The objective of the present study was to assess the effect of different times of exposure to the host on parasitoid emergence rate, parasitoid progeny sex ratio, and on host mortality, as a step toward the development of an efficient mass-rearing system for the braconid Diachasmimorpha longicaudata (Ashmead) on larvae of the VIENNA 8 Temperature Sensitive Lethal Ceratitis capitata (Wiedemann) strain. The role of host-rearing substrate cues in stimulating the host-searching behavior of parasitoid females was also evaluated. Three exposure times (40, 60, and 120 min) were tested. One hundred 7 d-old host larvae were exposed to 25 female parasitoids per treatment. Larvae mixed with wheat-based rearing medium and larvae without medium were used in each test. A second set of treatments with the same method described above was conducted using late third-instars of the C. capitata wild-type strain. These experiments were carried out to assess the quality of the larvae of VIENNA 8 strain in producing D. longicaudata adults by comparing them with the larvae of the wild-type strain. Results indicated that the use of larvae of VIENNA 8 strain on their rearing diet at 40 min exposure time significantly increased overall parasitoid offspring production and decreased the host mortality level. Nevertheless, parasitoid emergence recorded from VIENNA 8 strain was notably lower than that recorded from the wild-type strain. Low parasitoid emergence levels and the prevalence of male-biased progeny recorded in all assays are obstacles to development of a parasitoid mass-rearing system using larvae of VIENNA 8 strain as host. Additional studies focusing on host exposure to parasitoids are needed to verify the effect of host larval quality on the production of D. longicaudata.
The Mediterranean fruit fly (= medfly), Ceratitis capitata (Wiedemann) (Diptera: Tephritidae), is one of the major pests of fruit crops in Argentina and its presence in many fruit-growing regions is a phytosanitary barrier to the export of fresh fruits (Guillén & Sánchez 2007). In the Province of San Juan, located in the central-eastern region of Argentina known as Cuyo, control strategies against C. capitata in temperate pome and stone fruit-producing areas have been implemented by the National Fruit Fly Control and Eradication Program (ProCEM) jointly with the provincial government and the producers. This program aims to establish C. capitata-free and/ or low prevalence areas based on the concept of area-wide integrated pest management (Guillén & Sánchez 2007). In 1986 the San Juan provincial government constructed the BioPlanta San Juan mass-rearing facility, in order to produce sterile flies to suppress or eradicate C. capitata in both urban and rural areas of San Juan Province, Argentina. Therefore, the strategies applied against C. capitata through the ProCEM in San Juan, have been based on an integrated use of SIT (sterile insect technique), cultural and chemical controls, jointly with the implementation of a strict quarantine system (Díaz et al. 2008). Furthermore, in 2008 biological control has been incorporated into control activities of ProCEM San Juan by establishing, in a first phase, a laboratory colony of the exotic parasitoid Diachasmimorpha longicaudata (Ashmead) (Hymenoptera: Braconidae) at BioPlanta San Juan facility.
Diachasmimorpha longicaudata is a larvalprepupal, koinobiont endoparasitoid of several tephritid species, native to Southeast Asia (Wharton 1989) that has been introduced throughout much of Latin American and in the southern United States (Ovruski et al. 2000). It is considered one of the most important biological control agents for augmentative releases against economically important Anastrepha (Diptera: Tephritidae) species and C. capitata in tropical and subtropical America (Paranhos et al. 2008; Montoya et al. 2011). In 1999, D. longicaudata was introduced into Argentina via Mexico with the intent of renewing fruit fly biological control (Ovruski et al. 2003). Therefore, laboratory colonies of D. longicaudata were established in Argentina on both wild-type MI94 and CAST191 genetic sexing strains of C. capitata (Viscarret et al. 2006) and on both C. capitata and Anastrepha fraterculus (Wiedemann), both fly strains originated from wild guava fruits (Ovruski et al. 2011).
The suitability of the VIENNA 8 temperature sensitive lethal (TSL) C. capitata strain (Franz 2005) as a host for rearing D. longicauda is currently being studied at the BioPlanta San Juan facility. This research is part of a new program of ProCEM-San Juan that includes the use of environmentally friendly strategies to suppress or eradicate C. capitata.
