Introduction
Augmentative releases of parasitoids as pest control agents are increasingly relevant in the context of current agriculture worldwide (Wang et al., Reference Wang, Liu, Shi, Huang and Chen2019; Murali-Baskaran et al., Reference Murali-Baskaran, Sharma, Sridhar, Jain and Kumar2021). For the use of this control alternative, studies on the functional response and mutual interference of foraging females revealing their behaviour and effectiveness as natural enemies are critical (Ghorbani et al., Reference Ghorbani, Seraj, Allahyari and Farrokhi2019; DeLong and Uiterwaal, Reference DeLong and Uiterwaal2022). The functional response of a predator/parasitoid describes the relationship between the number of prey/hosts it consumes per unit of time and the abundance or density of the prey or hosts (Solomon, Reference Solomon1949; Holling, Reference Holling1959). Studies on the foraging ecology of natural enemies contribute to estimate their attack capacity (Savino et al., Reference Savino, Coviella and Luna2012; Guleria et al., Reference Guleria, Sharma, Verma, Chandel and Sharma2020), and allow inferences about their searching capacity, handling time, and intrinsic attack rate in a given interval of time and space (Fernández-Arhex and Corley, Reference Fernández-Arhex and Corley2004).
Mutual interference between foraging conspecific parasitoid females results from the aggregation behaviour around the location of their hosts (Hirose et al., Reference Hirose, Nakamura, Takagi, Mackauer, Ehler and Roland1990). Periodic releases of parasitoids can generate a high density of foraging females in the same patch, which would increase intraspecific competition at low host densities (Salas, Reference Salas2017; Rezaei et al., Reference Rezaei, Talebi, Fathipour, Karimzadeh and Mehrabadi2019). Mutual interference between female parasitoid wasps reduces their searching and parasitic efficiency due to a higher probability of antagonistic encounters (Kumar et al., Reference Kumar, Kashyap and Soni2019). It can also cause an increase in the occurrence of superparasitism (i.e., more than one immature parasitoid of the same species per host) (Skovgard and Nachman, Reference Skovgard and Nachman2015; Poncio et al., Reference Poncio, Montoya, Cancino and Nava2016; López et al., Reference López, Rosales, Flores and Montoya2021), which can cause an imbalance in their offspring sex ratio and emergence rates (Li et al., Reference Li, Gong, Chen, Pan, Dai, Hu and Liu2022) because of competition for the same host resource (Skovgard and Nachman, Reference Skovgard and Nachman2015). Functional response and mutual interference studies allow to determine the host or prey density that results in the lowest percentage of antagonistic encounters between conspecifics of natural enemies, which would then result in a more effective control of the pest (Rezaei et al., Reference Rezaei, Talebi, Fathipour, Karimzadeh and Mehrabadi2019; Soni and Kumar, Reference Soni and Kumar2021). This information would not only provide an idea of the efficiency that a parasitoid species would be able to achieve but would also allow us to understand the behaviour, benefits, and/or consequences in predator-prey population dynamics (Skovgard and Nachman, Reference Skovgard and Nachman2015; Nakamichi et al., Reference Nakamichi, Tuda and Wajnberg2020).
