Thelytokous Diglyphus wani: A more promising biological control agent against agromyzid leafminers than its arrhenotokous counterpart

DU Su-jie, YE Fu-yu, XU Shi-yun, , WAN Wei-jie, GUO Jian-yang#, YANG Nian-wan, , LIU Wan-xue#

1 State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing 100193, P.R.China

2 College of Life Sciences, Hunan Normal University, Changsha 410081, P.R.China

3 Western Agricultural Research Center, Chinese Academy of Agricultural Sciences, Changji 831100, P.R.China

Abstract Diglyphus wani (Hymenoptera: Eulophidae) is a dominant parasitoid that attacks agromyzid leafminers.Two reproductive types occur in D.wani: arrhenotoky (in which virgin females produce only male offspring; and virgin females mate with males to produce bisexual offspring) and thelytoky (in which virgin females produce female offspring).As a potential biological control agent, exploring the differences in the relevant biological parameters of both strains is necessary.In this study, comparisons between the two strains of D.wani were performed by evaluating the life table and host-killing rate.The thelytokous strain exhibited significantly better life table parameters than its arrhenotokous counterpart.Higher values for the intrinsic rate of increase, finite rate of increase, net reproductive rate, and fecundity were found in the thelytokous strain.The thelytokous strain also performed better than the arrhenotokous strain in terms of net parasitism, host-feeding, host-stinging, and total host-killing rates.Thus, populations of the thelytokous strain could grow fast and kill more hosts.In conclusion, the thelytokous strain of D.wani may be the more promising biological agent against agromyzid leafminers compared to its arrhenotokous counterpart.Also, since the thelytokous strain of D.wani is only known to produce females, it should be given priority in future biocontrol applications owing to the cost savings of breeding only females.

Keywords: parasitoid, life table, arrhenotoky, thelytoky, biocontrol applications

1.Introduction

Agromyzid leafminers comprise more than 3 000 species worldwide.Most species, including the commonLiriomyzaandPhytomyzaspecies, are polyphagous pests of crops and ornamental plants (Spencer 1973).These leafminers can feed on plant mesophyll tissue inside the leaves during the entire larval period (Spencer 1973).Extensive mining can result in premature leaf withering and dropping, resulting in significant economic losses (Parrella 1987).According to Prattet al.(2017),the average annual economic losses to smallholder agricultural production from invasiveLiriomyzaspecies in three African countries (Kenya, Tanzania, and Uganda)was estimated to be US$125.2-149.1 million.In recent years, the invasion and expansion ofLiriomyzaspecies(e.g.,L.huidobrensisandL.trifolii) have been rapid.Some countries and regions (e.g., Australia and Korea)have been newly invaded only recently (Maharjanet al.2014; Blacketet al.2015).Although several pest management approaches have been explored for agromyzid leafminers, the use of insecticides remains the main method for controlling them worldwide (Kanget al.2009; Gitongaet al.2010).However, the long-term side effects of chemical insecticides, such as the development of insecticide resistance in the pests and lethality to the natural enemies, have been reported (Hernándezet al.2011).By comparison, the application of parasitoids to control agromyzid leafminers is considered a better option because of their high population abundance and parasitism rates in the field (Ridlandet al.2020).

In field control applications of parasitoids, females often play the dominant role in pest control.Therefore, obtaining adequate females should be a major consideration during rearing or future biological control applications.The majority of hymenopteran parasitoids are arrhenotokous,whereby fertilized eggs develop as diploid females, and unfertilized eggs develop as haploid males (Heimpel and de Boer 2008).However, some species are thelytokous,whereby unfertilized eggs develop as diploid females and all offspring generally consist of females.There are over 500 thelytokous species of Hymenoptera (van der Kooiet al.2017), the majority of which are induced by bacteria(e.g.,Wolbachia,Rickettsia, andCardinium).Nevertheless,the reproductive stability of thelytokous parasitoids is usually influenced by bacterial titer, and titer decrease may lead to male production (Stouthameret al.1990).Thus,multiple releases may be required to supplement female wasps in practical applications.In addition, the bacteria may not only induce female production in thelytokous species, but they can also influence the fitness and competitive capacity of the species (Tagamiet al.2001;Huigenset al.2004).For example,Wolbachiareduces the immature survival rate, numbers of progeny and mature eggs, and capacities for assessing fresh and old hosts in thelytokous parasitoids (Hohmannet al.2001; Tagamiet al.2001; Huigenset al.2004; Farahaniet al.2015).

