High temperatures do not decrease biocontrol potential for the host-killing parasitoid Neochrysocharis formosa (Hymenoptera:Eulophidae) on agromyzid leafminers

XUAN Jing-li ,XIAO Yue ,YE Fu-yu ,ZHANG Yi-bo ,TAO Shu-xia,GUO Jian-yangLIU 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 Agronomy,Jilin Agricultural University,Changchun 130118,P.R.China

3 Department of Entomology and Plant Pathology,North Carolina State University,Raleigh 27607,USA

AbstractTemperature,as a critical abiotic factor,might influence the effectiveness of biological control by parasitoids in hostparasitoid systems. In this study,Neochrysocharis formosa (Westwood),a larval endoparasitoid,is used to investigate the efficacy of biological control on a vegetable agriculture pest,Liriomyza sativae Blanchard,reared on kidney bean(Phaseolus vulgaris L.),at four constant temperatures (26,29,32,and 35°C) under laboratory conditions. Our results show that high temperatures (29,32,and 35°C) do not significantly affect lifetime host-killing events of female adults by increased daily host-killing events compared to temperature of 26°C,although their lifespans decrease with an increase in temperatures. Each life-history trait of female adults (lifespan,parasitism,stinging,or non-reproductive host-killing events) present a linear relation with temperatures and host-feeding events,respectively. Our findings contribute to a better understanding of biocontrol efficacy of parasitoid N.formosa against agromyzid leafminers at high-temperature seasons or environments.

Keywords:leaf-mining fly,host-feeding parasitoid,heat stress,biological control

1.Introduction

Global climate warming has become a major concern for ecologists as it affects geographical synchronization,biodiversity,ecosystem functioning,physiological traits,and behaviors of ectothermic organisms,particularly insects (Heglandet al.2009;Walther 2010;Schmitz and Barton 2014;Abramet al.2017;Peclet al.2017). Some insects have evolved different adaptative strategies to respond to external environmental variations,such as modifications in morphology,changes in physiology,and shifts in behavior (Oswald and Arnold 2012;Seebacheret al.2015). Among insects,parasitoid wasps are potentially more susceptible to environmental changes due to their involvements in tri-trophic communities(Tylianakis and Binzer 2014;Furlong and Zalucki 2017).

Temperature is considered as a major external environment factor to influence life-history traits and behaviors of parasitoids,for instance,some parasitoids present negative effects on lifespan and reproduction(Wilkes 1963;Moirouxet al.2015;Zhanget al.2019),sex ratio (Nguyenet al.2013),host resources forage (Moirouxet al.2015),patch residence time on host (Abramet al.2015),as well as host-killing rates (Stokset al.2017)when exposed to temperature-stressed environments. To date,most studies mainly focused on the development time of immature stages (egg,larva,and pupa) as well as life-history traits of adult parasitoids at a range of constant temperatures or temporary periods of high-temperature treatments (Duncan and Pena 2000;Hansen and Jensen 2002;Bernardoet al.2006;Calvoet al.2013;Zhanget al.2019). However,few studies considered if constant high temperature environments would affect the life-history traits and biocontrol potentials of host-feeding parasitoids during their entire adult stage.

Thus,the objective of this study is to investigate the effect of constant high temperatures on the lifespan of female adults and their lifetime host-killing events. To achieve this goal,a destructive host-feeding parasitoidNeochrysocharisformosa(Westwood) (Hymenoptera:Eulophidae) was selected as a study subject to evaluate the biocontrol efficacy of female adults in suppressing agromyzid leafminers.Neochrysocharisformosais a solitary synovigenic,larval endoparasitoid,widely distributed in the world (Wanget al.2012),and is capable of adapting to a broad range of temperatures,especially tolerating the high-temperature environments (Moonet al.2004). The parasitoid is a dominant biological control agent against invasive vegetable pestsLiriomyza sativaeBlanchard andLiriomyzatrifolii(Burgress) in summer and early fall in the field (Wanget al.2012). It has been reported that optimal rearing temperature for the parasitoid is at around 26°C in the laboratory or greenhouse (Moonet al.2004;Salehet al.2010). The wasp exhibits three host-killing types:parasitism,host feeding,and stinging (Osmankhilet al.2010;Zhanget al.2011;Wanget al.2012;Liuet al.2015;Xuanet al.2018). Parasitoids with host-feeding behavior have been considered potentially significant biocontrol agents (Kidd and Jervis 1989;Briggset al.1995;Lewiset al.1998)since they can allocate host resources gained by feeding to survive longer,or to mature more eggs,or to provide themselves more energy to search for hosts (Jervis and Kidd 1986;Chan and Godfray 1993). In this study,we not only compared lifespan and host-killing events (parasitism,host feeding,and stinging) of female adults among four different temperatures (26,29,32,and 35°C),but also explored the relationships between host-feeding events and several other life-history traits (lifespan,parasitism,stinging,and non-reproductive host killing events) at distinct temperatures. The results in the present study can contribute to developing an effective biocontrol strategy system as a major component of integrated leafminer management programs applied in the field.

