The experiments were conducted at the Teaching, Research and Extension Farm (21º15’04” S; 48º17’04” W) and at the Biological Control and Integrated Pest Management Laboratory of the Agricultural and Veterinarian Sciences School of Universidade Estadual Paulista “Júlio de Mesquita Filho”, Jaboticabal Campus, State of São Paulo, Brazil (21º14’28” S; 48º17’23” W) during the 2015 and 2016 seasons. Samples were collected in January and February in both seasons. The Department of Exact Sciences of Universidade Estadual Paulista “Júlio de Mesquita Filho” provided the meteorological data used in the study.

Soybean cultivars SYN 1365 RR® and BRX Potência RR® (both are tolerant to glyphosate) were sown in 1 ha during the 2015 and 2016 seasons, respectively – crops were not treated with insecticide during the experiment. Anticarsia gemmatalis larvae were collected with a beat sheet (length: 1 m - width: 0.5 m) in a crop row (length: 2 m) according to Souza et al. (2018); in total, 20 points were randomly sampled on a weekly basis after plant emergence. The collected larvae were stored in plastic containers and sent to the laboratory.

The collected larvae were stored in the Biological Control and Integrated Pest Management Laboratory under controlled temperature (25 ± 2°C) and relative humidity (60 ± 10%) conditions, in a 14:8-h photoperiod. Larvae were grouped based on their field instar; and individually placed in Petri dishes. Leaflets, which were daily replaced, covered the bottom of the Petri dishes for larvae feeding. Soybean leaves were previously cleaned with sodium hypochlorite (2.5%) for five minutes and washed three times in water. Petri dishes were sealed with polyvinyl chloride film to avoid leaflet dehydration. It was not possible differentiating the parasitized larvae from those non-parasitized; therefore, it was only possible to separate the parasitized group from the non-parasitized one when parasitoid larvae left the hosts to form the cocoon. A brief description of the experimental groups is presented in Table 1.

Table 1.
Description of the groups.

The following parameters were measured after larval collection in the field in the 2015 and 2016 seasons: Length of each larval instar, larval period of A. gemmatalis and the sex ratio of M. anticarsiae. Larval instars of A. gemmatalis grown in the laboratory environment were observed on a daily basis and changes on them were recorded at the same frequency. Larvae molt was characterized by the presence of shed head capsules in each Petri dish. Sex ratio was determined based on the ratio of females to the total population. Gantt charts of larval instar length and the parasitoid egg-to-pupa period were plotted to determine the larval instar of A. gemmatalis parasitized by M. anticarsiae in the field.

Based on the Gantt charts, larval instar lengths were organized in ascending order (from first to fifth instar). The egg-to-pupa period of M. anticarsiae parasitizing A. gemmatalis larvae was characterized by the first instar of larvae parasitized in the field, and this variable was used to determine the larval instar of A. gemmatalis parasitized by M. anticarsiae in the field. Laboratory culture was based on measurements of the following parameters: Egg-to-pupa, pupal and egg-to-adult periods.

Microcharops anticarsiae emerged from larvae collected in the field in the 2016 season were used to start the culture in the laboratory environment. Adult specimens represented the parental generation of the three cultivated generations. In total, 10 larvae (3-day old) of A. gemmatalis were conditioned in plastic container (5 L) with one drop of pure honey and a coffee cup (10 mL) of wet cotton; the container had two lateral openings (10 x 5 cm) covered with a mesh for airflow purposes. A glass pot (10 mL) was used to hold the soybean leaf. Larvae were placed on soybean leaves before parasitism began. Microcharops anticarsiae females were mated at the age of 8 days, prior to parasitism. One mated female was placed in the experimental container, and stayed there for 2 hours; this procedure was repeated 10 times, but with different females and larvae in each repetition. Larvae were placed in Petri dishes in separate, and fed soybean leaflet, similar to the procedure adopted for larvae collected in the field. All parasitoid females were released at Jaboticabal Campus after parasitism. Microcharops anticarsiae males and females fed on pure honey and wet cotton.

Statistical analyses were carried out in the R software for Windows (R Core Team, 2016). Normality was checked by the Shapiro-Wilk test and the Levene test was used to evaluate homogeneity of variances (α ≤ 0.05). Data were tested by one-way ANOVA, followed by Duncan’s test for comparison between factors. Differences were significant at α ≤ 0.05. Data were transformed into yλ by Box-Cox transformation, whenever necessary. Data of the laboratory culture were analyzed by repeated measures ANOVA.

