Cold storage-mediated rearing of Trichogramma evanescens Westwood on eggs of Plodia interpunctella (Hübner) and Galleria mellonella L.

The egg parasitoid Trichogramma evanescens Westwood is considered as an efficient biological control agent for managing several lepidopteran pests and it is widely distributed throughout the world. Mass rearing protocols of parasitoids that are currently in use in biocontrol programs require a meticulous quality control plan, in order to optimize their efficacy, but also their progeny production capacity. In this paper, the effect of different factors on the quality control in mass rearing of T. evenescens, using Plodia interpunctella (Hübner) and Galleria mellonella L. as host species, were investigated. The impact of egg agewas significant in the rates of parasitism, for both host species tested. Significantly highest percent of parasitoid emergence was noticed in two day-old eggs for both host species, while one day-old eggs day exhibited the maximum emergence when both species were used togetherin the same trials. Age-dependent storage egg preservation at either 4 or 9°C significantly influenced the parasitism percentages on both species. The highest parasitism percentage was recorded in two day-old G. mellonella eggs that are kept for 15 days at 9°C while the lower in one day-old P. interpunctella eggs for 60 d storage. Moreover, the highest parasitoid mortality was recorded in T. evanescens reared either on P. interpunctella or G. mellonella at 20°C. Rearing of the parasitoid on a mixture of eggs of both host species resulted in higher parasitism, but not always in higher rates of parasitoid emergence. The results of the present work provide useful information that can be further utilized in rearing protocols of T. evanescens.


Introduction
The egg parasitoids of the genus Trichogramma (Hymenoptera: Trichogrammatidae) have been extensively used as efficient biocontrol agents in integrated pest management programs interpunctella individuals were also collected from the stock cultures maintained at the Post-Harvest Entomology Laboratory since 2014, and they were reared on a diet of corn meal, chick laying mash, chick starter mash, and glycerol at a volumetric ratio of 4:2:2:1, respectively [31]. Both cultures were maintained in an incubator set at 27˚C, 70% relative humidity (RH), with a photoperiod of 16:8 (L: D) h.

Parasitoid origin and rearing
The parasitoid was originally received from the Bangladesh Agricultural Research Institute (BARI), Gazipur, Bangladesh. The species was reared using eggs(<24 h old) of P. interpunctella, following the method of Hegazi et al. [32]. The host eggs were stacked in a paper strip (1.5×3 cm) with gum arabic glue and placed in a transparent glass vial (3.5 cm in diameter x 12 cm in length) exposed to parasitoids [33]. The glass vials were covered by cloth-wrapped cotton. The egg strips were renewed daily to avoid super parasitism.

Experimental procedures
Host age-dependent parasitoid performance. To estimate the age-dependent performance of T. evanescens, 1, 2 and 3d-old eggs of P. interpunctella and G. mellonella were used. A total of 25 eggs of each age and species were stacked in a paper strip with gum arabic glue and placed in a transparent glass vial, as noted above, containing one-day old male and female pair of T. evanescens. The host eggs of both species were exposed to the parasitoid for 24 h. The mouths of the vials were wrapped in a white filter paper with the help of rubber and then were kept separately in an incubator set at 25˚C, 16:8 (L:D) photoperiod and 70% RH. Twelve host eggs were taken for each age and individual species for carrying out the experiment of combination of host eggs. There were three replications for each age and species. The percent parasitism (%) of T. evanescens was recorded based on eggs that turned into blackened colour. The glass vials were checked every day until all the adult parasitoids were dead. During this period, we recorded parasitism and emergence, as well as the female: male proportion. Parasitism, emergence, adult longevity and developmental period of T. evanescens were recorded in the F 1 individuals. After death, the adult body length was also measured using ocular micrometer under the microscope.
Cold storage impact on the parasitoids. Two age groups, i.e. 1 and 2-old, of host eggs of P. interpunctella and G. mellonella were selected for different exposure times to low temperatures, i.e. 0 (control), 15,30,45 and 60 d at either 4or 9˚C [34]. Twenty-five eggs of each host species and age were stacked in a paper strip with gum arabic glue and placed in a transparent glass vial, as above. The mouths of the vials were wrapped in a white filter paper with the help of rubber. After that, they were kept separately in incubators set at either 4 or 9˚C, with the rest of the conditions were set at 16:8 (L:D) photoperiod and 70% for RH, for each of the different storage periods. The glass vials were removed after the completion of each storage period. Then a single pair of male and female T. evanescens was introduced separately into the vials containing the cold stored host eggs for each host age, species and storage period for parasitism and kept in an incubator set at 25˚C, 16:8 (L:D) photoperiod and 70% RH. A small cotton absorbing 10% aqueous sucrose-glucose-fructose mixture (1:1:1) was supplied in the vial for feeding the T. evanescens adults. After 2 days of parasitism, the host eggs were checked for the percent of parasitism. The effects of cold storage in the percentage of adult emergence and sex ratio in T. evanescens were recorded separately for each host species. The adult body length was also measured using an eye piece-micrometer (New York Microscope Company, Hicksville, NY, USA). There were three replicates for each cold temperature, host species, age group and storage periods.