The diapriid wasp Coptera haywardi (Oglobin) is a solitary idiobiont endoparasitoid of tephritid fruit fly pupae, native to and with a wide distribution in the Neotropical region, from Mexico to Argentina (Nuñez-Campero et al., Reference Núñez-Campero, Ovruski and Aluja2012). The females of this species can parasitise fruit fly pupae buried in soils that are difficult for other parasitoids to access (Baeza-Larios et al., Reference Baeza-Larios, Sivinski, Holler and Aluja2002; Guillén et al., Reference Guillén, Aluja, Equihua and Sivinski2002). In addition, this species has a high ability to discriminate against pupae previously parasitised by conspecific and heterospecific parasitoids such as Diachasmimorpha longicaudata (Ashmead) (Cancino et al., Reference Cancino, Liedo, Ruiz, López, Montoya, Barrera, Sivinski and Aluja2012; Reference Cancino, Pérez, Johnson and Reynolds2019). It has been documented that C. haywardi develops successfully in pupae of various species of the genus Anastrepha, such as A. serpentina, A. striata, and A. obliqua (García and Montilla, Reference García and Montilla2001), and that it can develop in wild and irradiated strains of Ceratitis capitata pupae (Wied) (Sivinski et al., Reference Sivinski, Vulinec, Menezes and Aluja1998; Núñez-Campero et al., Reference Núñez-Campero, Suárez, Buonocore-Biancheri, Cancino, Murúa, Molina, Laría, Aluja and Ovruski2020). In addition, it has been suggested that C. haywardi can be used together with other biological control agents against both Anastrepha and Ceratitis flies due to its great discriminatory capacity (Cancino et al., Reference Cancino, Montoya, Barrera, Aluja and Liedo2014; Van Nieuwenhove et al., Reference Van Nieuwenhove, Bezdjian, Schliserman, Aluja and Ovruski2016; Martínez-Barrera et al., Reference Martínez-Barrera, Toledo, Cancino, Liedo, Gomez, Valle-Mora and Montoya2021). However, the use of C. haywardi (or any other species of parasitoid) in augmentative biological control programmes must be carefully planned and analysed in advance to avoid failure and/or harmful effects of their interaction (Montoya et al., Reference Montoya, Gálvez and Díaz-Fleischer2019; Rezaei et al., Reference Rezaei, Talebi, Fathipour, Karimzadeh and Mehrabadi2019). Our purpose here was to characterise the functional response and mutual interference of C. haywardi females foraging on A. ludens pupae under laboratory conditions. These findings will reinforce existing knowledge on the biology and potential of C. haywardi as a natural enemy of tephritid fruit flies and may contribute to improving the rearing methods and field release strategies for the control of these important pests.
Materials and methods
Study area
The functional response experiment was carried out under laboratory conditions at El Colegio de la Frontera Sur, Tapachula Unit, Chiapas, Mexico, while the mutual interference experiment was carried out in the Biological Control laboratory of the Moscafrut Program (SADER-SENASICA) located in Metapa de Domínguez, Chiapas, Mexico. The experiments were conducted out according to the bioassay criteria established by Montoya et al. (Reference Montoya, Liedo, Benrey, Barrera, Cancino and Aluja2000, Reference Montoya, Gálvez and Díaz-Fleischer2019) and Poncio et al. (Reference Poncio, Montoya, Cancino and Nava2016).
Biological material
The Moscafrut facility, (SADER-SENASICA) located in Metapa, Chiapas, Mexico, provided 3–4-day old Anastrepha ludens pupae. The 7–8-day old C. haywardi females were provided by the Biological Control laboratory of the Fruit Fly Program located in the same locality. The laboratory strain was started in 1996 with specimens collected as described by Aluja et al. (Reference Aluja, Sivinski, Ovruski, Guillén, López, Cancino, Torres-Anaya, Gallegos-Chan and Ruíz2009). This colony has been maintained using A. ludens pupae produced in the Moscafrut facility, with an average of nine generations per year (approximately 234 generations under laboratory rearing).
Functional response experiment
Seven host densities (1, 5, 10, 20, 30, 40, and 50 pupae) were used as treatments, which were placed in a Petri dish bottom (60 × 15 mm) with 2 g of humid and sterile vermiculite to simulate soil. Each treatment was placed separately in 3.7 L plastic containers (25 cm high and 12 × 14 cm diameter) with a wide mouth. Petri dishes inside plastic containers were covered with a black cardboard roof to obtain an illuminance of ~20 lux (Cancino et al., Reference Cancino, Liedo, Ruiz, López, Montoya, Barrera, Sivinski and Aluja2012). Each pupal density was exposed to a 7-day-old mature C. haywardi female with previous oviposition experience and previously provided with water and food consisting of bee honey mixed with soft paper ad libitum (Montoya et al., Reference Montoya, Pérez-Lachaud and Liedo2012). The pupae were exposed to each female for 24 h with a 12:12 L: D photoperiod at 25 ± 1°C. After exposure, the pupae remained under the same environmental conditions for 3–4-days to later determine the number of pupae attacked, the number of oviposition scars per pupa, and the number of parasitoid larvae per pupa obtained by dissection using a stereomicroscope (Carl Zeiss®, Model Stemi 2000C). Sixteen replicates per host density were made with different females in each replicate.
