How long-term mass rearing affects the quality of the Trichogramma embryophagum (Hartig) (Hymenoptera: Trichogrammatidae) reared on Sitotroga cerealella (Olivier) eggs

Maintaining the quality and efficacy of biological control agents during long-term mass rearing plays a crucial role in the success of a biological control program. In this study, the biological traits of a local population colony of Trichogramma embryophagum (Hartig) with rearing on Sitotroga cerealella (Olivier) eggs for over 40 generations (G) were assessed. Female adult longevity was significantly different among generations, ranging from 7.98 d in G5 to 5.19 d in G40. The reared wasps showed highest fecundity (60.50 eggs/female) in G5 compared to the other generations. The female sex ratio varied from 63.16% in G5 to 49.31% in G40. Significant differences were observed in population growth parameters and the highest gross reproductive rate (GRR) (40.96 eggs/individual) and net reproductive rate (R0) (38.21 eggs/individual) were found in G5. However, a non-significant difference was found in the intrinsic rate of natural increase (r) until the 10th generation, but its values significantly declined with increasing the generation numbers. The finite parasitism rate (ω) ranged from 0.468 host/parasitoid/day in G5 to 0.274 host/parasitoid/day in G40. The results showed that the quality of T. embryophagum reared under continuous laboratory conditions declined after 10 generations, and for use of them in biological control programs under field conditions, the reared population should be refreshed by adding wild individuals from time to time.


Background
Among different recommended methods to control the pests in integrated pest management (IPM), biological control with the use of natural enemies has special importance due to its safety, selectivity and as a result producing healthy products. Trichogrammatid' wasps are one of the main groups of the insect's egg parasitoids that are widely used all over the world. The different species of genus Trichogramma has been used successfully as egg parasitoids to control more than 30 key pest species in inundative or augmentative biological control programs (Vinson et al. 2015). Mass rearing and release of these wasps leads to a significant decrease in damage caused by the larval stage of the pest.
Augmentative or inundative biological control programs rely on mass production and releases of natural enemies and in continuous rearing the quality of the produced individuals has much importance (van Lenteren 2003). The process of Trichogramma mass rearing often continues for a long time with many generations; therefore the quality of reared wasps in this prolonged process was also affected by reared conditions (Smith 1996). One of the concerns in the mass rearing process is that they are produced under ambient conditions and in limited space may adapt to laboratory conditions and lose some characters during long-term maintenance and experience inbreeding depression, resulting in poor performance of the released wasp in the field (Badran et al. 2020). Therefore, they may be unable to find and parasite their hosts under variable field conditions; therefore it is very important to study the changes that may occur during mass rearing of natural enemies.
Quality control in mass rearing of the parasitoids is one of the important measures that was used to prevent defeats in biological control programs (van Lenteren 2003). In quality control process some traits including parasitism rate, sex ratio, fecundity, longevity, body size and weight and flight activity or host-searching ability have to be evaluated (Lu et al. 2017). Reduction in the fecundity, survival and adult longevity of some trichogrammatid' wasps reared for several generations have been reported (Ghaemmaghami et al. 2021a,b).
Trichogramma brassicae (Bezdenko), T. embryophagum (Hartig) and T. pintoi Voegele are the most widespread species of the Trichogrammatidae family in Iran (Ebrahimi et al. 1998). T. embryophagum has a wide host range. This species is annually reared over more than 15-20 generations mainly to control the Ectomyelois ceratoniae Zeller in some regions of Iran.
Although several studies were accomplished on different aspects of T. embryophagum, no information was available on its quality during long-term mass rearing. Therefore, the effects of long-term mass rearing on the quality of T. embryophagum reared on Sitotroga cerealella (Olivier) over 40 generations were assessed.

Insect host rearing
This study was conducted during the years of 2018 to 2020. In order to establish the colony of S. cerealella, its eggs were obtained from an insectary, in the Agriculture and Natural Resources Research Institute of Hamadan Province, Hamadan, Iran. The stock culture of S. cerealella was initiated under constant temperature of 26 ± 1 °C in the Entomological laboratory of the Agriculture and Natural Resources Research Institute of Hamadan Province, Hamadan, Iran. The prepared S. cerealella eggs were reared in cylindrical containers (19 cm diameter and 8 cm height) on barley grains until the end of pupal stage. The newly emerged adults were transferred to funnels (diameter 150 mm) covered with a fine mesh net at top and kept up-side-down. After mating, the moth eggs were collected on pieces of paper placed as egg laying substrate to ease egg collection under the wider side of the funnel.