Temperature-dependent fitness of parasitoids.
To evaluate the temperature-dependent fitness of T. evanescens,1 d-old host eggs of each species was exposed at different temperatures, i.e. 15, 20, 25 and 30˚C. Twenty-five eggs of each species and age were stacked in a paper strip with gum arabic glue and placed in a vial, as noted above. The mouths of the vials were wrapped in a white filter paper with the help of rubber. After that, a single pair of male and female T. evanescens was introduced separately into the vials containing host eggs for each species and keptseparately in incubators set at the temperatures mentioned above, and at 16:8 (L: D) photoperiod and 70% RH. Percentage of parasitism, adult emergence, sex-ratio, developmental periods and adult body length were evaluated following the procedures described above for cold storage experiments. There were three replicates for each cold temperature, host species, age group and storage period.
Data analysis. The assumptions of normality and homogeneity of variance were determined using Levene's test [35]. The percentages of parasitism, adult emergence and female progeny of parasitoids for all the trials were transformed to arcsine square root for stabilizing variances before being subjected to ANOVA, but the untransformed data are presented in the results for clarity. For each of the different trials including the effects of host age, cold storagedependent and temperature, the data were subjected to analysis by multi-factorial ANOVA with the "host egg age" and "host species" as factors using the PROC ANOVA [36]. A k-valuewas calculated for different temperatures that are primarily responsible for an increase or decrease in the number of parasitoids in a given population. Means were compared by Tukey-Kramer HSD test at the 5% level.

Host age-dependent parasitoid performance
Parasitism rates varied significantly among the different egg age groups (F = 17.60; df = 2,24; P<0.001). The highest parasitism rate was recorded in one day eggs for both host species (Table 1). However, there were no significant variationsinthe species and their combination for all age groups (F = 0.16; df = 3,24; P = 0.92). Moreover, the interaction age � species was not significantly different in the percent of parasitism (F = 0.77; df = 6,24; P = 0.60). Adult emergence inT. evanescens varied significantly in all egg age groups (F = 5.96; df = 2,24; P = 0.007) ( Table 1). The highest percentage of emergence was recorded in 2 d-old eggs for both species while 1 d-old eggs showed the maximum emergence in the combination of species (Table 1). In addition, the adult emergence did not vary significantly in the species and their combinations (F = 0.75; df = 3,24; P = 0.53), but the interaction of age � species was significant (F = 2.61; df = 6,24; P = 0.04). Percent female progeny of T. evanescens did not differ significantly among treatments (F = 0.68; df = 11,24; P = 0.74) and the maximum (67.41) percent of female progeny was recorded in the case of P. interpunctella eggs, resulting from the combination with P. interpunctella in the case of the 2d-old eggs ( Table 1). There were no significant variations in the percent female progeny among the different egg age group (F = 0.54; df = 2,24; P = 0.59) and species including their combinations (F = 0.74; df = 3,24; P = 0.54). Adult longevity of T. evanescens was significantly lengthened in 1 d-old eggs, compared to 2 and 3-dold eggs (F = 22.06; df = 11,24; P<0.001) ( Table 1). The longevity was found to be shorter in the case of the eggs of P. interpunctella, especially in 2 and 3 d-old eggs. Also, parasitoid longevity was significantly different among the egg age groups (F = 81.33; df = 2,24; P<0.001) and species combinations (F = 10.67; df = 3,24; P<0.001). The interaction of age � species also varied significantly (F = 8.00; df = 6,24; P<0.001) for adult longevity. The different egg age groups significantly influenced the developmental periods in the case of both host species (F = 262.91; df = 11,24; P<0.001) ( Table 1). The highest (14 d) developmental period was recorded for the combination of G. mellonella at the 1 d-old eggs. The developmental periods of T. evanescens also showed a significant variation among the egg age groups (F = 1324.00; df = 2,24; P<0.001) and species combinations (F = 69.33; df = 3,24; P<0.001). There were also significant variations in the interaction of age � species for developmental period (F = 5.83; df = 6,24; P<0.001). The adult body length of T. evanescens resulting from different egg age groups varied significantly (F = 6.45; df = 11,128; P<0.001) ( Table 1). The maximum body length was recorded in T. evanescens resulting from G. mellonella in all age groups. The body length also varied significantly among the age (F = 21.04; df = 2,128; P<0.001) and species combinations (F = 7.53; df = 3,128; P<0.001). The interaction age � species was not significant for adult body length (F = 1.04; df = 6,128; P = 0.40).