Parasitoid emergence and host mortality after exposure are 2 key parameters in the quality control system of the D. longicaudata mass-rearing process that must be continuously evaluated (Montoya et al. 2007). Therefore, the present study examined the effect of parasitoid exposure time durations to the host on the parasitoid emergence rate, the parasitoid progeny sex ratio, on the host mortality level, and on the role of host-rearing substrate cues in stimulating host-searching behavior in parasitoid females. The ultimate goal of this study was to identify the factors needed to maximize yields of D. longicaudata. This study serves as a first step toward development of an efficient mass-rearing technique for D. longicaudata on larvae of VIENNA 8 strain at BioPlanta San Juan facility.
MATERIAL AND METHODS
Study Site and Insects Rearing
The study was performed in the Parasitoid Rearing Laboratory at BioPlanta San Juan facility located in Marquesado, San Juan Province, Argentina. Diachasmimorpha longicaudata adults were kept in rectangular iron-framed, mesh-covered cages (0.5 × 0.5 × 0.6 m) at a capacity of 2,000 pairs per cage at 24 ± 1 °C; 65 ± 5% RH and 12:12 h L:D. Each parasitoid rearing cage was provided with water and honey on a daily basis. The colony of D. longicaudata was initiated originally in 2008 with 20,000 individuals from the Biological Control Division of the Pilot Plant of Industrial Microbiology Processes and Biotechnology (PROIMI), where D. longicaudata had been reared in the laboratory on A. fraterculus larvae (Ovruski et al. 2011). Two strains of D. longicaudata were used in the experiments. One of them was reared and maintained on third-instar larvae of VIENNA 8 strain, whereas the other one was reared on C. capitata larvae of a wild-type strain that had originated from peach (Prunus persica (L.) Batsch; Rosales: Rosaceae) fruit collected in rural areas. The larvae of both C. capitata strains were reared on a wheat-based diet fortified with yeast, poplar (Populus sp.; Malpighiales: Saliaceae) shavings, sugar, hydrochloric acid, sodium benzoate, nipagin and water.
Exposure times of 40, 60, and 120 min were tested independently to choose between these 3 alternatives a suitable time duration to expose larvae of VIENNA 8 strain to D. longicaudata for achieving the greatest parasitoid yield with the highest proportion of female. Late third-instars of VIENNA 8 strain (7 d-old) were used for exposure to the parasitoid. To keep larval weight constant, only male larvae coming from the first batch that dropped out of the larvae rearing tray were collected for assays. In addition, these larvae were not irradiated. Tests were carried out by first placing a group of host larvae into an oviposition device without artificial rearing medium (referred to as “naked” host larvae in the text). This oviposition unit was positioned on the top screen of an experimental cage (0.5 x 0.5 x 0.6 m). Concurrently, another group of host larvae provided with 50 cm3 wheat-based medium (referred to as “covered” host larvae in the text) was placed into an oviposition device on the top screen of another experimental cage. Each oviposition unit was composed of an organdy screen-covered dish (10 cm diameter and 1 cm deep). One hundred of either naked or covered host larvae per oviposition unit were exposed to 25 female parasitoids in each test. This host/parasitoid ratio was based on previous assays for optimizingD. longicaudata mass rearing on A. ludens at the Moscafrut facility (Metapa de Dominguez, Chiapas, Mexico) (Montoya et al. 2000). Parasitoid females used in the tests were 5 d-old, mated, and had no prior oviposition experience as recommended by Montoya et al. (2000). Parasitoids were released on the floor in the central part of the test cage. The cages were positioned at random on a table and kept under the above-mentioned laboratory conditions. After exposure to the parasitoids, oviposition units were removed from the cages. Host larvae were then placed in emergence cups (0.7 m diam, 0.8 m deep) with meshscreen covers and containing wheat bran as the pupation medium on the bottom. The cups were kept under the above mentioned laboratory conditions until the emergence of adult parasitoids.
A second set of treatments following the same method above described was conducted using late third-instars of the C. capitata wild-type strain. These experiments were carried out to assess the quality of the larvae of VIENNA 8 strain in producing D. longicaudata adults by comparing with the larvae of wild-type strain.
The number of emerged male and female parasitoids, and the number of noneclosed puparia was recorded daily. Each treatment was replicated 14 times with fresh parasitoids and host larvae. Before starting assays in each replicate, the quality of batches of both strains of host larvae used in the assays was evaluated. For this, a subsample of 10 larvae of both C. capitata strains was taken out from each sowing tray and weighed for host quality evaluation. Also, unexposed oviposition units containing either naked or covered larvae of both medfly strains were kept individually to determine the percentages of host mortality. Moreover, all host larvae batches with adult fly emergence percentages below 80% were discarded and were not considered for this study.