2.Materials and methods

2.1.Insect cultures

Colonies of bisexualN.formosa(parasitoid) andL.sativae(host) were established and maintained in climate chambers ((26±1)°C;(50±5)% relative humidity (RH);14 h L:10 h D photoperiod)) in the laboratory at the Department of Biological Invasion (DBI),Institute of Plant Protection(IPP),Chinese Academy of Agricultural Sciences (CAAS),Beijing,China. The colonies were originally collected from vegetable fields near the IPP,CAAS (116.28°E,40.02°N) in August 2013. The cultures ofL.sativaewere maintained on kidney beans (PhaseolusvulgarisL.) in the gauze cages (length×width×height=50 cm×50 cm×60 cm,mesh number=100),while parasitoidN.formosawere reared using kidney bean leaves infested byL.sativaelarvae in the gauze cages (length×width×height=40 cm×40 cm×40 cm,mesh number=100).

2.2.Lifespan and host-killing events of female adults

To investigate if high temperature affects the lifespan and host-killing events of female wasps,four temperature treatments (26,29,32,and 35°C) were designed and simultaneously conducted in the laboratory. To collect host-killing data,the late second instar to early third instar larvae ofL.sativaewere selected as hosts (Wanget al.2007). Kidney bean leaves infested withL.sativaelarvae(approximately 30/leaf) were detached from plants. Extra host larvae were removed using a dissection needle under a binocular microscope (Olympus;Tokyo,Japan)to obtain the same density of host larvae in each leaf. To keep the leaf fresh,1% agar gel was introduced into each glass Petri dish (9.0 cm in diameter×2.2 cm in height),and covered around one-third height of a Petri dish.Leaves within host larvae were immediately transferred to each Petri dish as soon as the agar gel mostly solidified,simultaneously,each leaf petiole was also embedded into the gel. After that,all the Petri dishes were sealed with parafilm. To maintain normal air circulations between the inside and outside of the Petri dish,the parafilm was pierced with numerous holes by a needle. Petri dishes with the leaves were made prior to transfer of the parasitoids.

To obtain freshly emerged parasitoid adults,all the parasitoid adults were removed from the colony cage using an aspiration tube at 7:30 a.m.every day. Newly emerged parasitoid wasps were obtained at the peak of emergence from 8:00-9:00 a.m.in the colony cage.Parasitoid adults with uniform body sizes were selected and used in the experiment. A pair of parasitoids (a female and a male) were randomly assigned to a readymade Petri dish,then Petri dish was placed into a growth chamber with our designated temperature settings((50±5)% RH;14 h L:10 h D photoperiod)).

After exposure to 24 h in a Petri dish,both a female and a male were transferred into a new one with a fresh leaf infested with host larvae if they both survived.Otherwise,only a female was transferred if a male was dead earlier than a female. The transfer was stopped if female died in the Petri dish,and the survival of each female adult was recorded in the experiment. Petri dishes affected by females were placed into a growth cabinet set within optimal conditions ((26±1)°C;(50±5)% RH;14 h L:10 h D photoperiod)) for the development of parasitoids’eggs (Zhanget al.2014). After 48 h,each leaf in the Petri dish was examined to count the numbers of host-killing events including parasitism,host feeding,and stinging under a binocular microscope based on the appearance of dead host larvae.

