Bracon wasps for ecological pest control–a laboratory experiment

Biological control of pest insects by natural enemies may be an effective, cheap and environmentally friendly alternative to synthetic pesticides. The cosmopolitan parasitoid wasp species Bracon brevicornis Wesmael and B. hebetor Say (Hymenoptera: Braconidae) use lepidopteran species as hosts, including insect pests like Ephestia kuehniella or Ostrinia nubilalis. Here, we compare the reproductive success of both Bracon species on E. kuehniella in a laboratory experiment. We asked (1) how the reproductive success on a single host larva changes with temperature, (2) how it changes with temperature when more host larvae are present and (3) how temperature and availability of host larvae influence the efficacy of Bracon species as biological control agents. In general, differences between B. brevicornis and B. hebetor have been small. For rearing both Bracon species in the laboratory on one host larva, a temperature between 20–27 °C seems appropriate to obtain the highest number of offspring with a female-biased sex ratio. Rearing the braconid wasps on more than one host larva revealed a higher number of total offspring but less offspring per host larva on average. Again, highest numbers of offspring hatched at 27 °C and the sex ratio was independent from temperature. Although no parasitoids hatched at 12 °C and only few at 36 °C, host larvae were still paralyzed. The efficacy of B. brevicornis was higher than 80% at all numbers of host larvae presented at all temperatures while the efficacy of B. hebetor was less than 80% at 12 °C and 27 °C at low numbers of host larvae presented. In conclusion, practitioners can use either B. brevicornis or B. hebetor at low and high temperatures and at varying host densities to achieve high pest control efficacy.

Bracon brevicornis and B. hebetor are promising species for biological pest control since females lay eggs on various pest host species, stop their development by paralyzing them and finally reduce pest densities (Taylor, 1988;Akman Gündüz & Gülel, 2005). However, both Bracon species are very similar morphologically (separated by wing venation, larval morphology and genital characters; Matthews, 1974) and also show similar life cycles (Alam et al., 2016;Srinivasan & Mohan, 2017). This parasitic life cycle can be influenced by several factors (e.g., . There is evidence that Bracon females can regulate clutch size based on the size and/or quality of the host (Taylor, 1988;Godfray, 1994;Milonas, 2005). Furthermore, host density and density of the parasitoid population seem to influence life cycle statistics and sex ratios; i.e., B. hebetor produced more females if densities were high (Galloway & Grant, 1988;Singh, Singh & Tripathi, 2016). Additionally, temperature plays an important role in the developmental biology of Bracon species and also affects the efficacy of parasitization (Rao & Kumar, 1960;Thanavendan & Jeyarani, 2010). As agents for biological control, braconid wasps are often used in addition to trichogrammatid parasitoids which parasitize eggs of insect pests (Brower, 1988) which is also done to dam E. kuehniella (Ayvaz & Karabörklü, 2008; St- Onge et al., 2015). If high pest pressure occurs and many eggs remain unparasitized and develop to larvae, B. brevicornis and B. hebetor can be used in support. Ideally, the rearing of biological control agents should be inexpensive and simple, with no special requirements or culture conditions. Furthermore, in Bracon wasps a high amount of female offspring is desired since only female wasps paralyze and kill the host larvae. For practitioners, high efficacy, i.e., a high number of paralyzed and killed pest larvae, at low cost is important. To further investigate and compare the best rearing conditions and efficacy of B. brevicornis and B. hebetor, we asked: (1) how reproductive success of female wasps on a single host Ephestia kuehniella larva changes with temperature, (2) how reproduction success of female wasps changes with the availability of more host larvae at different temperatures, and (3) how temperature and varying availability of hosts affect the number and mortality of paralyzed host larvae and therefore the potential efficacy of braconid wasps as biocontrol agents.

Rearing of Bracon brevicornis and Bracon hebetor
The laboratory studies were conducted between January and June 2019. The first generation of B. brevicornis and B. hebetor was provided by AMW Nützlinge, Pfungstadt, Germany. Cultivation started with three strains of B. hebetor and two strains of B. brevicornis collected from different hosts (Appendix 1). For rearing, two females of either B. hebetor or B. brevicornis and ten Ephestia kuehniella larvae (last stage before pupation L6) were placed in Petri dishes with 5 cm diameter, using five to ten Petri dishes per strain and generation. Until pupation, Petri dishes were placed in climate chambers at 27 C, 70% relative humidity (RH) and a light-dark photoperiod of 16:8 (L:D). Subsequently, the cocoons were transferred in glass bottles and left at room temperature until hatching. Female and male wasps were kept together after hatching and fed with sugar water. At the earliest 48 h after hatching but not later than 5 days after hatching, the female wasps were used for the experiments. Unused specimens were frozen after 5 days, and only young females were ever used.