Upon release of parasitoids into each test cage, the number of females that landed on the oviposition unit (= females/visit) and also the number of females that inserted their ovipositor into the oviposition unit (= ovipositor insertions) were recorded (García-Medel et al. 2007). The female parasitoids were observed once every 10 min during the first 30 min, each observation lasting 30 sec. Both behavioural observations were used as suitable variables to measure parasitoid preference between host larvae mixed with rearing diet and naked host larvae.
Percent host mortality was calculated as the fraction of dead host larvae from the total number of larvae exposed to parasitoids plus the fraction of non-eclosed puparia from the total number of recovered puparia. The percentage of emerged parasitoids was determined as the ratio of total number of parasitoid adults divided by the total number of recovered puparia. Parasitoid progeny sex ratio was calculated as the percentage of female offspring from the total number of emerged parasitoids. Data on host mortality, parasitoid emergence, sex ratio of parasitoid offspring, and number of female visits to and ovipositor insertions on the oviposition devices containing either naked or covered host larvae, were compared statistically between treatments by three-way multivariate analysis of variance (MANOVAs) followed by univariate ANOVAs for each response variable at P = 0.05. The use of MANOVA followed by ANOVAs, as described above, reduces the probability of inflating the Type I error rate (Zar 1999).
Moreover, host mortalities of each C. capitata strain were statistically compared between controls (host larvae not exposed to parasitoids) and between controls and treatments (host larvae exposed to parasitoids) by means of three-way ANOVAs. The exposure times, medfly strains, and the modes of exposing hosts to parasitoids were treated as categorical factors in MANOVAs, and also in the ANOVA performed to compare controls. Regarding ANOVAs used to contrast host mortality between controls and treatments, the host exposure times and modes, and the host larval status (host larvae exposed to, or not exposed to parasitoids) were the 3 categorical factors. Means were separated with a Fisher's Least Significant Difference (LSD) multiple comparison test at P = 0.05. To meet parametric analysis assumptions, arcsine square root and loge (x+1) transformations were applied to the percentage data and to the number of visits and ovipositor insertions, respectively (Zolman 1993). However, the untransformed data are shown as means (± SE) in tables and figures. Pupal weight differences between the 2 C. capitata strains were analyzed by a twosample t-test (P= 0.05).
The three-way MANOVA revealed significant effects of the 3 categorical factors and their interaction on D. longicaudata performance (Table 1). Corresponding ANOVAs revealed that both host exposure times and medfly strains significantly influenced both parasitoid emergence and host mortality, whereas only the mode in which hosts were exposed to parasitoids (i.e., naked or covered) notably impacted parasitoid emergence (Table 2). The interactions between host exposure times and medfly strains and also between these 2 categorical factors and the host exposure modes were highly significant only for the parasitoid emergence (Table 2). Significantly greater parasitoid emergence (Fig. 1) and lower host mortality (Fig. 2) percentages were recorded from both naked and covered larvae of the 2 medfly strains exposed to parasitoid females at 40 min than at longer exposure times. Furthermore, the percentages of emerged adult parasitoids (Fig. 1) and host mortality (Fig. 2) recorded from both naked and covered wild-type strain larvae were noticeably greater and significantly lower, respectively, than recorded from both naked and covered VIENNA 8 strain larvae exposed to D. longicaudata at 40, 60 or 120 min. In addition, parasitoid emergence was significantly higher in covered VIENNA 8 strain larvae at 40 min and at 120 min exposure times, as well as in covered wild-type strain larvae exposed to parasitoids at 60 min (Fig. 1).
The sex ratio of parasitoid offspring did not differ significantly among exposure times, or between the 2 modes of exposing host larvae to parasitoids, or between the 2 medfly strains (Table 1). The mean proportion of parasitoid female offspring recorded from VIENNA 8 strain and from wild-type strain ranged between 32–68% and 47– 52%, respectively.
The percent mortalities recorded from host larvae not exposed to parasitoid females (controls) did not differ significantly between the 2 medfly strains, or among exposure times of host larvae or between the 2 modes of exposing host larvae to parasitoids (Table 3). In addition, the interactions among these categorical factors were not significant (Table 3). In contrast, host mortalities recorded from both exposed and non-exposed larvae of the 2 medfly strains differed significantly between controls and treatments (host larvae exposed to parasitoids) and among host exposure times (Table 3). Host mortalities recorded in the controls of both VIENNA8 and wild-type strains were respectively 3 and 2.3 times lower than for the treatments.