Specifically,for parasitic event (parasitism),a dead host larva was swollen due to the development of an egg inside. To confirm this,each dead host larva was dissected. For host-feeding events,a dead host larva was flat and dry,even with only larva skin left under the leaf epidermis in that the tissue and hemolymph of host larva were completely consumed by the parasitoid. For host stinging events,the body of a dead host larva was full,and seems to be a living host larva except for the punctures on the surface of host larva generated by the parasitoids,because parasitoid killed the host larvae by injecting venom instead of laying eggs and feeding.Typically,parasitism was considered as a reproductive host-killing behavior,while other two host-killing events(host feeding and stinging) were regarded as nonreproductive host-killing behaviors (Barrett and Brunner 1990;Duncan and Pena 2000;Bernardoet al.2006;Grabenwegeret al.2009;Zhanget al.2011;Liuet al.2015;Xuanet al.2018).

2.3.Statistical analysis

Survival curves of female adults at four temperature treatments were plotted using the Kaplan-Meier method,and were compared within a log-rank test implemented in the GraphPad Prism (version 9.0.0).

A one-way analysis of variance (ANOVA) was used to compare the means of lifetime and daily host-killing events(including parasitism,host-feeding events,host-stinging events,non-reproductive host-killing events,and total mortality) among different temperature treatments since the data were subjected to normal distribution and homogeneity of variances. Multiple comparisons of host-killing events were conducted using the Bonferroni method in SAS 9.3.

Age-specific curves of host-killing events among different temperatures were analyzed over time using a generalized linear modeling approach conducted in IBM SPSS Statistics ver.26. This test was conducted not only to compare the differences among the temperatures independent of time using Bonferroni adjustment for multiple comparisons,but also to test if the curves differed over time. To avoid missing out most values of host-killing events due to the variations in lifespans of female adults at the distinct temperature treatments,the shortest mean lifespan (9 days at 35°C) was selected as a timepoint to truncate the data,namely,the data for only the first 9 days were used to analyze the age-specific curves of hostkilling events among four temperature treatments.

To explore if there exists a relation between hostfeeding events and any other life history trait,we plotted 3D scatters for each life-history trait (lifespan,parasitism,host-stinging event,and non-reproductive host-killing event) linked with host-feeding event and temperature conducted in Origin 9.0,then multiple linear regression was adopted to analyze the relationships among them implemented in SAS 9.3.

Any statistical significance was determined under a threshold level of 0.05 (P＜0.05) in the comparison analysis.

3.Results

3.1.Lifespan of female adults

Median survivals of female adults at 26,29,32,and 35°C are 19,16,11 and 10 days,respectively. Survival curves under four temperature treatments are significantly different (χ2=17.55,df=3,P=0.0005),however,there are no significant differences between 26 and 29°C(χ2=0.7147,df=1,P=0.3979),as well as between 32 and 35°C (χ2=1.132,df=1,P=0.2874;Fig.1).

Fig.1 Survival curves of Neochrysocharis formosa female adults at four different temperature treatments.

3.2.Host-killing events of female wasps

There are no significant differences on the lifetime host-killing events among four temperature treatments(specifically,parasitism events:F=2.65,df=3,P=0.0538;host-feeding events:F=2.70,df=3,P=0.0500;hoststinging events:F=1.26,df=3,P=0.2939;non-reproductive host-killing events:F=1.41,df=3,P=0.2444;total mortality:F=1.77,df=3,P=0.1593). The means of lifetime oviposition and lifetime non-reproductive host-killing events at 29°C are higher compared to any other temperature (Table 1).

Daily host-killing events of the parasitoids first show an increasing trend,then tend to be stable followed by an increase in temperatures (Table 1). Specifically,daily oviposition and daily host-feeding events are separately significantly higher at 29°C than those at 26°C (daily oviposition:F=33.60,P＜0.0001;daily host-feeding events:F=45.17,P＜0.0001),but not significant among high-temperature treatments at 29,32,and 35°C(daily oviposition:F=0.09,P=0.9162;daily host feeding events:F=0.16,P=0.8504).The other three daily host-killing traits (host-stinging events,non-reproductive hostkilling events,and total mortality) separately exhibit an increase trend from 26 to 32°C,otherwise,there are no significant differences between 32 and 35°C (daily host-stinging events:F=0.23,P=0.6337;daily non-reproductive host killing events:F=0.10,P=0.7509;daily host mortality:F=0.12,P=0.735).