In comparison, genetic thelytokous parasitoids(e.g.,Venturiacanescens) are regarded as stable female producers and undergo non-bacterial induced reproduction (Schneideret al.2001).Therefore, they may show wider and more sustainable biocontrol application prospects.Thelytoky without bacterial induction and arrhenotoky has been found in some conspecific parasitoids (Beukeboom and Pijnacker 2000; Tsutsuiet al.2014; Duet al.2022).In this context, the potential advantages of the thelytokous and the corresponding arrhenotokous strains should be compared completely.However, at present, few studies focus on the comparison of thelytokous strains without bacteria-induced reproduction and their arrhenotokous counterparts.In addition, many studies have focused on the life history traits of the two strains of parasitoid species, especially adult longevity and fecundity.There are still only a few comparative studies that have combined the life table parameters and host-killing rates of the two strains of parasitoids to accurately assess their relative biological control performance (Yeet al.2023).

Diglyphuswaniis a synovigenic, idiobiont, and ectoparasitic wasp that parasitizes larvae of agromyzid leafminer species (Fig.1), particularlyLiriomyzaandPhytomyzaspecies (Yeet al.2018; Duet al.2021).Diglyphuswanihas both thelytokous and arrhenotokous strains (Duet al.2021, 2022), and shows three types of host-killing behaviors: reproductive parasitization(parasitism) and two non-reproductive host killing behaviors (host feeding and host stinging) (Fig.1).In addition, thelytoky ofD.waniis not induced byWolbachia,Rickettsia, orCardinium(Duet al.2022).Moreover, no males were found during nearly six years of rearing in the laboratory, suggesting that this species stably produces only females.As the dominant natural enemy of the field agromyzid leafminer, whether there are differences in biological characteristics between the two strains ofD.waniremains unclear.Exploring and assessing these differences between the two strains ofD.waniwould provide strong support for improving future classical and augmentative biological control practices.

Life tables are widely adopted to comprehensively describe the survival, development, host-killing rate, and fecundity of parasitoid species (Chiet al.2020).In addition,the host-killing rate has been applied to studies on the control capacity of parasitoids with reproductive and nonreproductive host-killing behaviors (Yeet al.2018, 2023;Ramos Aguilaet al.2021; Zhaoet al.2021).In this study,we collected the age-stage, two-sex life table data, as well as reproductive (parasitism) and non-reproductive (host feeding and host stinging) host-killing rates to investigate the differences in population parameters between the two strains ofD.wani.Our results will contribute to a better understanding of mass rearing techniques and the potential application of the two strains ofD.waniin controlling agromyzid leafminers.Furthermore, the results will lay a foundation for the comparative life table study of nonbacterially-induced thelytokous strains and the corresponding arrhenotokous strains of conspecific parasitoids.

2.Materials and methods

2.1.Insect cultures

Thelytokous and arrhenotokousD.waniwere originally collected from Xining, Qinghai, China (36°39´N,101°45´E) in 2015 and Kunming, Yunnan, China(24°53´N, 102°47´E) in 2018, respectively.The colonies were reared on kidney bean (Phaseolusvulgaris)leaves infested withL.sativaelarvae, and were maintained in climate chambers (at (26±1)°C, (50±5)%RH, and L:D=14 h:10 h) in the laboratory.The colony ofL.sativaewas obtained in Langfang, Hebei, China(39°35´N, 116°47´E).The populations of leafminers and parasitoids have been maintained in gauze cages prior to and during the study.