PLOS ONE
Cold storage-mediated rearing of T. evanescens species (Table 3). In most combinations, adult body length was higher in the individuals that were developed from G. mellonella, as compared with the respective figures of P. interpunctella. The body length varied significantly among temperatures (F = 171.68; df = 3,41; P = 0.001) and between the species (F = 1752.03; df = 1,41; P = 0.001). There were also significant variations (F = 9.32; df = 3,41; P = 0.001)in the interaction of species � temperature for the adult body length. Finally, the highest mortality levels were recorded in T. evanescens reared at 20˚C for either host species, as indicated by the higher k-values (Table 3).

Discussion
Egg parasitoids have to be efficient in locating hosts as the life span of the egg stage is generally short under normal conditions. Moreover, previous studies indicate that old eggs are less suitable as hosts as they adversely affect several parasitoid biological parameters, such as percentage of parasitism, developmental time [37], adult emergence [38], body size and sex ratio [39].  These effects may be due to changes in the chemical composition as nutrients that are gradually consumed by the host embryo and change, or there are changes in the physical characteristics of the chorion, which becomes more rigid as the egg ages [40,41]. In this context, it is well established that preference for hosts towards a particular host age group may enhance the fitness of egg parasitoids [42,43]. Thus, parasitoids might make use of the physical characteristics and/or chemicals in and on the surface of eggs, which drastically change with age, as critical cues indicating the suitability of hosts. Nevertheless, our results demonstrate that egg age may not be of critical importance, at least in the case of the species range tested here, and can be moderated by other factors, such as the host species and previous exposure to low temperatures. In this context, the differences between the age groups utilized here were not very diverse, in terms of rates of parasitism.
We have found an interesting interaction between egg age and storage period. As a general principle, we saw that the percentage of parasitism was gradually decreased with the increase of the cold storage period. However, this decrease exhibited similar trends for both parasitoids only in the case of 1d-old eggs. In contrast, for the 2 d-old eggs G. mellonella performed better as a host at 4˚C, providing a considerable percentage of parasitized eggs, even at the longer storage period (60 d). It is generally considered that G. mellonella eggs take longer to hatch, as compared with those of P. interpunctella [44,45], although comparative studies are not available towards this direction. Hence, the difference between the two egg age groups for G. mellonella can be narrower as compared with the respective figures of P. interpunctella, which indicates that G. mellonella can be more suitable as a host for T. evanescens, in terms of more gradual egg "maturation". In an earlier comparison of these two species as hosts of the larval ectoparasitoid Harbobracon hebetor (Say) (Hymenoptera: Braconidae), Hasan et al. [29] considered G. mellonella as a superior host, due to certain larval characteristics, such as size and longevity. Paradoxically, our data illustrate that 2 d-old eggs of this host species maintain a good level of suitability for parasitism by T. evanescens, for a long period of time, which means that 2 d-old eggs are less affected by storage periods as compared with those of P. interpunctella. Nevertheless, morphological variations may drastically contribute to egg suitability for parasitism. For instance, the egg parasitoid Telenomus remus Nixon (Hymenoptera: Scelionidae) is able to parasitize the eggs in the basal layers of Spodopteraspp. (Lepidoptera: Noctuidae) egg masses, even when they are covered by moth scales, much more effectively than Trichogramma spp. [46], suggesting that the latter is more "selective" regarding egg preference.
In contrast with storage at 4˚C, storage at 9˚C differentiated the parasitism rates in a more uniform way for both host species. However, even in this temperature, parasitism rates were higher at G. mellonella eggs, as compared with P. interpunctella eggs. Storage of G. mellonella at this temperature provided very high percentages of parasitism up to 30 d; in fact, storage for 15 and 30 d provided even higher percentages than untreated eggs (0 d). For the larval endoparasitoid Venturia canescens (Gravenhorst) (Hymenoptera: Ichneumonidae), Eliopoulos et al. [47] noted that there was a critical egg maturation stage, which was directly related with viability and progeny production capacity. For the same species, Andreadis et al. [48] found that cold tolerance was negatively correlated with parasitoid age. We are unaware if egg exposure to low temperatures increases their suitability for parasitism, in the same way that larval irradiation positively affects parasitism in moth larvae by H. hebetor [49], but this hypothesis needs additional investigation. Considering the overall data for this series of bioassays, we assume that older eggs were more susceptible to low temperatures as compared to young eggs.