3.3.Age-specific curves of host-killing events of female wasps

Age-specific curves of parasitism among four temperature treatments differ substantially,and there is a significant interaction between temperature and parasitoid age(temperature:F3,63=7.7534,P＜0.0001;age:F8,56=9.568,P＜0.0001;temperature×age:F24,163=2.136,P=0.003). The parasitism curves in the first six days at high temperatures(29,32,and 35°C) are above the one at 26°C. The peak of oviposition occurs earlier at 35°C than those at any other temperature treatment,specifically,the peak occurs on day 16 at 26°C (4.66 eggs),day 9 at 29°C (7.47 eggs),day 4 at 32°C (6.78 eggs),and day 3 at 35°C (6.35 eggs),respectively. Interestingly,female adults are inclined to not lay eggs at the last one or two days of the life with the exception for 32°C (Fig.2-A).

Temperature and parasitoid age significantly influence the daily dynamics of hostfeeding events,and there is a significant interaction between them (temperature:F3,63=42.654,P＜0.0001;age:F8,56=2.844,P=0.010;temperature×age:F24,643=1.817,P=0.016). In contrast with optimal temperature 26°C,the curves of host-feeding events at high temperatures (29,32,and 35°C) are more fluctuated and above the curve at 26°C. The numbers of host feeding at the last day in a life separately show increased trends at 29 and 35°C compared to 26 and 32°C (Fig.2-B).

The age-specific curves of host-stinging events are significantly influenced by temperature and parasitoid age,however,there is no significant interaction between them (temperature:F3,63=22.646,P＜0.0001;age:F8,56=4.295,P＜0.0001;temperature×age:F24,163=1.319,P=0.159). The curves in the first nine days at 32 and 35°C are over those at 26 and 29°C,respectively (Fig.2-C).

Temperature and parasitoid age as well as the interaction between them have significant impacts on the age-specific curves of non-reproductive host-killing events(temperature:F3,63=53.498,P＜0.0001;age:F8,56=6.592,P＜0.0001;temperature×age:F24,163=2.834,P＜0.0001).The numbers of non-reproductive host-killing events are slightly elevated in the last one day of a life at 29 and 35°C,while they tend to decline at the last four days in a life at 26 and 32°C,respectively (Fig.2-D).

Fig.2 Age-specific curves of the number of parasitism (A),host-feeding events (B),host-stinging events (C),and non-reproductive host mortality (D) for Neocheysocharis formosa females at 26,29,32,and 35°C,respectively. Bars on the plots represent the standard errors of the means (n=26,23,23 and 23,respectively).

3.4.Relationships of life-history traits associated with host-feeding events and temperature

The relationships between each life-history trait (lifespan,parasitism number,host-stinging number and nonreproductive host-killing events) and host-feeding events are linear based on a visual inspection of the residuals(Fig.3). The regression model (host-feeding events and temperature) provides better fits to the data for lifespan and non-reproductive host-killing events than those for parasitism and host-stinging events according to the coefficient of determinationR2(Table 2). Nevertheless,there are statistically significant associations between the model and each life-history trait (P＜0.0001;Table 2).

4.Discussion

Temperature is perhaps the most critical environmental factor affecting insect life-history traits,foraging behaviors,and geographic distributions (Angilletta and Angilletta 2009;Régnièreet al.2012;Sánchez-Guillénet al.2016).Previous studies have reported that high temperature might induce many negative effects for some parasitoids,such as decreasing the lifespans (Morales-Ramos and Cate 1992;Hentzet al.1998;U?kan and Ergin 2002;Zhanget al.2019),lowering the reproductions by damaging their oocytes and ovarian developments(Mironidis and Savopoulou-Soultani 2010),as well as affecting sex biases in their progenies since heat stress dramatically decreases the sperm stock of male adults(Nguyenet al.2013). Nevertheless,for host-feeding parasitoidN.formosain our study,high temperature plays a positive role on daily oviposition,that is,the parasitoid lays more eggs instead of extending its lifespan in face of high temperature stress (Table 1). This is probably because high temperatures might speed up the maturation of eggs inN.formosa,causing an increase in daily oviposition. The phenomenon of tradeoffs between current maintenance and future reproduction has been also found in other host-feeding parasitoids (Sevensteret al.2000;Desouhantet al.2005;Rivero and West 2005;Marshall and Sinclair 2010).