The host larva exposure mode was the only categorical factor that significantly affected the number of visits as well as the number of ovipositor insertions (Table 4). The mean number of D. longicaudata female's visits to and ovipositor insertions on the oviposition units containing covered host larvae were 2–3 and 3–4 times significantly higher, respectively, than those recorded in the devices with naked host larvae (visits, F (1, 156) = 132.088, P < 0.0001; ovipositor insertions, F (1, 156) = 202.606, P < 0.0001) (Table 5).
Wild-type strain larvae were significantly larger than VIENNA 8 strain larvae (t = 56.501, df = 839, P < 0.0001, N = 840). Wild-type strain larvae weighed 12.3 ± 0.1 g, while VIENNA 8 strain larvae weighed 10.8 ± 0.1 g.
SUMMARY OF THREE-WAY MANOVA ON THE EFFECT OF HOST EXPOSURE TIME AND MODE, MEDFLY STRAINS, AND INTERACTIONS OF THESE CATEGORICAL FACTORS ON DIACHASMIMORPHA LONGICAUDATA EMERGENCE (EMERGED ADULT PARASITOIDS), PARASITOID SEX RATIO (PROPORTION OF FEMALE OFFSPRING), AND HOST MORTALITY (DEAD HOST LARVAE PLUS NON-ECLOSED HOST PUPARIA).
SUMMARY OF UNIVARIATE THREE-WAY ANOVA ON THE EFFECT OF HOST EXPOSURE TIME AND MODE, MEDFLY STRAIN, AND INTERACTIONS OF THESE CATEGORICAL FACTORS ON PARASITOD EMERGENCE, PARASITOID PROGENY SEX RATIO, AND HOST MORTALITY.
Results from this study indicated that the use of 7 d-old larvae of VIENNA 8 strain with their rearing diet at 40 min exposure time to parasitoid females significantly increased the overall parasitoid offspring production. Nevertheless, the results also revealed that the maximum yield of parasitoids achieved with the 40 min host exposure period was only about 9%. With regard to the latter, it should be noted that the parasitoid emergence rate in the current study were slightly lower than the 13.4% reported for D. longicaudata by Lopes & Paranhos (2006), who also worked with third instars of VIENNA 8 strain collected from the first batch at the EMBRAPA Semi-Arid Laboratory, Brazil. Different experimental procedures or differential adaptability to laboratory conditions or distinct host larva quality may be responsible for this difference.
SUMMARY OF THREE-WAY ANOVAS ON THE EFFECT OF HOST EXPOSURE TIME AND MODE, MEDFLY STRAIN CONTROLS (HOST LARVAE NOT EXPOSED TO PARASITOIDS) OR HOST LARVA STATUS (HOST LARVAE OF THE SAME MEDFLY STRAIN EXPOSED OR NOT EXPOSED TO PARASITOIDS), AND INTERACTIONS OF THESE CATEGORICAL FACTORS ON HOST MORTALITY.
Results from the current study also showed that parasitoid emergence recorded from VIENNA 8 strain was notably lower than that recorded from wild-type strain. This difference in parasitoid emergence rates was previously mentioned by Lopes & Paranhos (2006). Thus, it is possible to consider that the difference between the 2 medfly strains may be the result of distinct internal physiological conditions of host larvae for parasitoid larvae survival or disparities in the quality between larvae of genetic sexing and wild strains. These presumptions may be supported by host mortality data recorded in the present study. Host mortality percentages recorded from both naked and covered larvae of VIENNA 8 strain exposed to D. longicaudata females were significantly 1.4–2.0 times greater than those obtained from wild-type strain larvae exposed to parasitoids. In contrast, host mortality percentages recorded from non-exposure control tests in both medfly strains were substantially similar. Nevertheless, more detailed experimental studies are needed to verify this conclusion.
Data on host mortality showed that the increased exposure time of D. longicaudata to the host increased the level of non-eclosed puparia, regardless of the mode in which the host larvae were exposed (i.e., naked or covered). Therefore, the difference between parasitoid emergence and host mortality found in the treatments may be attributed to the action of D. longicaudata females during their time of exposure to the host larvae. The high larval mortality levels detected in both C. capitata strains at longer exposure times (1and 2-h) might be caused either by a detrimental effect of the puncture produced when ovipositing parasitoid females insert their ovipositors more than once in an unsuitable host larva, or perhaps as a consequence of superparasitism, as previously pointed out by González et al. (2007, 2010) and Van Nieuwenhove et al. (2012) to D. longicaudata reared on A. ludens and A. fraterculus, respectively.
