Insect mass rearing

Strains of both C. tomentosicollis and G. fulviventre used in this study were derived from a laboratory mass rearing facility at the International Institute of Tropical Agriculture (IITA) in Cotonou, Benin. In the rearing room, the average temperature was 26.15°C ± 1.54°C and the mean relative humidity was 65.9% ± 7.6%. Clavigralla tomentosicollis nymphs and adults were supplied with fresh pods of a continuous crop of the sensitive Benin landrace cowpea variety, Kpodji-guê-guê, and were reared in cylindrical boxes (1900 mL each). Two strains (one from Burkina Faso and one from Benin) of the parasitoid G. fulviventre were reared on eggs of C. tomentosicollis. For this purpose, one mated G. fulviventre female was introduced in a cylindrical box (360 mL) for 48 hours with fifty fresh eggs of C. tomentosicollis. Emerging bug larvae hatching from nonparasitized eggs were removed from the box. The parasitized eggs were then monitored, and the subsequent generation of parasitoids emerged 16 days after parasitism occurred. Both strains of parasitoids were used in rearing and experiments as well.

For the olfactory tests, naive emerging G. fulviventre females were kept by pairs with males for mating prior to testing them. Burkina Faso and Benin are quite different in their climatic conditions which can influence insect ecology and behavior. That is the reason why both parasitoid strains were used in the hypothesis of a behavioral difference.

Obtaining the odors for olfactory tests

Several sources were used to obtain the tested odors. First of all, cowpea pods were collected from field at the IITA station. The harvested pods were previously washed with tap water to clean any other odor before use. The pods were placed under a stream of tap water for a few minutes without scrubbing to remove any insect odors that may be initially present. This does not change the structure of the pod and its odors and allows infestation with a controlled number of insects.

Infested pods were obtained by placing freshly harvested and cleaned pods with C. tomentosicollis adults and nymphs for 24 hours (24-hour infested pods) and 48 hours (48-hour infested pods). A total of five pods of each type were used for the experiment. Pods infested in this way for 48 hours are damaged enough to change the volatile compounds they emit. The healthy pods were not exposed to the insect prior to the trial. Three-day-old males and females of C. tomentosicollis collected from the rearing boxes were also used as source of odor. The insects used to produce odors were not provided with food over the course of the test.

Olfactometer setup

The response of G. fulviventre to volatiles produced by cowpea pods and C. tomentosicollis adults was investigated using a four-armed olfactometer previously described.¹⁴ This device was expected to determine the discriminatory ability of insects to more than two sources of odor at the same time. The device allows testing three different odors and one control at the same time as in a previous study.¹⁵ The four-armed olfactometer is a complex device that includes a central unit called the orientation chamber in which the insect can move freely. The chamber has four openings 90° apart that allow the entry of air. A hole which ensures the exit of the air is in the middle of the room. The latter is finally covered with a transparent glass plate. From the end to the orientation chamber, the olfactometer consists of a glass tube containing charcoal for the purification of air, another container of distilled water for humidifying the air, and four flowmeters that calibrate the air inlet. Each flowmeter was connected to an odor source (Figure 1). Clean airflow (wind speed: 300 mL/min) was divided into four subflows, and each subflow passed through one of the four odor sources connected to the four arms of the olfactometer. The device thus made it possible to deliver air loaded with each of the four test odors into the exposure chamber. The air exit hole in the middle of the orientation chamber prevents an accumulation of odors inside.

Figure 1.

Four-arm olfactometer device used for the different tests.

Picture: Apolline Sanou.

Experimental procedure

Mated, three-day-old, naive G. fulviventre females (without any prior contact with cowpea pods or C. tomentosicollis adults) were introduced individually into the exposure chamber and their behavior was observed for a maximum of 5 minutes. Six sets of four-odor combinations (three odors and one control) were tested (Table 1). The choice of three, five, ten, and fifteen individuals of male and female for pheromone experiment was based on preliminary tests. The increasing number of individuals was expected to yield increasing amount of pheromone which affects parasitoid attraction.¹⁶ A test began once the female started moving, and the time spent in each odor field was recorded. Indirectly, the frequencies of entering in each odor field were recorded. Females remaining motionless for more than 2 minutes at the release point were discarded from the analysis. After testing five naive females, the positions of the odor sources were exchanged to correct for any potentially unforeseen asymmetry in the experimental setup. After testing ten naive female wasps, odor sources were renewed. In total, sixty naive female parasitoids from each strain were tested for each combination. Before each test, the arena was cleaned with laboratory alcohol. In the olfactometer room, the average temperature was 27.2°C and the mean relative humidity was 63.4%.