In our study,high temperature decreases the lifespan of parasitoidN.formosa,but do not significantly decrease its biological control efficacy by increased daily mortality(Fig.1;Table 1). There are two potential reasons for the interesting findings. First,it is probably related to selfbiology/bionomics characteristics ofN.formosain that the parasitoid can tolerate high-temperature environments(Moonet al.2004). Second,the parasitoid is possibly more active at a high-temperature environment that speeds up its metabolic rate,further resulting in more host consumptions (Olsonet al.2000;Acaret al.2001),eventually increasing the non-reproductive host-killing events especially for host-feeding events (Table 1).Host-feeding behavior of parasitoids not only plays a significant role on biological control,but also exhibits the extension of lifespan and the promotion of egg maturation for synovigenic parasitoids that are required to gain the protein materials from host tissues by feeding (Chan and Godfray 1993;Briggset al.1995;Heimpel and Collier 1996;Gironet al.2002;Rivero and West 2005;Kapranas and Luck 2008;Miksanek and Heimpel 2020). In the present study,host-feeding events inN.formosashow positive relations with other life-history traits (Fig.3;Table 2).

Fig.3 Scatter plots for life-history traits including lifespan (A),parasitism (B),stinging (C) and non-reproductive host mortality (D),associated with host-feeding events and temperatures.

Table 2 Linear regression analysis of each life-history trait linked with host-feeding event and temperature

Additionally,temperature not only affects the behaviors of parasitoids,but also simultaneously impacts on the host behaviors,especially for host larvae defense against parasitoids in the host-parasitoid system (Jeffs and Lewis 2013;Jerbi-Elayedet al.2015;Moirouxet al.2016;Furlong and Zalucki 2017). Some researchers have found that high temperature generally increases the capacity of hosts to defend against the immature parasitoids,resulting in a low survival rate in the progeny of parasitoids (Hanceet al.2007). In contrast,Iltiset al.(2018) found that high temperature reduces behavioral and immune defenses of host larvae,further enhancing the effectiveness of biological control of parasitoids. In our study,biocontrol efficacy ofN.formosado not significantly decrease at high temperatures (Table 1),probably because high temperature stress might not significantly influence the defense behavior ofL.sativaelarvae.

Data were collected under laboratory conditions in our study,but different results may be observed in this interaction when data are collected in the field since there exist more complicated and unpredictable factors such as relative humidity,photoperiod,host density,community competitions in the organism network and so forth. These factors should be also concerned before parasitoidN.formosais released into agricultural fields in the future. Nevertheless,our results presented in this study are a basic evaluation on the effect of temperature on the biological control efficacy ofN.formosa,which provides significant evidence to apply the parasitoid in the biological control program.

5.Conclusion

Neochrysocharisformosais a predominant biological control agent of agromyzid leafminers. The results in this study emphasize the effect on lifespan and biocontrol effectiveness of female adults at four distinct temperatures.Specifically,increased temperatures decrease the lifespan of female adults,however,do not significantly influence their biocontrol efficacy by increasing daily host-killing events,especially non-reproductive host-killing events including host feeding and stinging. Each life-history trait presents a linear relation with host-feeding events and temperatures. The study contributes evidence to promote the release of host-feeding parasitoidN.formosain agriculture greenhouses or fields.

Acknowledgements

We are grateful to Prof.Brian M.Wiegmann,Ph D candidates Samuel Buzuleciu and Peter Willadsen,at the Department of Entomology and Plant Pathology,North Carolina State University,USA for not only editing the English text of a draft manuscript but also providing constructive and invaluable comments on the manuscript.We greatly thank two anonymous reviewers for helpful and constructive comments and suggestions on improvement the quality of the manuscript. This work was supported by the National Natural Science Foundation of China (31772236),the Science and Technology Innovation Program of Chinese Academy of Agricultural Sciences (caascx-2017-2022-IAS),and the Program of China Scholarship Council (201807990002).

Declaration of competing interest

The authors declare that they have no conflict of interest.