Parasitoid colony
To obtain the initial colony of T. embryophagum, the egg traps [S. cerealella eggs placed on pieces of white papers (200 × 100 mm)] were placed at a height of about 2 m above the ground on the leaves of pomegranate trees, in Tange-Siab, Kuhdasht County, Lorestan Province, Iran, in May 2018. The traps were collected after 24 h and transferred to the laboratory. The parasitized eggs were reared in a growth chamber set at 26 ± 1 °C, 60% ± 5% RH and a 16: 8 h (L: D) photoperiod until the emergence of T. embryophagum adults, which were then reared in rectangular plastic containers (300 × 200 × 150 mm) on S. cerealella eggs under the same conditions as mentioned above. The collection process was repeated several times in a limited time to supply a sufficient amount of parasitoids.

Experimental design
After emergence of the adult wasps and oviposition, more than 100 newly parasitized eggs of hosts (less than one day old) were selected from the colony and kept in a glass container (100 mm in diameter, 60 mm in height) until the adult wasps emerged. After the emergence of the adults, each pair of male and female wasps was placed in a glass cylinder (100 × 16 mm) containing about 120-150 eggs (less than one day old host glued on a piece of paper 10 × 50 mm). The parasitoid wasps were regularly supplied by diluted honey on cotton rolls after emergence. Dead males during the experiments were replaced by newly emerged ones (< 24 h old) from the stock colony. These individuals were excluded from the statistical analyses. Exposed egg papers were collected after 24 h and replaced daily with refresh ones until the death of the last females. The exposed eggs were separately maintained in a growth chamber at 26 ± 1 °C, RH of 65 ± 5% and a photoperiod of 16L: 8D hrs. The time interval between egg exposure and the time of emergence of adults was considered as the duration of the immature stages development. The glass cylinders were monitored and longevity, mortality and survivorship of adult females and males were recorded by visual examination every 12 h until the death of the last individual. Fecundity of each parasitoid wasp was also calculated by counting and recording the parasitized eggs of the host daily (as evidenced by blackening) over its life span. Sex ratio was estimated as a percentage of females. Total life span of females was calculated for those wasps that died as female adults and presented as a total duration of immature stages and adult longevity for females. These procedures were repeated at the generations of 5, 10,15,20,25,28,30,32,34,37 and 40 of the T. embryophagum colony. The Page 3 of 12 Sadat et al. Egypt J Biol Pest Control (2021) 31:119 counts of the studied generations were 5 generations until the 25 th one. By increasing the number of generations and the effects of long-term rearing on their characteristics, the counts among the studied generations decreased to 2 or 3 generations. All experiments were carried out in an incubator set at 26 ± 1 °C, 60% ± 5% RH and a photoperiod of 16: 8 h (L: D).

Statistical analysis
The life history data were analyzed using age-stage, twosex life table theory (Chi 1988). The age-stage survival rate (s xj ) (the probability that a newly laid egg will survive to age x and stage j), age-stage-specific fecundity of female (f xj ) (the number of offspring produced by female adult of age x and stage j), the age-specific survivorship (l x ) (the probability that a newly laid egg would survive to age x), age-specific fecundity (m x ) (the mean number of eggs produced per individual at age x) and also the age-stage life expectancy (E xj ), as well as the population parameters, include the intrinsic rate of natural increase (r), the finite rate of increase (λ), the gross reproductive rate (GRR ), the net reproductive rate (R 0 ) and the mean generation time (T) were calculated by TWOSEX-MSChart program (Chi 2019b). Bootstrap procedure was used to estimate the variances and the standard errors of population growth parameters (Huang and Chi 2013).
To obtain the stable estimates, 10,000 bootstrap samples were used. Bootstrap values of different generations of T. embryophagum were then compared with paired-bootstrap procedure (Bahirai et al. 2019). Daily parasitism rates of the cohort were used to estimate the parasitism parameters for each generation. The age-specific parasitism rate (k x ) is the mean number of host parasitoid at age x that was calculated by the following equation (Chi and Yang 2003): where β is the number of stages. The net parasitism rate (c 0 ) shows the mean number of hosts parasitized by an individual parasitoid during its entire life span, and was calculated as (Chi and Yang 2003): The transformation rate from host population to parasitoid offspring (Q p ) that shows the number of hosts that is needed for a parasitoid to produce an offspring (Chi et al. 2011) is calculated as: The stable parasitism rate (ψ) is the total parasitism capacity of a stable population which total size is unity (Chi et al. 2011), and is calculated as: a xj is the proportion of individuals belonging to age x and stage j in stable age-stage distribution (SASD). The finite parasitism rate (ω) shows the parasitism potential of a parasitoid population by combining its finite rate of increase (λ), age-stage parasitism rate (c xj ) and stable agestage structure (a xj ) (Chi et al. 2011;Yu et al. 2013), and is calculated as: Parasitism rate data were analyzed using the computer program CONSUME-MSChart (Chi 2019a). The bootstrap resampling method (10,000 bootstraps) was used to estimate the variances and the standard errors of the parasitism parameters. Comparison of parasitism parameters was done based on paired-bootstrap test using TWOSEX-MSChart program (Ghaemmaghami et al. 2021a, b). The parasitism capacity was compared among the generations 5, 10,15,20,25,28,30,32,34,37 and 40 of T. embryophagum. Microsoft Excel 2013 was used to draw the graphs.