The combined use of both host species generally provided higher parasitism rates, which, paradoxically, did not always yield in higher parasitoid emergence rates. Adult parasitoid emergence in the combined use of both host species, was reduced in eggs that were 2 and 3 dold, as compared with the use of one single host species. Spatial segregation of hosts may result in different parasitism rates, resulting in host preference patterns and parasitoid aggregation in specific host groups [50]. Adaptation and counter-adaptation of a parasitoid in different hosts that can coexist locally is directly related with host rate advantage [51]. Practically, the simultaneous utilization of more than one host species may offer specific benefits to the overall parasitoid rearing technique, as parasitism may be shifted to an alternative host, if there is a breakdown in the case of the superior host species, e.g. change of egg suitability.
In contrast with egg age and previous exposure to cold, parasitism rates were similar on both host species, regardless of the temperature level for the rearing of the parasitoids, at least at the range of temperatures examined here. Moreover, despite the fact that we observed a reduction in parasitism at 20˚C, the increase of temperature beyond that level further increased the parasitism. In a similar series of tests with Trichogramma turkestanica Meyer, reared of eggs of the Mediterranean flour moth, Ephestia kuehniella Zeller (Lepidoptera: Pyralidae), Hansen and Jensen [52] found similar temperature-dependent parasitism trends. Still, different species of Trichogramma exhibit dissimilar developmental and parasitism rates in relation with temperature, but there are species of this genus for which population growth is benefited by the increase of temperature [53]. Consequently, our data demonstrate that maximum parasitism can be achieved at temperatures that are higher than 20˚C, which should be taken into account in T. evanescens rearing protocols, or enhancement of the parasitoid progeny production when using eggs that have been previously exposed to low temperatures.
Paradoxically, exposure of eggs to low temperatures changed the female ratio, in a different way for each of the host species. In some of the combinations tested, the female ratio was higher in untreated (not exposed to cold) eggs, which partially explains the higher parasitism rates that were recorded in these combinations, despite the fact that the highest female ratio was not always correlated with higher rates. Moreover, for 1 d-eggs that had been exposed to either 4 or 9˚C, we observed a gradual increase of the female ratio, while the reverse was noted for 2 d-old eggs. It has been well established that the female ratio of some species of Trichogramma can be significantly altered at different temperatures, resulting in different parasitism and longevity rates [53]. In contrast with the female ratio, adult body length was decreased with the increase of the exposure interval in low temperatures. "Better eggs", in terms of size and nutrients, are known to provide larger parasitoids, but the size of parasitoids may not necessarily result in a better parasitism performance [54].

Conclusions
Our work demonstrated that storage of eggs for a certain internal can be used with success for rearing protocols of T. evanescens. Between the two hosts used here, G. melonella was more suitable for parasitoid rearing, mainly due to the fact that this species provides eggs that are more suitable for parasitism when exposed at low temperatures. Maintaining moth eggs at low temperatures has been proposed as a means to control parasitoid rearing, and to provide large numbers of parasitoids whenever it is needed, utilizing both moth species. Having alternative host species in egg parasitoid rearings can prevent unexpected turnovers when there are problems with the rearing of the basic host. In this effort, 2 d-old eggs can be more suitable than 1 d-old eggs, and this age could be proposed for selection to increase parasitoid performance, especially when maintained at 9˚C.
Rajshahi University for extending the laboratory facilities. We also thank the two anonymous reviewers that provided useful comments on the original manuscript.