Mutual interference experiment
The experiment was carried out using the same type of containers as in the previous experiment, which were conditioned with water and food ad libitum. The bottom of a Petri dish (60 × 15 mm) with 2 g of moist vermiculite to simulate soil was also placed in the containers. Ten unparasitised 3–4-day old A. ludens pupae were placed in each Petri dish, which were exposed to six different parasitoid densities: 1, 2, 4, 8, 12, and 16 mated C. haywardi females with previous oviposition experience, which were free in the above described container. The Petri dishes with pupae were covered with a black cardboard cutout to obtain a lighting level of ~20 lux to stimulate the searching and oviposition behaviour of the females. The exposure time of the pupae was 24 h with a photoperiod of 12:12 L: D at 25 ± 1 °C. Each treatment was repeated ten times in duplicate to obtain two groups of host pupae (groups 1 and 2) for further analysis. All host pupae utilised were from the same production batch.
Four days after exposure, the pupae of the first group were observed to determine the number of pupae attacked, the number of oviposition scars per pupa, and the level of superparasitism by determining the number of parasitoid larvae per dissected pupa using a stereoscopic microscope (Carl Zeiss®, Model Stemi 2000C). The pupae of the second group were used to wait for the emergence of adults (~ 34–35 days) and then determine the survival and sex ratio in each replicate and treatment. Interference behaviour among foraging C. haywardi females was randomly filmed between treatments using a Motorola smartphone (Mod.G60). In each recording, the black cardboard roof was removed for 30 min.
Statistical analysis
Functional response data were analysed following the models and parameters proposed by Trexler et al. (Reference Trexler, McCulloch and Travis1988) and Juliano (Reference Juliano, Scheiner and Gurevitch1993). Mutual interference data were analysed according to the model proposed by Skovgard and Nachman (Reference Skovgard and Nachman2015). The data from both experiments were analysed using ‘R Commander’ (Version 4.2.0) (Fox, Reference Fox2005, Reference Fox2017; Fox and Bouchet-Valat, Reference Fox and Bouchet-Valat2019).
Functional response
The functional response analysis was performed by fitting a logistic regression model to the proportion of pupae attacked in relation to the density of pupae offered. The values of the linear P1 and quadratic P2 parameters were used to differentiate between type II and type III functional responses (Holling, Reference Holling1959). If the linear parameter P1 was negative, it would indicate a type II functional response, whereas if the P1 parameter was positive and the square coefficient was negative (P2), then it would indicate a type III functional response (Poncio et al., Reference Poncio, Montoya, Cancino and Nava2016). Attack rate (a’) and handling time (t) were estimated by fitting the number of hosts attacked (Na) against the number of hosts offered (No) to a nonlinear regression of least squares. The expected number of pupae attacked was calculated using Holling's (Reference Holling1959) ‘disc equation’:
where Na is the expected number of hosts attacked, a' is the constant attack rate, Tt is the exposure time of the host, No is the number of hosts available, and Th is the handling time.
Mutual interference. The mutual interference analysis was performed by modelling the functional response of a single female parasitoid according to the following formula:
where Na is the number of hosts attacked at least once during the exposure time t, No is the number of exposed hosts, E is the rate of encounters between a parasitoid and the hosts that caused attacks, and P is the number of parasitoids. E was calculated based on parasitoid density as E = Emax f (P); if all the experimental factors are optimal, Emax will represent the maximum attack rate achieved. It is assumed that the effect of parasitoid density (f (P)) decreases monotonically with the number of parasitoids, so that f (P) = 1 when P = 1. The model used by Skovgard and Nachman (Reference Skovgard and Nachman2015) has two parameters (r and q) used in the description of the relationship between parasitoid density and parasitoid attack rate, as follows:
Parameter r is the result obtained from the rate of encounters with conspecifics and the wasted time per encounter (P−1), A is the size of the area where the hosts and parasitoids are confined. The r values reflect the effect of parasitoid density, while the q values denote the degree of interference as the parasitoid density increases. The effect of mutual interference is evidenced by the difference between the number of attacked pupae observed and the number of attacked pupae expected by the model.