Adult longevity, reproductive parameters and survival
Although the duration of the immature stages of T. embryophagum varied slightly over successive generations, it was non-significantly affected by the increase in generation numbers. It was between 10 and 11 days. Unlike, significant differences in female and male longevity were observed among generations. The longest female longevity (7.98 ± 0.36 d) was estimated in G5, while in G40 showed the shortest female longevity (5.19 ± 0.40 d). The longest male longevity was estimated in G5 (6.54 ± 0.40 d), while the shortest was in G32 (4.22 ± 0.39 d) and G40 (4.22 ± 0.33 d). The longest total life span of females was counted in G5 (18.40 ± 0.38 d) then decreased over the generations. The shortest female total life span (15.35 ± 0.42 d) was in G40 (Table 1). However, a non-significant difference was observed in total life span until the 15thgeneration. Ovipositional days differed significantly among generations, ranging from 5.83 ± 0.19 d in G5 to 3.53 ± 0.16 d in G40 (Table 1). Non-significant differences were observed between the ovipositional days of the 5th and 10th generations. Fecundity varied from 60.50 ± 2.35 eggs/female in G5 to 19.24 ± 1.88 eggs/ female in G40. As shown in Table 1, the fecundity reduced to less than half after the 25th generation. Also, different generations showed significant variations in sex ratio, ranging from 63.16% females in G5 to 49.31% females in G40. According to the results after G30, the sex ratio of females was less than 55% (Table 1).

Population growth parameters
As shown in Table 2, all population growth parameters of T. embryophagum on S. cerealella were affected by sequential generations. The highest values of GRR (40.96 ± 3.71 eggs/individual), R 0 (38.21 ± 3.65 eggs/ individual), r (0.296 ± 0.008 d −1 ) and λ (1.345 ± 0.011 d −1 ) were found in G5. By increasing generation numbers from G5 to G40, the values of GRR, R 0 , r and λ decreased and their lowest values were obtained in G40. However, non-significant differences were observed among the values of GRR , R 0 , r and λ in the 5th and 10th generations. The longest and shortest values of T were recorded in G10 with 12.63 ± 0.13d and 11.63 ± 0.15 d in G37, respectively.

Table 1 Duration of adult longevity and total life span (d), sex ratio (female ratio) and fecundity (eggs) of sequential generations of Trichogramma embryophagum reared on Sitotroga cerealella eggs
The means marked by the same letters within the same column are not significantly different (Paired-bootstrap test, P < 0.05)