Reproductive success of female wasps on a single host larva at different temperatures
The number of offspring from each individual female wasp of either B. brevicornis or B. hebetor on one larva of E. kuehniella was recorded. Therefore, one female of each Bracon strain and one host larva were set in a Petri dish with 5 cm diameter and placed in climate chambers with 12 C, 20 C, 27 C or 36 C with steady humidity of 70% and 16:8 L:D photoperiod. Each temperature was replicated per species and strain; the final number of replications varied due to difference in the developmental cycle and the survival rate (see Appendix 2 for initial numbers of replications). Samples were frozen when all wasps had hatched, but no later than 5 weeks after the start of the experiment. Subsequently, the number of hatched female and male wasps was determined.

Reproductive success of female wasps on variable numbers of host larvae at different temperatures
The number of offspring of one female wasp of either B. brevicornis or B. hebetor on varying numbers of E. kuehniella larvae was recorded. Therefore, one female wasp of each Bracon strain and 5, 10, 15, 20 or 25 E. kuehniella larvae, respectively, were kept in Petri dishes or plastic tubes with rectangular shape (5 cm diameter, 2.0 cm × 15.2 cm × 2.0 cm height × width × depth) in the climate chamber regimes as described above. Since female wasps have not been fed during the experiments, they were removed after 4 days. Furthermore, we expected 4 days to be the timeframe of the main oviposition period (Kabore et al., 2019). Larvae were tested for paralysis through a short squeeze of the head capsule with a featherweight forceps, showing no movement when paralyzed. Numbers of paralyzed and still active larvae were documented. Subsequently, paralyzed larvae were kept in climate chambers until all parasitoid larvae hatched but no longer than 5 weeks after the start of the experiment. Afterwards the samples were frozen and the number of hatched female and male wasps was documented. As a control, the same number of host larvae was incubated over the same period without female wasps, and moth development was documented. To determine the efficacy of pest control by the braconid wasps, paralyzation rates were calculated as paralyzation rate = number of paralyzed host larvae/ number of presented host larvae.

Data analysis
Data were analysed with RStudio Version 1.1.383. First, the impact of strain affiliation was tested using linear model ANOVA. Since no differences between strains were observed (temperature Ã strain: F 3,617 = 0.688 p = 0.560, strain: F 3,617 = 1.146 p = 0.330), all strains were pooled for further analyses. Before statistical analyses, normal distribution of residuals was tested by Shapiro test and variance homogeneity by Levene's test The response of total offspring of either B. brevicornis or B. hebetor hatched from one host larva from one female to changing temperatures was tested using general linear model (GLM) with poisson distribution and Chi-test. Tukey pairwise comparison was used subsequently. Similarly, sex ratio of offspring in response to different temperatures and Bracon species was tested using GLM with Poisson distribution (see Appendix 3 for raw data).
For testing the response of total offspring or sex ratio when different numbers of host larvae were presented, temperature, species and number of host larvae were included as variables in a GLM analysis with Poisson distribution and Chi-test. The number of hatched parasitoids (total, male, female) was square root transformed before statistical analysis. For pairwise comparisons, Tukey post hoc test was used (see Appendix 4 for raw data).
Differences in the efficacy of the braconid wasps between temperature and number of host larvae was tested using Kruskal-Wallis tests because of lacking normal distribution.

RESULTS
How does the reproductive success of female wasps of B. brevicornis and B. hebetor on a single host larva change with temperature?
In both species, the sex ratio was not temperature dependent (species × sex: χ 2 (11,299) = 272.57, p = 0.105; temperature × sex: χ 2 (11,299) = 272.57, p = 0.907). Although not significant, in B. brevicornis, the sex ratio was balanced at 20 C (♂/♀ 1.02), became Presenting different numbers of host larvae at different temperatures to females of B. brevicornis and B. hebetor revealed a similar pattern of offspring for both species (host larvae × species × temperature: χ 2 (7,627) = 1073.78, p = 0.512; Fig. 3, Table 2). In general, the number of offspring in Bracon species differed with the number of presented host larvae and temperature (host larvae respectively temperature: χ 2 (7,627) = 1073.78, p < 0.001). At 12 C, no offspring hatched at any number of presented host larvae, but most host larvae were paralyzed and did not develop to imagoes within 5 weeks. At 27 C, the highest total offspring was observed in both species (Fig. 3, Appendix 5). At 36 C, oviposition was observed in 20 of 140 samples, and only few parasitoids hatched; therefore, statistical analysis for hatched offspring within this treatment seemed inappropriate.
In B. brevicornis, the number of total offspring at 20 C did not change with the number of presented larvae (F 1,98 = 0.05, p = 0.826; Fig. 3A, Appendix 5). Mean values varied between 11.3 ± 6.6 at 15 host larvae and 19.4 ± 11.5 at 10 host larvae. At 27 C, the number of total offspring varied significantly with the number of presented host larvae, changing from 27.9 ± 10.7 at 5 host larvae to 45.7 ± 19.3 on average at all higher numbers of host larvae (F 1,88 = 21.46, p < 0.001; Fig. 3A, Appendix 5).