Table 1.

Details of experiments to determine the influence of volatiles produced by cowpea pods and Clavigralla tomentosicollis adults on the host location behavior of Gryon fulviventre females.

Data analysis

Data were analyzed to determine differences in the effect of the odor source on the behavior of female parasitoids. The Shapiro-Wilk and Levene tests in R (R Development Core Team 2014) were used to analyze the times spent by the parasitoids in each odor source to check for normality and homogeneity of variances, respectively. Several series of transformations $(\log x,\log (x+1),\sqrt{x,}\sqrt{(x+0.5)},\ln x,\ln (x+1))$ were also used without normalizing the data. So, the Kruskal-Wallis test, as an alternative to analysis of variance, was used with a = .05. Friedman tests^17,18 with a = .05 were used to analyze the frequency of choice of G. fulviventre females for each of the odors tested using XLSTAT752 software (Addinsoft; XLSTAT752).

Response of parasitoid females to odors from healthy and infested cowpea pods (Experiment 1)

The time spent by the parasitoids in the arms receiving odors from cowpea pods and the clean air ranged from 56 to 98 seconds for both strains of parasitoid (Burkina Faso and Benin). Females of both G. fulviventre strains were not significantly attracted by volatiles from either healthy or damaged cowpea pods (P = .495 and P = .558, respectively; Table 2). Similarly, mean frequencies of parasitoid choice among the 4 odors tested did not significantly differ for the Burkina Faso (P = .492) and Benin (P = .321) strains.

Table 2.

Response of both strains of Gryon fulviventre females to clean air and odors from healthy pods, 24-hour and 48-hour infested cowpea pods.

Response of parasitoid females to odors from Clavigralla tomentosicollis females (Experiment 2)

The parasitoid females were not significantly attracted by odors emitted by females of C. tomentosicollis (Table 3). Differences in time spent in the four arms of the olfactometer, as well as frequencies of choice between odors, did not significantly differ for both strains (Table 3).

Table 3.

Response of Gryon fulviventre females to clean air and odors from increasing number of Clavigralla tomentosicollis females.

Response of parasitoid females to odors from C. tomentosicollis males (Experiment 3)

Gryon fulviventre females were attracted by odors from males of C. tomentosicollis depending on not only the strain tested but also concentration (Table 4). For the Burkina Faso strain, the time spent by the parasitoids in the arm diffusing volatiles from C. tomentosicollis males was significantly different (P = .037), whereas the frequency of choice among the four odor sources did not differ (P > .05). For the Benin strain, only the frequency of choice of G. fulviventre females for the odors from C. tomentosicollis males was significantly different (P = .001).

Table 4.

Response of Gryon fulviventre females to clean air and odors from increasing numbers of Clavigralla tomentosicollis males.

Discriminatory response of parasitoid females between combined odors from healthy pods, infested pods, and C. tomentosicollis males (Experiments 4-6)

As the compounds released by males of C. tomentosicollis attracted G. fulviventre females, we also tested their response to odors from an increasing number of males in comparison with volatiles from healthy and infested cowpea pods (Tables 5 to 7).

Table 5.

Discriminatory response of Gryon fulviventre females to volatiles from cowpea pods and five males of Clavigralla tomentosicollis.

Table 6.

Discriminatory response of Gryon fulviventre females to volatiles from cowpea pods and ten males of Clavigralla tomentosicollis.

Table 7.

Response of Gryon fulviventre females to discriminate volatiles from cowpea pods and fifteen males of Clavigralla tomentosicollis.

Females of both strains showed no significant difference for the time spent or the frequency of choice for the odor sources tested with five males (Table 5). However, females were attracted by the odors released by ten males of C. tomentosicollis as evidenced by the higher frequencies of choice in Burkina Faso strain (P = .028). Females in Benin strain showed a discriminatory capacity of the different odors (P = .008), but they did not discriminate odors between infested pods and ten males (Table 6). Again parasitoid females of both strains were not significantly attracted when the tested odors were provided by cowpea pods and 15 C. tomentosicollis males (Table 7).