The total number of pupae attacked, the number of pupae attacked per female, the number of oviposition scars per pupa, and the number of immature parasitoids per pupa were analysed with a one-way ANOVA followed by a Tukey's multiple range test. The values of the last three parameters were log-transformed [x + 1] for the analysis. Adult emergence percentage and sex ratio were arcsine square-root-transformed prior to the ANOVA. The relationship between the number of oviposition scars per pupa and the number of parasitoid larvae was analysed using the Pearson correlation method. The correlation between the number of oviposition scars per pupa and the percentage of adult emergence was evaluated by means of a simple linear regression. In addition, a logistic regression was used to analyse the relationship between the number of oviposition scars and sex ratio.
Results
Functional response experiment
The values of the linear and quadratic coefficients of the logistic regression (P1 = −2.4736, P < 0.00379; P2 = −0.61860, P > 0.3456) suggest that the proportion of pupae attacked decreases as host density increases, and thus female C. haywardi exhibit a type II functional response. The attack rate (a') and handling time (Th) of C. haywardi were 0.0134 host pupa/h and 1.8429 h, respectively. The functional response and the parasitism average curves are shown in fig. 1, where it is possible to observe that the highest of attacked pupae and the highest number of parasitised pupae occurred in the treatment of (1:50) female parasitoid: host pupae. The highest average of parasitised pupae was observed at 10 host pupae per female.
Mutual interference experiment
The highest attack rate (Emax) was 0.153 ± 0.055 h−1. The r value was 2.265 ± 0.83 P-1, which differs from 0 (t = 2.699, P = 0.00914, df = 57), while the q value of 0.356 ± 0.229 was significantly different from 1 (t = 2.812, P = 0.00673, df = 57), which indicates the existence of a mutual interference effect dependent on the density of C. haywardi females. The expected average values of pupae attacked without mutual interference and the average of pupae attacked with the presence of mutual interference are shown in fig. 2. The difference between the curves reflects the magnitude of the mutual interference.
The number of parasitised pupae per female was negatively affected by parasitoid density (F = 4.52, df = 5, 54, P < 0.0016), where treatments with 12 and 16 foraging females showed the lowest values. The number of oviposition scars per pupa (F = 4.652, df = 5, 54, P < 0.0013) and the number of parasitoid larvae per pupa (F = 4.059, df = 5, 54, P < 0.0033) increased as the number of foraging females increased (Table 1). A positive Pearson correlation was observed between the number of oviposition scars per pupa and the number of immature parasitoids (F = 8.414, df = 1, 49, P = 0.00556) (fig. 3). There was no significant difference in the adult emergence percentage between treatments (F = 0.6764, df = 5, 54 P > 0.6432) or in the sex ratio of adults emerging from the different treatments (F = 1.4682, df = 5, 54 P > 0.2155) (Table 1).
Values followed by the same letter in each column indicate that differences are not significant (one-way Anova, Tukey's test, P > 0.05).
Discussion
Several authors have stated (e.g., Merkel, Reference Merkel2014; Sereno et al., Reference Sereno, Salvo and Battan-Horenstein2016; Francati, Reference Francati2018; Stucchi et al., Reference Stucchi, Giménez-Benavides and Galeano2019) that functional response and mutual interference are important attributes of natural enemies because they provide valuable information about parasitoid-host interactions, which can be used to infer the impact they may have on the regulation of target pests, and even assist to improve their laboratory rearing (Poncio et al., Reference Poncio, Montoya, Cancino and Nava2016; López et al., Reference López, Rosales, Flores and Montoya2021).
Our results show that C. haywardi exhibits a type II functional response, where the number of pupae attacked increased as host density increased. A type II functional response has also been determined in other pupal parasitoids such as Dirhinus giffardii (Hymenoptera: Chalcididae) attacking Bactrocera zonata and B. cucurbitae (Khan et al., Reference Khan, Khuhro, Awais, Memon and Asif2020), as well as in the parasitoid Spalangia endius (Hymenoptera: Pteromalidae) attacking Bactrocera dorsalis and Musca domestica pupae (Zheng et al., Reference Zheng, Song, Zhang and Li2021). However, unlike D. giffardii and S. endius, our results with C. haywardi show a longer handling time that is reflected in a lower attack rate. This low attack rate is a consequence of a long searching time that results in a slow increase in attacked pupae as the number of offered pupae increases (Fernández-Arhex and Corley, Reference Fernández-Arhex and Corley2004).