Comparsion
Bracon 0.020 * offspring were smaller at 27 C compared to 20 C and smaller in B. hebetor as compared to B. brevicornis (Fig. 4). While offspring was generally male biased, females dominated in B. brevicornis at 10 presented host larvae at 27 C and in B. hebetor at 25 presented host larvae at 20 C. A sex ratio of approximately 1:1 was observed at 15 host larvae at 20 C in both Bracon species and at 15 host larvae at 27 C in B. brevicornis (Fig. 4, Appendix 5).
Considering the offspring output per host larva, the highest number of offspring per larva was observed when only one host larva was present which was independent of the temperature (Table 3). If more than one host larva was present, a higher number of   offspring per larva was observed at 27 C in both Bracon species, being highest-although lower than on one host larva-when 5 or 10 host larvae were present (Table 3).

Do different temperatures and varying numbers of host larvae affect the efficacy of braconid wasps?
Paralyzation rates differed significantly with temperature and species (temperature × species: F 2,641 = 7.14, p < 0.001; Fig. 5, Table 4). In B. brevicornis, rates were continuously higher than 0.8 for all numbers of host larvae and temperatures. In B. hebetor, rates varied strongly between 0.5 at 12 C with 10 or 15 host larvae and 1or almost 1 (0.99) at 36 C for all numbers of host larvae (Fig. 5, Table 4).

DISCUSSION
How does the reproductive success of female wasps of B. brevicornis and B. hebetor on a single host larva change with temperature?
While presenting one host larva to one female wasp in our study, B. hebetor produced slightly higher numbers of offspring compared to B. brevicornis and oviposition was reduced by increasing temperature. Yu et al. (2003) reported an average of 12.8 eggs oviposited by one B. hebetor female on one host larva (Plodia interpunctella, 28 ± 0.5 C, 70-75% relative humidity, 16:8 L:D photoperiod). Since the average numbers of eggs per one host larvae in our study were lower (~9 at 20 C and 27 C), the choice of the host species seems to have an impact on the reproductive outcome. Previous studies reported that the best host species for rearing Bracon wasps in the laboratory is the greater wax moth Galleria mellonella (Farag et al., 2015;Khalil et al., 2016) which can produce about 260 total offspring within about 38 days (Nikam & Pawar, 1993). However, highest number of females, which are needed to control host pest species, were produced at 27 C, with a sex ratio of nearly 1:1 in B. hebetor and even 0.86 in B. brevicornis while nearly all other temperatures showed male-biased sex ratios. Since only female wasps contribute to the effective success of biological pest control, the sex ratio of a laboratory strain of Bracon wasps is of critical importance. Mohamad, Mansour & Ramadan (2015) reported a correlation between sex ratio and temperature whereby high Table 3 Mean number of offspring per one host larva when emerged from different numbers of host larvae at different temperatures. Mean number of offspring per one host larva when emerged from different numbers of host larvae (5,10,15,20,25) at different temperatures (20 C, 27 C) in Bracon brevicornis and Bracon hebetor. Standard deviations are given in brackets.