2.2.Developmental durations of immature stages

Fifty thelytokous females, 50 arrhenotokous females, and 50 males were collected within 2 h of emergence.To promote ovarian maturation inD.wani, the parasitoids were placed in round glass Petri dishes (90 mm in diameter and 18 mm in height) and fed for 2 days with cotton balls containing 20% honey solution.Meanwhile,we mixed females and males of the arrhenotokous strain to allow them to fully mate.

Then, a group of 50 arrhenotokous females and a separate group of 50 thelytokous females were placed in the new glass Petri dishes.The glass Petri dishes contained approximately 35-40 hosts which were at the end of the second to the beginning of the third instar of 3-4 days old, and the parasitoids were allowed to parasitize the hosts for 1 h.After that, the parasitoid wasps were removed.The Petri dishes with the infested hosts on the leaves were transferred to a climatic chamber (at (26±1)°C, (50±5)% RH, and L:D=14 h:10 h).The development of eggs, larvae, prepupae, and pupae was monitored under a stereomicroscope (Olympus Corporation, SZX-16, Tokyo, Japan) and recorded at 6 h intervals (8:00, 14:00, 20:00, and 2:00).

2.3.Records of parasitoid adult data

Emerged parasitoid adults were transferred to a wellventilated cage designed by Zhanget al.(2022) (Fig.2).For the thelytokous strain, one female was introduced per cage; for the arrhenotokous strain, one female and one male were introduced per cage.A leaf containing 35-40L.sativaehosts was supplied to the parasitoids.A cup(450 mL) with tap water was placed in the cage to avoid desiccation.The cages were then transferred to a 26°C artificial climate chamber.Leaves were replaced daily.The leaves from the previous day were kept in a glass Petri dish containing 10 mL of 1.2% agar gel.After 48 h,the data on the fecundity (the number of eggs produced by females), parasitism, host-feeding events, and hoststinging events were recorded.The criteria for identifying the three host-killing behaviors described by Zhanget al.(2014), Yeet al.(2018), and Xuanet al.(2022) were followed in this study.Moreover, longevity was checked dailyuntil the parasitoids died.

Fig.2 The flow chart for the observations of the life history and life table parameters of Diglyphus wani in the immature and adult stages.TH, thelytokous strain; AR, arrhenotokous strain.

2.4.Life table analysis

The raw data for the immature period and female daily fecundity of the two strains ofD.waniindividuals were analyzed based on the age-stage, two-sex life table using the TWOSEX-MSChart Program (Chi and Liu 1985; Chi 1988, 2022c; Chiet al.2020).The age-stage specific survival rate (sxj) (wheresxjis the probability of a newly laid egg surviving to agexand stagej) and age-stage specific fecundity (fxj) (i.e., the number of eggs that will be laid by a female individual at agexand stagej) were obtained according to the method of Chi and Liu (1985).The agespecific survival rate (lx) and age-specific fecundity (mx)were calculated as follows:

whereβis the number of stages.

The net reproductive rate (R0), which is the mean number of offspring that an individual can produce throughout its lifespan, was calculated as follows:

The intrinsic rate of increase (r) was estimated from the Euler-Lotka formula using the iterative bisection method with the age indexed from 0 (Goodman 1982), following the formula:

The finite rate of increase (λ) was calculated as follows:

The mean generation time (T) is the amount of time required for a population to increase toR0-fold its size when the population is in the stable age-stage distribution.It was calculated as follows:

2.5.Host-killing rate analysis

The CONSUME-MSChart Program (Chi 2022a) was used to analyze the host killing rate.For the thelytokous and arrhenotokous strains ofD.wani, females were found to have three different host-killing behaviors: parasitism, host feeding, and host stinging.To assess the differences in these three host-killing behaviors between the two strains,we calculated the age-specific parasitism rate (kx,p), hostfeeding rate (kx,f), and host-stinging rate (fx,s) according to Chi and Yang (2003), Yuet al.(2005), Zhaoet al.(2021),and Yeet al.(2018, 2023), using eq.(7):

wherecxjis replaced with either the age-stage parasitism rate (cxj,p), host-feeding rate (cxj,f), or host-stinging rate(cxj,s), for the calculations of the age-specific parasitism rate (kx,p), host-feeding rate (kx,f), or host-stinging rate(kx,s), respectively.