SUMMARY OF THE THREE-WAY MANOVA ON THE EFFECT OF HOST EXPOSURE TIME AND MODE, MEDFLY STRAINS, AND INTERACTIONS OF THESE CATEGORICAL FACTORS ON THE NUMBER OF PARASITOID FRMALE VISITS TO AND OVIPOSITOR INSERTIONS ON THE OVIPOSITION DEVICES CONTAINING EITHER NAKED OR COVERED LARVAE OF EITHER THE VIENNA 8 OR WILD-TYPE CERATITIS CAPITATA STRAIN.
The offspring sex ratio, which was preferentially biased toward males in D. longicaudata found in almost all tests in the present study involving both C. capitata strains, may indicate differential survival of the sexes during parasitoid post-embryonic development (Paranhos et al. 2008) and/or may also reflect low host larva quality that would encourage parasitoid females to lay a reduced amount of fertilized eggs (Montoya et al. 2011). Previous studies conducted by Ovruski et al. (2011) found that D. longicaudata reared from A. fraterculus larvae, which are twice as large as C. capitata larvae, had a considerably higher proportion of female offspring than parasitoids that had developed from larvae of a wildtype C. capitata strain. Thus, physical characteristics, such as size, and/or chemical cues derived from host larvae could influence offspring sex ratio in D. longicaudata (Wong 1993; Messing et al. 1993; Eben et al. 2000; López et al. 2009).
Results of behavioral observations provided evidence to explain the high percentages of parasitoid emergence recorded from covered larvae of both medfly strains. Diachasmimorpha longicaudata females were more strongly attracted to oviposition units containing host larvae plus rearing medium than to units with only host larvae, albeit the fact that the devices were not offered in a choice situation. These findings are in agreement with earlier observations by Duan & Messing (2000) on ovipositor-probing behaviour in D. longicaudata reared on C. capitata. It is well known that D. longicaudata females exhibit oviposition behaviors in response to chemical cues derived from rotting fruits (Messing & Jang 1992; Eben et al. 2000; Carrasco et al. 2005) or from fermentation of the artificial rearing medium (Duan & Messing 2000) and also from the host larva themselves (Stuhl et al. 2011).
MEAN (±SE) NUMBER OF DIACHASMIMORPHA LONGICAUDATA FEMALE VISITS TO, AND OVIPOSITOR INSERTIONS INTO THE OVIPOSITION DEVICES CONTAINING EITHER NAKED OR COVERED LARVAE OF EITHER VIENNA 8 OR WILD-TYPE CERATITIS CAPITATA STRAINS.
In conclusion, low parasitoid emergence levels and the prevalence of male-biased progeny found in the present study are potential obstacles to development of a mass-rearing system using larvae of VIENNA 8 strain as host. Certain qualities of the host larva, such as size, age, instar, and species (Montoya et al. 2007; López et al. 2009; Cancino et al. 2009) are among the factors that influence parasitoid emergence, offspring sex ratio, and host mortality, among other quality control parameters in the mass-rearing process of D. longicaudata. Therefore, as pointed out by Cancino & Montoya (2004), weight, volume, and age of host larva must be carefully and continuously monitored in the quality control system to achieve more efficient mass rearing of D. longicaudata. Following up on the suggestions proposed by Cancino & Montoya (2004) and Montoya et al. (2007), additional studies focusing mostly on the assessment of the VIENNA 8 larvae exposure to parasitoids are needed to verify the effect of host larval quality on production of D. longicaudata. Accordingly, a series of future experiments involving different sizes and ages of third instars of VIENNA 8 will be performed.
We are thankful to Segundo Nuñez-Campero, Patricia Albornoz Medina, Carolina Chiappini, and Josefina Buonocore (PROIMI-CONICET), Cristina Coria and Victoria Beorchia Nigris (ProCEM-San Juan) for valuable technical assistance. Financial support was provided by the San Juan Eradication and Control Medfly Program (ProCEM San Juan), the Consejo Nacional de Investigaciones Científicas y Técnicas de la República Argentina (CONICET) (grant PIP/2008 No. 112–200801–01353) and by the Agencia Nacional de Promoción Científica y Tecnológica de Argentina (ANPCyT) through Fondo Nacional de Ciencia y Tecnología (FONCyT) (grant PICT/2010 No. 0393).