According to Martínez-Barrera et al. (Reference Martínez-Barrera, Toledo, Cancino, Liedo, Gomez, Valle-Mora and Montoya2021), C. haywardi females have an average oviposition time of 24.6 min and an average refractory period of 10.6 min. These results agree with our observations, since we recorded times fluctuating between 25 and 45 min per oviposition. This amount of time invested by females in oviposition is considered a characteristic of the species (Cancino et al., Reference Cancino, Liedo, Ruiz, López, Montoya, Barrera, Sivinski and Aluja2012), which is much longer compared to other pupal parasitoids (e.g., Dirhinus giffardii, Khan et al., Reference Khan, Khuhro, Awais, Memon and Asif2020). Cancino et al. (Reference Cancino, Liedo, Ruiz, López, Montoya, Barrera, Sivinski and Aluja2012) mention that C. haywardi females carefully examine their hosts before laying eggs in them, which is also consistent with our observations.
According to Holling (Reference Holling1959), a type II functional response is representative of invertebrates including parasitoids, whereas a type III functional response is more representative of vertebrate predators, where learning and switching from one prey to another is more frequent. However, type III functional responses have also been determined in some braconid species (e.g., Montoya et al., Reference Montoya, Liedo, Benrey, Barrera, Cancino and Aluja2000, Yazdani and Keller, Reference Yazdani and Keller2016), where density-dependent mortality at low prey densities has been observed (Fernández-Arhex and Corley, Reference Fernández-Arhex and Corley2004). In our results with C. haywardi, we also observed density-dependent mortality at low host densities (fig. 1), but unlike the results with the braconid D. longicaudata, the fit of the quadratic term was not significant in our results, and thus the prediction of our model indicated a type II functional response. In addition, due to the long foraging time and low attack rate by C. haywardi, we consider that a type II functional response might become more evident under field conditions where parasitoids are free to choose the foraging patch. De Pedro et al. (Reference De Pedro, Beitia, Ferrara, Asis, Sabater-Munoz and Tormos2016) observed that Aganaspis daci (Hymenoptera: Figitidae) exhibited a type III functional response when Ceratitis capitata larvae were supplied with fruit and an artificial diet under laboratory conditions, but exhibited a type II functional response when the flies were under greenhouse conditions (more similar to natural conditions). Changes in the functional response type have also been shown in D. longicaudata allowed to freely choose among different host density patches under laboratory conditions (Núñez-Campero et al., Reference Núñez-Campero, Benitez-Vieyra, Gorla and Ovruski2016).
Suitable host pupae available for C. haywardi females under natural conditions may not be abundant, since both the larval and pupal stages of fruit flies are attacked by multiple competitors (Aluja et al., Reference Aluja, Sivinski, Rull and Hodgson2005), and most host pupae in the soil may have been previously parasitised. We believe that this condition may lead to the type of foraging and attack rate shown by C. haywardi in our study, since the distribution of suitable pupae in the ground would tend to be random, reducing thus the possibilities of interference between foraging females. However, it must be considered that the type of soil and its characteristics, together with the abiotic conditions, are factors that can also influence the search time and parasitism success (Aluja et al., Reference Aluja, Sivinski, Rull and Hodgson2005; Okuyama, Reference Okuyama2012). Guillen et al. (Reference Guillén, Aluja, Equihua and Sivinski2002) found that C. haywardi females performed better in clayey soils, where foraging and host parasitism were not affected by moisture or clay compaction. This may be related to the characteristics of C. haywardi, such as its hypognathous head and great physical resistance (Sivinski et al., Reference Sivinski, Vulinec, Menezes and Aluja1998).
The results of the mutual interference experiment show that the attack efficiency of C. haywardi females is negatively affected by an increase in the density of foraging females, which became evident in the reduction in the number of pupae attacked by female. The effect of mutual interference between foraging females under laboratory conditions may be caused by high parasitoid densities or low host densities (Yang et al., Reference Yang, Campos-Figueroa, Silva and Henne2015; Rezaei et al., Reference Rezaei, Talebi, Fathipour, Karimzadeh and Mehrabadi2019), as well as by the small dimensions of the experimental arena (Ridout, Reference Ridout1981). These conditions reduce the searching capacity of the parasitoid (indirect interference) and may force it to migrate to new foraging areas under natural conditions (Visser and Driessen, Reference Visser and Driessen1991). However, in some cases, parasitoid females can physically fight (direct interference) for control of the host resource (Mi et al., Reference Mi, Zhang, Haye, Zhang, Zhao, Lei, Li and Zhang2021). We did not detect such direct interference in our observations, and thus our results may rather be due to a low tolerance to the presence of other foraging females in the experimental arena.