By incorporating the age-specific survival rate (lx), the age-specific net parasitism rate (qx,p), net feeding rate(qx,f), and net stinging rate (qx,s) were calculated as follows:

wherekxis replaced withkx,p,kx,f, orkx,s, for the calculation ofqx,p,qx,f, orqx,s, respectively.The cumulative parasitism rate (Cx,p) cumulative feeding rate (Cx,f), and cumulative stinging rate (Cx,s) were calculated as follows:

wherekxis replaced withkx,p,kx,f, orkx,s, for the calculation ofCx,P,Cx,f, orCx,s, respectively.

The total numbers of hosts parasitized, fed upon, and stung by an average parasitoid during its lifespan are defined as the net parasitism rate (C0,p), net feeding rate(C0,f), and net stinging rate (C0,s), respectively.They were calculated as follows:

The finite parasitism rate (ωp), finite feeding rate (ωf),and finite stinging rate (ωs) were calculated according to Chiet al.(2011) and Yuet al.(2013) as:

whereψis the stable parasitism rate (ψp), stable feeding rate (ψf), or stable stinging rate (ψs), andaxjis the proportion of individuals belonging to agexand stagejin a stable age-stage distribution.

To reflect and compare the overall host-killing capacity of the two strains, we also calculated the total host killing rate.In the present study, the number of total host-killing events was calculated as the sum of the numbers of parasitism,host-feeding, and host-stinging events.Therefore, the agestage specific total host-killing rate of individuals at agexand stagej(pxj) based on the descriptions of Yeet al.(2018,2023) and Zhaoet al.(2021) was calculated as follows:

The age-specific total host-killing rate (ux) followed the formula described by Chi and Yang (2003) and Yuet al.(2005):

By incorporating the age-specific survival rate (lx), the age-specific net total host killing rate (wx) was calculated as follows:

The cumulative total host-killing rate (Zx), net total host-killing rate (Z0), stable host-killing rate (ψ), and finite host-killing rate (ω) were calculated using the following formulas, as described by Yeet al.(2018, 2023):

The total transformation rate (Qp), which is the number of hosts needed for a parasitoid to produce a single individual, was calculated according to Chi and Yang(2003), Zhaoet al.(2021), and Yeet al.(2023):

2.6.Population projection

The population growth and host-killing capacity of the two strains ofD.waniwere projected using the TIMINGMSChart Program (Chi 2022b).Since the parasitoids are usually stored and released by the pupae of parasitoids,the population projections ofD.waniwere started with 10 pupae at 10-day age.

2.7.Data analyses

The standard errors (SE) of the life table and host-killing parameters were estimated using 100 000 bootstraps samples in the TWOSEX-MSChart, CONSUME-MSChart,and TIMING-MSChart programs (Chi 2022a, b, c).All statistics were compared using a paired bootstrap test.

3.Results

3.1.Biological characteristics

The durations of the egg, larval, preadult, and female adult stages, and total female longevity between the two strains were not significantly different, whereas the durations of the prepupal and pupal stages were significantly different between the two strains (Table 1).Thelytokous females showed shorter prepupal stage (1.1 daysvs.1.3 days)and longer pupal stage durations (5.0 daysvs.4.7 days)than the arrhenotokous strain.In addition, the longevity of male adults (5.0 days) and total duration of males (17.6 days) were significantly shorter than that in females of the two strains.

Table 1 Life history parameters (mean±SE) of the thelytokous and arrhenotokous strains of Diglyphus wani

The total pre-oviposition period was shorter in thelytokous females (13.0 days) than in arrhenotokous females (13.7 days).Meanwhile, thelytokous femalesexhibited higher fecundity (87.4 eggs/female) than arrhenotokous females (57.8 eggs/female).However,there were no significant differences in the adult preoviposition and oviposition periods between thelytokous and arrhenotokous females.