Discussion
Maintaining the quality of natural enemies that reared for many generations is one of the most important problems which should be considered, because continuous mass rearing in insectaries with stable conditions may decrease their performance under field conditions (Bertin et al. 2017). Although based on the information obtained from this study, T. embryophagum was able to complete its life cycle on S. cerealella eggs over 40 generations, but its life table parameters and parasitism performance significantly affected over generations. As a result of this work, the immature developmental time of T. embryophagum was not influenced by increasing the generation number and varied from 10 to 11 days. In line with these results, the duration of immature stage of T. brassicae reared on S. cerealella eggs over 45 generations was not affected by increasing generation numbers (Ghaemmaghami et al. 2021a). The developmental times of T. embryophagum on Plutella xylostella (Akbari et al. 2012) and on E. ceratoniae eggs (Mohseni et al. 2016) were 10 d, which are close to the present findings.
Different generations of T. embryophagum showed significant variations in adult longevity. Obtained results were in agreement with those of Ghaemmaghami et al. (2021a, b), who reported a significant decrease in female's adult longevity of T. brassicae reared on S. cerealella and E. kuehniella Zeller over 45 generations. Also, similar results reported for Trichogramma dendrolimi Matsumura over 30 generations (Lu et al. 2017). Ranjbar Aghdam and Attaran (2015) reported 7.15 d for adult longevity of T. embryophagum on S. cerealella, close to the obtained in G5 (7.98 d). But Haghani (2001) reported 12.37 and 11.29 d as adult longevity for T. embryophagum on E. kuehniella and S. cerealella, respectively, which are greater than our data.
In present study, a significant difference in ovipositional days was observed among sequential generations and the longest ovipositional days were observed in G5 with 5.83 d and with increasing the generation number, its value declined reaching 3.53 d in G40. However, a non-significant difference was observed between the ovipositional days in the 5th and 10th generations. On the contrary, Ghaemmaghami et al. (2021a) reported that the ovipositional days of T. brassicae reared on S. cerallella eggs were non-significantly different among generations until G35. Mohseni et al. (2016) reported 8.20 d for ovipositional days of T. embryophagum on E. ceratoniae eggs which was longer than those obtained in this study. These differences can be due to the different host species.
Fecundity is a very important criterion for assessing the quality of reared parasitoids, because it determines the cost of mass production and can be influenced during long-term mass rearing. In this work, the fecundity of T. embryophagum decreased over 40 generations, which is similar to those reported for T. brassicae reared on S. cerealella and E. kuehniella (Ghaemmaghami et al. 2021a,b). The values of fecundity obtained in the different generations in this study were less than those obtained by Mohseni et al. (2016) for this parasitoid on E. ceratoniae (74.40 eggs/female). Different factors may contribute to such a discrepancy such as the used population of parasitoids, host species and some reared conditions.
The sex ratio in mass rearing and inundative release of parasitoid wasps is an important factor (Ebrahimi 1996) that influencing the economic profitability of  . 4 The age-stage parasitism rate (c xj ) of sequential generations of T. embryophagum reared on S. cerealella eggs Page 10 of 12 Sadat et al. Egypt J Biol Pest Control (2021) 31:119 mass rearing (Badran et al. 2020). In accordance with IOBC guidelines, that sex ratio of Trichogramma species in mass production programs should be higher than 50% (van Lenteren et al. 2002). In the present study, the ratio of females was greater than 50% female in all generations, except in G34 and G40. But contrary to those reported for some species of Trichogramma (Nordlund et al. 1997;Pratissoli et al. 2004;Ghaemmaghami et al. 2021a, b), current study shows that sex ratio percentage of T. embryophagum is affected by rearing generations. Intrinsic rate of natural increase (r) along with net reproductive rate (R 0 ) is a key parameter in life table studies (Bahirai et al. 2019). R 0 is defined as the average number of female offspring that would be born by each female during its life span. The value of R 0 decreased from 38.21 eggs/individual in G5 to 9.75 eggs/individual in G40. R 0 in this study in G5 (38.21 eggs/individual) was close to those reported for the parasitoid on S. cerealella by Haghani and Fathipour (2003) (37.63 eggs/ individual) and on E. cerotoniae (37.20 eggs/individual) by Mohseni et al. (2016).
Also the mean generation time (T) is defined as the time required for a newborn female to replace herself by R 0 -folds at the stable age-stage distribution. Although T values in the present study differed among generations, they did not show a regular trend. Accordingly, each female required 12.63 d to be replaced by 31.32 females in G10, but in G37 each female was replaced by 10.65 females in 11.63 d. According to these results, the profitability of producers was greatly reduced due to the role of females in reproduction in recent generations. Haghani and Fathipour (2003) reported 16.49 and 16.37 d for mean generation time (T) of T. embryophagum on S. cerealella and E. kuehniella, respectively, that are higher than those obtained in our study. According to Mohseni et al. (2016), the mean generation time (T) of this parasitoid on E. cerotoniae was 11.64 d that is close to the obtained finding.
Intrinsic rate of natural increase (r) combined the effects of mortality, duration of life stage and fecundity; therefore, it is influenced by preimaginal survival, developmental time, longevity of females, fecundity value and sex ratio, which are all affected in long-term mass rearing. In the present study, the intrinsic rate of natural increase (r) of T. embryophagum decreased from 0.296 d −1 in G5 to 0.192 d −1 in G40. Also, in accordance with those reported for T. brassicae reared on S. cerealella, we found no significant difference between the r values in the 5 th and 10 th generations. Similar to the present study, Ghaemmaghami et al. (2021a, b) showed the r value of T. brassicae reared on S. cerealella and E. kuehniella eggs decreased from 0.250 d −1 and 0.242 d −1 in G5 to 0.167 d −1 and 0.174 d −1 in G45. Haghani and Fathipour (2003) reported 0.218 d −1 as r for this parasitoid reared on S. cerealella. Also Mohseni et al. (2016) reported that the value of intrinsic rate of natural increase was 0.311 d −1 that is higher than the obtained results.
Finite rate of increase (λ) also varied among successive generations. The results showed that the population of T. embryophagum in the 5 th generation increased 1.345 times daily than the previous day, while in the 40th generation, its population increased by 1.212 times every day than the previous day. Haghani (2001) reported that the finite rate of increase for this species on S. cerealella was 1.244 (day −1 ) which is close to our finding in 30th generation.
In the present study, numbers of hosts parasitized by each female (c 0 ) decreased with increasing generations,