Our data also show that the number of attacked pupae, number of oviposition scars, and number of immature parasitoids per pupa increased as the number of foraging females increased. This behaviour of superparasitism in experiments of mutual interference under laboratory conditions has also been observed in other parasitoids of dipteran larvae and pupae (Skovgard and Nachman, Reference Skovgard and Nachman2015; Poncio et al., Reference Poncio, Montoya, Cancino and Nava2016; López et al., Reference López, Rosales, Flores and Montoya2021). These authors have found that, at high parasitoid densities in the presence of a low number of hosts, superparasitism increased exponentially, which is consistent with the positive correlation between the number of oviposition scars and the number of immature parasitoids per pupa obtained in our study.
Cancino et al. (Reference Cancino, Liedo, Ruiz, López, Montoya, Barrera, Sivinski and Aluja2012) observed that C. haywardi females are capable of discriminating A. ludens pupae previously attacked by conspecific and heterospecific parasitoid species under laboratory conditions. We believe that the observed superparasitism may be caused by the competitive pressure among foraging females, the low host densities, and the ovarian loads of C. haywardi females (as in Montoya et al., Reference Montoya, Gálvez and Díaz-Fleischer2019), but it has been shown that C. haywardi females also superparasitised A. fraterculus pupae in field cages with a 5:1 host-parasitoid ratio (Van Nieuwenhove et al., Reference Van Nieuwenhove, Bezdjian, Schliserman, Aluja and Ovruski2016). It has been reported that superparasitism has a negative effect on the emergence of adults (Tunca and Kilincer, Reference Tunca and Kilincer2009), which may be related to competition between parasitoid larvae for the host resource (Luo et al., Reference Luo, Li, Lu, Zhang, Haye, Kuhlmann and Wu2014). However, in the cases of D. longicaudata and U. anastrephae, moderate superparasitism has been associated with a favourable, biased female sex ratio (González et al., Reference González, Montoya, Pérez-Lachaud, Cancino and Liedo2007; Poncio et al., Reference Poncio, Montoya, Cancino and Nava2016). Unlike the above cases, the emergence and sex ratio of C. haywardi were not significantly affected by superparasitism. Interestingly, sex ratio was male biased when C. haywardi attacked host previously parasitised by D. longicaudata (Van Nieuwenhove et al., Reference Van Nieuwenhove, Bezdjian, Schliserman, Aluja and Ovruski2016) while in the present study sex ratio was in general female biased. It is unknown if adults of this species emerging from superparasitised hosts present limitations in parameters such as flight ability, longevity, or fecundity. However, Montoya et al. (Reference Montoya, Gálvez and Díaz-Fleischer2019) reported that C. haywardi adults emerging as hyperparasitoids from A. ludens pupae that had been previously parasitised by the braconid D. longicaudata do no differ in size, fecundity, or longevity when compared to those that emerge as primary parasitoids.
Our results show that C. haywardi exhibits a type II functional response, and that mutual interference between foraging females decreases the attack efficiency of individual females even at low densities. They also suggest that this species could act as a complementary parasitoid against fruit fly pests, since the random distribution of fruit fly pupae in the soil would facilitate their location while reducing the possibility of aggregation and mutual interference between foraging females. This would render a better performance of C. haywardi as biocontrol agent.
Acknowledgments
The authors thank Amanda Ayala (SADER-SENASICA Moscafrut Program) for her support during the experiments. We also thank Azucena Oropeza Cabrera (El Colegio de la Frontera Sur, Tapachula) for her valuable help in obtaining the necessary materials for the experiments. The study was supported by the National Council of Science and Technology (CONACYT) with a PhD scholarship for GC, and by the Moscafrut Program (SADER-SENASICA) with the donation of biological material for the experiments.