3.2.Population parameters

The intrinsic rate of increase (r) (0.2371 d-1), finite rate of increase (λ) (1.2676 d-1), and net reproductive rate(R0) (78.6 offspring) of the thelytokous strain reared onL.sativaewere significantly higher than those of the arrhenotokous strain (r=0.1659 d-1,λ=1.1805 d-1,R0=27.1 offspring; Table 2).By comparison, the thelytokous strain had a significantly shorter mean generation time(T=18.4 d) than the arrhenotokous strain (T=19.8 d).

Table 2 Means of the population parameters in the thelytokous and arrhenotokous strains of Diglyphus wani

3.3.Parasitism rate

Fig.3 shows the parasitism rate curves for the two strains ofD.waniobserved in this study.Gaps in the parasitism rate curves were found because at any immature stage,the parasitoids are incapable of killing hosts for either of the two strains ofD.wani.The daily age-specific parasitism rate (kx,p) of thelytokous females was higher than that of the arrhenotokous strain from 10- to 23-day age (Fig.3).The daily age-specific net parasitism rate(qx,p) and cumulative parasitism rate (Cx,p) of thelytokous females were higher than those of the arrhenotokous strain at various ages (Fig.3).The peaks of thekx,pandqx,pvalues of the thelytokous strain were 9.8 hosts at 19-day age and 8.8 hosts at 19-day age, respectively,which were higher and appeared earlier than those of the arrhenotokous strain (5.5 hosts at 21-day age and 2.2 hosts at 21-day age, respectively) (Fig.3).

Fig.3 Cumulative parasitism rates (Cx,p), age-specific parasitism rates (kx,p), and age-specific net parasitism rates(qx,p) of the thelytokous and arrhenotokous strains of Diglyphus wani.TH, thelytokous strain; AR, arrhenotokous strain.The gaps represent that parasitism rate is all zero before the adult stage, because the parasitoids are incapable of killing hosts for either of the two strains of D.wani.

The net parasitism rate (C0,p) of the thelytokous strain(78.6 hosts/parasitoid) was ~2.91 times greater thanthat of the arrhenotokous strain (27.0 hosts/parasitoid)(Table 3).Additionally, the stable parasitism rate (ψp)(0.2725) and finite parasitism rate (ωp) of the thelytokous strain (0.3458) were significantly higher than those of the arrhenotokous strain (0.1912 and 0.2260, respectively).

Table 3 Means of the parasitism, host-feeding, host-stinging,and host-killing rates of the thelytokous and arrhenotokous strains of Diglyphus wani in the adult stage

3.4.Host-feeding rate

The daily age-specific host feeding rate (kx,f), daily agespecific net parasitism rate (qx,f), and daily cumulative parasitism rate (Cx,f) of thelytokous females were generally higher than those of the arrhenotokous strain throughout their lifespan (Fig.4).The maximumkx,fandqx,fvalues of the thelytokous strain were 13.7 hosts at day 20 and 11.9 hosts at day 20, respectively, while these parameters were 9.6 hosts at day 21 and 4.6 hosts at day 17, respectively, for the arrhenotokous strain.

Fig.4 Cumulative host-feeding rates (Cx,f), age-specific hostfeeding rates (kx,f), and age-specific net host-feeding rates(qx, f) of the thelytokous and arrhenotokous strains of Diglyphus wani.TH, thelytokous strain; AR, arrhenotokous strain.The gaps represent that parasitism rate is all zero before the adult stage, because the parasitoids are incapable of killing hosts for either of the two strains of D.wani.

The net host-feeding rate (C0,f) of thelytokous females(115.7 hosts/parasitoid) was ~2.02 times larger than that of its arrhenotokous counterpart (52.3 hosts/parasitoid).The stable host-feeding rate (ψf) and finite host-feeding rate (ωf) of the thelytokous strain were not significantly different from those of their arrhenotokous counterparts.