Parasitism condition and the sex of larvae collected in the field were unknown at sampling. Parasitoids emerged from larvae collected in the field were fed on water and honey (10%), and were used in laboratory studies. Sample collection in the field was carried out based on the Brazilian legislation for scientific activities. The first author was granted with Authorization 49516-1 to carry out the research between 2014 and 2016. Adult parasitoids were sent to the American Entomological Institute for identification and those identified were deposited in the Museum of Entomology of Universidade Estadual Paulista “Júlio de Mesquita Filho” at the Agricultural and Veterinarian Sciences School, Jaboticabal Campus, State of São Paulo, Brazil.

The egg-to-pupa period (11.07 ± 0.46 days) was assessed through Gantt chart plots and results were used to determine the larval instar parasitized by M. anticarsiae in the field in the 2015 season (Table 2). Considering Figure 1, A. gemmatalis larvae in G1p were parasitized at their first larval instar and larvae at G2p, G3p and G4p were parasitized at their second instar, as shown in the Gantt chart. In total, 79 larvae from groups G2p (N = 17), G3p (N = 28) and G4p (N = 34) were parasitized by M. anticarsiae at their second larval instar in the field (Table 2). In addition, 14 larvae of A. gemmatalis from G1p were parasitized at their first larval instar; therefore, the host larvae were parasitized at the first (15%) and second larval instars (85%) in the 2015 season. Unfortunately, it was not possible to arrange groups with individuals collected in the 2016 season, given the low natural infestation with A. gemmatalis in the field (Tables 4 and 5).

Table 4.
Length of the first to fourth instar (mean ± SD) of non-parasitized and parasitized Anticarsia gemmatalis larvae, 2016.

Table 5.
Length of the fifth, sixth instar and total larval development (mean ± SD) of non-parasitized and parasitized Anticarsia gemmatalis larvae and the sex ratio of Microcharops anticarsiae, 2016.

7N: Total number of insects sampled in the field. 8Means with the same letters in columns indicate no significant differences (Duncan, p ≤ 0.05).

Sex ratio and development time of M. anticarsiae parasitizing A. gemmatalis

The sex ratio of M. anticarsiae was female-biased in the 2015 season, except for G3p, which recorded sex ratio of 0.49 (Table 2). Based on the observed sex ratio in the 2016 season, the population of M. anticarsiae was composed of 50% females, except for G2p, which was female-biased (Tables 4 and 5). Adult M. anticarsiae emerged in the 2016 season were used to start the laboratory culture, based on the preliminary results of the 2015 season.

The sex ratios of the first, second and third cultured generations of M. anticarsiae parasitizing 3-day old A. gemmatalis larvae were 0.56, 0.48 and 0.58, respectively; therefore, parasitism rate in these generations reached 92%, 81% and 90%, respectively (Table 6). The sex ratio of M. anticarsiae was female-biased when A. gemmatalis larvae were parasitized at the age of 3 days.

Figure 1
Gantt chart illustrating the larval instar lengths of Anticarsia gemmatalis parasitized by Microcharops anticarsiae in the field, during the 2015 season. L1, L2, L3, L4, and L5 represent the first, second, third, fourth, and fifth larval instar of A. gemmatalis. G1p, G2p, G3p, and G4p represent the parasitized larvae sampled at the first, second, third, and fourth instar in the field. G1 represents the non-parasitized larvae sampled at the first instar in the field.

Table 6.
Sex ratio and parasitism percentage of Microcharops anticarsiae in the laboratory environment.

The ‘generation’ factor affected the egg-to-pupa, pupal and egg-to-adult periods, and the factor ‘sex’ affected the egg-to-pupa period of M. anticarsiae parasitizing A. gemmatalis (Table 7).

Table 7.
Repeated measures ANOVA for the development time of Microcharops anticarsiae in the laboratory environment.

The egg-to-pupa, pupal and egg-to-adult periods were longer in the third generation of M. anticarsiae, and the egg-to-pupa period was shorter in females than in males (Figure 2).

Figure 2.
Development time of Microcharops anticarsiae parasitizing Anticarsiae gemmatalis larvae based on generation (A) and sex (B). Similar letters and ns on the columns indicate no significant differences and * on the columns indicates significant differences (Duncan, p ≤ 0.05). F1, F2, and F3: First, second and third generation of M. anticarsiae. Female and Male: Female and male adults of M. anticarsiae.

The egg-to-pupa period of M. anticarsiae in group G1p was used to determine the larval instar of parasitized A. gemmatalis larvae. This development time did not change when A. gemmatalis was parasitized at its first, second, third or fourth larval instars. This phenomenon occurs in other endoparasitoids; for instance, the development time of Microplitis mediator (Haliday) (Hymenoptera: Braconidae) is not affected by parasitizing host larvae of Mamestra brassicae (Linnaeus) (Lepidoptera: Noctuidae) at the first, second or third instar (Malcicka & Harvey, 2014), which indicates that endoparasitoids strategically maintain the development time changing other parameters like weight and sex ratio. Therefore, we considered the egg-to-pupa period needed by M. anticarsiae at the first instar (G1p) of A. gemmatalis was not different from that needed by M. anticarsiae to parasitize the second, third, and fourth larval instars of the assessed wasp species.