Bracon brevicornis
Bracon hebetor 12 C 2 0 C 2 7 C 3 6 C 1 2 C 2 0 C 2 7 C 3 6 C 5 0.93 ± 0.14 0.91 ± 0.24 1 ± 0 1 ± 0 0. The production of offspring and the pattern of sex ratios in response to different numbers of host larvae and temperature was very similar between B. brevicornis and B. hebetor. Presenting more than one host larva to one female wasp resulted in higher numbers of total offspring while the average number per larva was lower as when a single larva was present. Since host quality affects different performance parameters of parasitoids (Bernal, Luck & Morse, 1998;Mody et al., 2017), reducing the number of eggs per host larva could result in increased fitness per parasitoid larva due to reduced competition for food. Furthermore, the processing time one female wasp needs to detect and paralyze a host and to oviposit eggs increases with the number of hosts presented, resulting in less eggs in average (Yu et al., 2003).
As seen for only one host larva, highest numbers of offspring were observed at 27 C in both species when more larvae were present. Higher temperatures (36 C) seem to negatively affect the development, probably by heat stress. Direct effects of heat stress were reported by Klockmann, Kleinschmidt & Fischer (2017), showing that heat stress reduced survival rates and fitness of a Lepidoptera larva (Bicyclus anynana). Furthermore, there are indirect effects of heat stress for parasitoids (Singh, Singh & Tripathi, 2014). Bracon wasps belong to the holometabolous insects that undergo a metamorphosis from egg over larva and pupa to adult. For the metamorphosis and the pupation of the larvae the so-called 'critical weight' must be achieved. When the critical weight is attained, the larvae begin to pupate and start the metamorphosis (Harrison, Woods & Roberts, 2012). At 36 C, the loss of water of host larvae may cause a reduction of the food quality for the parasitoid larvae. If parasitoid larvae starve, the critical weight is not reached; therefore, the larvae cannot develop into an adult but die. On the other hand, no oviposition and nearly no activity of the female wasps was observed at 12 C, indicating that they entered chill-coma, a reversible state that is caused by a reduction of physiological processes (MacMillan & Sinclair, 2011;Singh, Singh & Tripathi, 2014).
Besides temperature, braconid wasps showed adjustments in their oviposition behavior according to differences in host density in previous studies (e.g., Godfray, 1994;Yu et al., 2003;Milonas, 2005;Mohamad, Mansour & Ramadan, 2015). In B. brevicornis, numbers of offspring increased with increasing numbers of presented host larvae at 27 C. In B. hebetor on Corcyra cephalonica, searching efficiency of female parasitoids was best when five host larvae were presented (Singh, Singh & Tripathi, 2016).
In this study, this pattern could be confirmed when Bracon species were reared on one host larva only, but when more host larvae were present the sex ratio of offspring was independent from temperature. On the other hand, sex ratio of offspring slightly changed with varying numbers of host larvae. While Yu et al. (2003) found a sex ratio of approximately 0.5 irrespective of the number of presented host larvae, other studies reported labile sex ratios of B. hebetor in response to density, but with different results: if parasitoid densities were high more females were produced (Galloway & Grant, 1988), but on Galleria mellonella the proportion of females decreased with increasing parasitoid density (Alam et al., 2016).
Do different temperatures and varying numbers of host larvae affect the efficacy of braconid wasps?
Practitioners like farmers care about the efficacy of pest control and not about the number of offspring of the control agent. Nevertheless, they benefit from an economical breeding process, as this allows lower prices per agent application. To replace or at least supplement pesticides, biological control agents should be nearly as effective as pesticides but provide additional advantages.
The current study showed high effectiveness of Bracon species at all temperatures (12-36 C) and host numbers (five-25) with high paralyzation rates under laboratory conditions. However, concerning the efficacy both Bracon species showed the greatest differences. While B. brevicornis was effective at all temperatures, B. hebetor was less effective at low temperatures. However, although no offspring was observed at 12 C, female wasps must have injected venom. Since female wasps were kept at room temperature before they were confronted with the host larvae and exposed to 12 C, they could have paralyzed the host larvae before turning to the chill coma. The paralysis and efficacy at low temperatures represent particular advantages for practitioners. When pest-infested grain-or other stored-products are chilled and not frozen during storage, pests turn over to chill coma, but can develop when conditions become favorable again (Evans, 1987;Locatelli, Papale & Daolio, 1990;Andreadis, Eliopoulos & Savopoulou-Soultani, 2012). Preventive usage of B. brevicornis or B. hebetor could protect grain-or other stored-products before and during chilling against pests like Ephestia kuehniella or Plodia interpunctella.
Furthermore, parasitization over a broad range of temperatures and host densities makes both Bracon species more effective for their use in open fields, where environmental conditions like temperature can vary. In Germany, there are guidelines ('good agricultural practices') regulating the use of pesticides in connection with environmental conditions. Pesticide use should hence be avoided when temperatures permanently exceed 25 C, which becomes more common in summer, or air humidity is below 30%, otherwise penalties will have to be paid. B. brevicornis or B. hebetor could be used effectively in either case.
To summarize the useful facts of this study for breeders: using ten host larvae and one female wasp at 27 C is recommended to increase the number of offspring in total and to keep the number of hatched parasitoids per host at a high level. Practitioners can use either B. brevicornis or B. hebetor at low and high temperatures as well as with varying host densities to achieve a high level of effectiveness of pest control.