3.5.Host-stinging rate

The host-stinging rate curves of parasitoids on each day of their life are plotted in Fig.5.The highest dailykx,svalue (1.5 hosts) of the thelytokous strain occurred on day 16, which was a greater value and occurred earlier than that of its arrhenotokous counterpart (1.1 hosts on day 26).The daily age-specific host-stinging rate (qx,s) of the thelytokous strain reached its peak (qx,s=1.3) on day 16,whereas the arrhenotokous strain reached its highestqx,svalue on day 17 with 0.4 hosts.

Fig.5 Cumulative host-stinging rates (Cx,s), age-specific hoststinging rates (kx,s), and age-specific net host-stinging rates(qx, s) of the thelytokous and arrhenotokous strains of Diglyphus wani.TH, thelytokous strain; AR, arrhenotokous strain.The gaps represented that parasitism rate is all zero before the adult stage, because the parasitoids are incapable of killing hosts for either of the two strains of D.wani.

The net host-stinging rate (C0,s) (9.8 hosts/parasitoid)of the thelytokous strain was significantly higher than that of its arrhenotokous counterparts (4.8 hosts/parasitoid)(Table 3).For the two other comparisons (the finite host stinging rate (ωs) and stable host stinging rate (ψs)), there were no significant differences between the two strains.

3.6.Total host-killing rate

The daily age-specific total host-killing rate (ux), daily agespecific net total host-killing rate (wx), and daily cumulative total host-killing rate (Zx) of the thelytokous strain were generally higher than those of the arrhenotokous strain during the entire adult period (Fig.6).Compared to the arrhenotokous strain, the daily total host-killing rate increased faster than in its arrhenotokous counterparts from days 10 to 14, and then increased moderately until reaching a peak.The peak daily age-specific total killing rate (ux) of the thelytokous strain was 23.2 hosts per individual at 19-day age, which was greater than that of the arrhenotokous strain (15.4 hosts per individual at 21-day age).Although the maximum value of the daily agespecific net total host-killing rate (wx) of the arrhenotokous strain occurred earlier than that of the thelytokous strain, it was lower in the arrhenotokous strain (20.9 hosts on day 19vs.15.5 hosts on day 21).

Fig.6 Cumulative total host-killing rates (Zx), age-specific total host-killing rates (ux), age-specific net total host-killing rates(wx) of the thelytokous and arrhenotokous strains of Diglyphus wani.TH, thelytokous strain; AR, arrhenotokous strain.The gaps represented that parasitism rate is all zero before the adult stage, because the parasitoids are incapable of killing hosts for either of the two strains of D.wani.

The net total host-killing rate (Z0) (204.1 hosts/parasitoid) of the thelytokous strain ofD.waniwas ~2.4 times greater than that of the arrhenotokous strain (84.2 hosts/parasitoid) (Table 3).Moreover, the transformation rate (Qp) of the thelytokous strain was significantly lower than that of the arrhenotokous strain.The stable total hostkilling rate (ψ) and the finite total host-killing rate (ω) of the thelytokous strain were not significantly different from the corresponding values of the arrhenotokous strain.

3.7.Population projection

The projection of the population growth of the two strains ofD.wani, when starting with an initial population of 10 pupae at 10-day age, is shown in Fig.7.The growth rate of the thelytokous population was several thousand times that of the arrhenotokous strain (Fig.7).

Fig.7 Population projections of the two strains of Diglyphus wani with an initial population of 10 pupae at 10-day age.TH,thelytokous strain; AR, arrhenotokous strain.

The total killing rates of the two strains ofD.waniwere also simulated to predict their control potential over the next 30 days (Fig.8).The host-killing potential of thelytokous females was stronger than that of arrhenotokous females because the logarithmic curve of thelytokous females was always higher than that of arrhenotokous females (Fig.8).

Fig.8 Projections of the total host-killing rates of the two strains of Diglyphus wani.