Based on Figure 1, M. anticarsiae parasitized the first and second larval instars of A. gemmatalis in the field; therefore, it is possible to state that M. anticarsiae more often parasitizes the early instars of A. gemmatalis in order to minimize mortality risks inherent in parasitism in older instars. Older larvae of A. gemmatalis are more efficient in attacking and harming the parasitoid during the parasitism process (Ameri, Rasekh, & Michaud, 2014; Firlej, Lucas, Coderre, & Boivin, 2010; Zhou et al., 2017); furthermore, parasitoid larvae have more time to develop better physiological conditions inside young hosts, since these hosts have less effective immune responses and/or produce fewer immune defenses than larger and more developed hosts (Kaeslin et al., 2010).

Other studies investigated different host-parasitoid systems and showed that ichneumonid parasitoids rather parasitize young larvae; for example, second instar larvae of Trichoplusia ni Hübner (Lepidoptera: Noctuidae) (3-5 days old) represent the most favorable stage for the development of Campoletis sonorensis (Cameron) (Hymenoptera: Ichneumonidae) (Murillo, Hunt, & VanLaerhoven, 2012). In addition, the second larval instar of Spodoptera littoralis (Boisduval) (Lepidoptera: Noctuidae) has proven to be suitable for the culture of Hyposoter didymator (Thunberg) (Hatem, Shawer, & Vargas-Osuna, 2016).

The present study sheds light on unanswered questions about the parasitic behavior of M. anticarsiae in the field, since the recorded findings provide important information about large-scale culture production methodologies focused on growing biological control agents to be applied to pests in crops.

Microcharops anticarsiae offspring was female-biased in most groups. The sex ratio is a critical factor to the mitigation of costs with the culture of this parasitoid, as well as an important attribute of biological control agents (Pandey, Kumar, & Tripathi, 2004). According to the literature, the sex ratio of Hymenopteran is influenced by factors such as diet (Ongaratto et al., 2019), instar (Poncio et al., 2018; Chu et al., 2014), superparasitism (Alvarenga, Dias, Stuhl, & Sivinski, 2015) and size/age of the host (Van Nieuwenhove & Ovruski, 2011; Ueno, 2015). Also, chemical cues are associated with the host characteristics, such as feces (Faraone, Svensson, & Anderbrant, 2017) and its ability to feed on different host plants (Morawo & Fadamiro, 2019). According to the present results, female M. anticarsiae individuals parasitize A. gemmatalis larvae at the end of the first and at the beginning of the second instar; these are the most favorable periods for parasitism for two reasons: larvae attach to the substrate at the end of the first instar and stop feeding, and the larvae stops moving at this point, M. anticarsiae has the advantage in parasitizing immobilized larvae; second reason, larvae remain weak at the beginning of the second instar and are easily exposed to parasitism - parasitoids can take advantage of this weak condition. Assumingly, M. anticarsiae can detect the volatiles emitted by early A. gemmatalis larvae (first and second instars) at molting; therefore, the end of the first and the beginning of the second instar may be the most favorable periods to parasitism, because they enable M. anticarsiae to produce female-biased sex ratio.

The effect of the factor ‘generation’ on the development time of M. anticarsiae indicates that the laboratory culture procedure needs improvements in order to avoid negative effects on the parasitoid development as reported by Bueno, Romero, Osorio, and Zaviezo (2017) for Mastrus ridens Horstmann (Hymenoptera: Ichneumonidae), because there were no introduced new parasitoids to the laboratory culture due to the lack of larvae in soybean crops in the field and laboratory. Using wind tunnels to help parasitoids finding the hosts, as well as olfactometers to determine the age of females to mate, could be instruments used to achieve such improvements. Otherwise, consanguinity had negative influence on the development time of third-generation parasitoids, because there were no new parasitoids in the laboratory culture due to the lack of larvae in soybean crops in the field. The difference between the egg-to-pupa period of males and females helps separating parasitoids by sex, even before emergence. The rejection of recently emerged M. anticarsiae females indicates that they rather feed or develop before mating. Our results provide important information to those coordinating the production of M. anticarsiae females in large-scale culture production programs focused on biological control.

The current study is expected to help and motivate new researchers to conduct other investigations on this topic, since there are more questions to be answered about parasitic behavior and host-parasitoid interactions in the field. This research broadens the knowledge about the development of biological control programs focused on agricultural applications.