key role in minimizing the capture of non-target organisms. For conservation purposes and from an area-wide integrated pest management approach, developing integral strategies favoring improved attractants for A. ludens can contribute to environmentally friendly practices and greater efficiency in trapping activities.
Figure 2: A comparison of the capture of non-target arthropods (a) and beneficial organisms (b) between types of monitoring traps based on linear mixed models and LS-means (emmeans package) in citrus agro- ecosystems in Mexico. Means were contrasted through Tukey multiple comparisons (p < 0.05). Different letters indicate statistical significance between the capture. BC = plastic bottle with Cera Trap®; MC = multilure trap with Cera Trap®; and MT = multilure trap with torula.
Figure 3: Mosaic plots for evaluating the non-dependence between control treatments and type of monitoring traps regarding the capture proportions of non-target arthropods (a) and beneficial organisms (b) in citrus agroecosystems in Mexico. Significant cells are colored based on Pearson residuals (p < 0.05). Rectangular areas represent the propor- tional size to the corresponding observed frequencies. Treatments: CT25 = 25 PET traps with Cera Trap®; CT40 = 40 PET traps with Cera Trap®; MC25 = 25 PET traps with the mix malathion 50 % + Captor 300®; GF25 = 25 PET traps with GF-120®; BSGF = bait sprays with GF-120®; BSMC = bait sprays with malathion and captor; and CT = control treatment. Monitoring traps: BC = plastic bottle with Cera Trap®; MC = multilure trap with Cera Trap®; and MT = multilure trap with torula yeast borax.
Acknowledgments: The authors are grateful to the citrus growers of the association Marte R. Gómez and E. Romo, owner of the “El Maravasco” orchard, for the facilities granted to carry out this study. Also, the authors appreciate the support of J. A. Carrizales, L. A. Guardiola, N. Álvarez, and field technicians for their participation in the field, taxonomic, and laboratory work. Finally, thanks to the Bioiberica® company through Agrotecnal S.A. de C.V., which contributed to the materials used in this work.
Research ethics: Not applicable.
Informed consent: Not applicable.
Author contributions: The authors have accepted respon- sibility for the entire content of this manuscript and approved its submission.
Use of Large Language Models, AI and Machine Learning Tools: None declared.
Conflict of interest: The authors state no conflict of interest. Research funding: None declared.
Data availability: The raw data can be obtained on request from the corresponding author.