4.Discussion

4.1.Population dynamics parameters

The life table is a recommended tool for evaluating differences in population dynamics parameters in terms of the two reproductive modes of a parasitoid (Chiet al.2020; Yeet al.2023).In this study, the thelytokous strain ofD.wanidemonstrated a higher net reproductive rate,finite rate of increase, and intrinsic rate of increase than its arrhenotokous counterpart.In addition, the thelytokous strain had a significantly shorter mean generation time than the arrhenotokous strain.The comparisons of the two strains in this study revealed that the thelytokous strain ofD.wanimay reproduce more quickly and establish a larger population in a shorter period than its arrhenotokous counterpart.The population projection results also support these discoveries.Except for the mean generation time, our results agree with the findings for the leafminer parasitoidNeochrysocharisformosaand theTrichogrammaparasitoids of lepidopteran pests (T.pretiosum,T.evanescens, andT.embryophagum) with two reproductive patterns (Roopaet al.2009; Xiaoet al.2021; Yeet al.2023).However, our results were different from those reported for two strains ofT.kaykai, because the thelytokousT.kaykaiinfected byWolbachiadid not outperform the antibiotically cured arrhenotokous strain in terms of the population dynamics parameters (Miura and Tagami 2004).

4.2.Fecundity, parasitism rate, and adult longevity

In the present study, thelytokousD.wanishowed a higher fecundity and finite parasitism rate than the arrhenotokous strain, whereas there were no significant differences in adult longevity between the females of the two strains.The adult longevity and fecundity results in our study were similar to those found in the two strains ofT.pretiosum(Prabhulingaet al.2016).Conversely, the longevity of thelytokous females was shorter than that of arrhenotokous females, although thelytokous females also had higher fecundity inV.canescens(Pelosseet al.2007).For another parasitoid of agromyzid leafminers,N.formosa,there are no significant differences in adult longevity,fecundity, or the finite parasitism rate between its two strains (Yeet al.2023).Nevertheless, at 26°C and under the sameL.sativaehost sources, thelytokousD.wanihad higher fecundity than thelytokousN.formosadespite having a shorter lifespan.This difference in the results may indicate that thelytokousD.waniand thelytokousN.formosahave different trade-offs between fecundity and longevity.The reason for the difference in the trade-offs may be that the two parasitoids have different biological adaptations.In addition, thelytokousD.wanimay have a stronger population reproductive potential than the thelytokousN.formosa, although the different experimental settings may also affect this inference.

Fecundity reflects the potential for the establishment and expansion of a population.Understanding the factors that potentially influence the higher fecundity of thelytokous strains ofD.waniwill help us to better understand their population propagation and breeding.Here, we propose two possible reasons for the higher fecundity of the thelytokous strains.On the one hand,the amount of nutrients used for reproduction may differ between the two strains.For the two strains ofD.wani,the main nutrient resource is host-feeding.Nutrients gained from host-feeding are closely related to fecundity,especially for synovigenic parasitoids (Jervis and Kidd 1986).Although there were no significant differences in the number of host-feeding events between the thelytokous and arrhenotokous strains, there was greater feed intake in the thelytokous strain (unpublished data).Feeding intake is defined as the proportion of the host’s body fluids consumed by the parasitoids to the total amount of body fluid in the host, which can be roughly determined from the degree of host desiccation.Thelytokous strains may invest more nutrients or energy in fecundity by increasing their feed intake.Therefore,in a relative sense, they may tend to eat the host more thoroughly when there are fewer resources and repeat feeding on dead hosts.Similar results have been found in thelytokousV.canescens, which allocates more energy to egg production (Pelosseet al.2007).On the other hand,thelytokous females are not disturbed by males and do not need to spend time mating with them when searching for and dealing with hosts.Therefore, the thelytokous strain may have more adequate and efficient time to complete the process of parasitism, and thus achieve a higher fecundity, than the arrhenotokous strain.

4.3.Host-feeding rate

In recent decades, host-feeding parasitoids have received a great deal of attention in the study of host-killing behavior.Some host-feeding parasitoids, especially eulophids, have three different host-killing behaviors(parasitism, host feeding, and host stinging) (Bernardoet al.2006; Liuet al.2015; Chenget al.2017; Yeet al.2023).In this study, the most dominant host-killing behavior for both the thelytokous and arrhenotokous strains was host feeding.In the present study, there was no significant difference in the finite host feeding rate between the two strains ofD.wani.However, inN.formosa, the thelytokous strain showed a higher finite host feeding rate than the arrhenotokous strain.The most likely reason is that theRickettsia-induced thelytoky ofN.formosarequires the parasitoid to feed on more hosts in order to obtain more nutrients to compensate for its deficiency in nutrient synthesis (Yeet al.2023).

4.4.Total host-killing rate

Another important finding is that the finite total host killing rate of thelytokousD.waniwas significantly higher than that of the arrhenotokous strain during the entire adult stage.In the present study, thelytokous females killed 204.1 hosts per individual, whereas arrhenotokous females killed 84.2 hosts per individual.The population projection results also demonstrated that the thelytokous strain could sustainably kill more hosts than the arrhenotokous strain when releasing 10 pupae at 10-day age.Therefore, thelytokous strains may have a greater opportunity to produce offspring when host resources are scarce.The above results suggest that the thelytokous strain ofD.wanihas better host killing and population breeding potential than the arrhenotokous strain.

4.5.Preadult duration

There were no significant differences in the total preadult duration.For preadult duration, the two strains ofD.waniwere similar to the two strains ofV.canescens(Schneideret al.2001; Pelosseet al.2010),T.pretiosum(Prabhulingaet al.2016) andN.formosa(Yeet al.2023).However,our results differed from those observed inT.minutum.Nevertheless, contradictory results regarding the immaturity of the two strains ofT.pretiosumhave been reported by Roopaet al.(2009), Prabhulingaet al.(2016), and Xiaoet al.(2021).Thus, to address the variations in life history traits between thelytokous and arrhenotokous strains,Ameriet al.(2015) proposed that experiments should include replication at the strain level, comparing multiple sexual and asexual strains.However, due to the limited geographical distributions of strains, or their cross-country or cross-regional distributions, repeated acquisition of different strains may not be feasible (Ameriet al.2015).

4.6.The prospect for the application of thelytokous parasitoids

Finally, from the perspective of low-cost biological control application, applying thelytokous parasitoids for biological control could be a preferred strategy because they quickly establish new populations that have a low density, higher population growth rate, and lower production cost (Stouthameret al.1993; Stouthamer 2003; Ramirez-Romeroet al.2012; Prabhulingaet al.2016; Yeet al.2018, 2023).However, many factors, such as different temperatures,habitats, and host densities, have been found to influence the comparative evaluation of thelytokous and arrhenotokous strains of parasitoids (Stouthamer 1993; Rahimi-Kaldehet al.2017).Thus, these factors should be considered when determining the actual control efficiency of parasitoids.Future research should expand on these aspects in order to better and more systematically evaluate the field application potentials of the two strains ofD.wani.

5.Conclusion

As a result of reproductive differentiation, thelytokous and arrhenotokous strains of parasitoids may have different ecological constraints, which may result in different biological characteristics between the strains (Ameriet al.2015).In the present study, we compared the life history and life table parameters of two strains ofD.wani.The results indicated that the thelytokous strain has greater biocontrol effectiveness for controlling agromyzid leafminers than the arrhenotokous strain.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (31972344 and 31772236), the National Key R&D Program of China (2021YFC2600400)and the Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences (caascx-2022-2025-IAS).We are grateful to Professor Chi Hsin(National Chung Hsing University, China) for his valuable comments.We also would like to thank MS student Lü Xi(Hunan Agricultural University, China) for helping to draw the Fig.2 and undergraduate students Liu Caiqin (Jining Normal University, China), Li Yu (Jining Normal University,China), and Liu Lei (Shanxi Agricultural University,China) for helping to replace plant leaves daily.We also would like to thank Editage (www.editage.cn) for English language editing.

Declaration of competing interests

The authors declare that they have no conflict of interest.