Indirect and direct interactions between grain aphid and parasitoid in the presence of symbiont Regiella insecticola

Abstract

Background

Aphids often harbor bacterial symbionts that confer resistance to biotic and abiotic stress. Previous studies have primarily examined the direct effects of symbiont infection on parasitoid success but less on aphid population dynamics under indirect parasitic situations, for example when exposed to parasitoid wasp odor. Deciphering this type of indirect effect is essential for understanding dynamics of insect ecosystems and communities and to improve IPM success.

Methods

We generated Sitobion avenae aphid clonal lines that are genetically identical but differ in Regiella insecticola infection. Then, the indirect odor effect of female parasitoid wasp Aphidius gifuensis (Ashmaed), one of its major natural enemies in the fields, was examined on the aphid lines fitness using different parasitoid densities. With these lines we also tested the direct effect of symbiont presence on aphid resistance against different parasitoid densities.

Results

Our study found fitness costs for the aphid line hosting Regiella, mainly via an increase in the development time and a reduction in population increase rate. Some of these fitness traits were influenced by the indirect exposure to parasitoid wasp odor with a density effect. Presence of the symbiont also reduced A. gifuensis parasitic success, increased the wasp development time and decreased its emergence weight with low effect of the parasitoid density used for parasitism.

Conclusions

These results showed that aphid population dynamic was mainly affected by the symbiont presence, but not by parasitoid odor. Symbiont presence also protected aphid from parasitism and affected parasitoid offspring weight and hence their future individual fertility and fitness.

Background

Wheat is one of the most important food resources for human beings, and its yield loss is mainly dependent on the magnitudes of pest attack and plant virus infection. Piercing-sucking insect pests can suck wheat phloem sap but also transmit many plant viruses, which seriously affect the wheat growth and reduce wheat production (Li et al. 2019; Luo et al. 2020a; Zogli et al. 2020). Evidences have shown that parasitoids may be the most important agents for controlling piercing-sucking pests (Schmidt et al. 2003; Stell et al. 2022). However, parasitoid’s hosts often shelter endosymbionts that confer biotic and abiotic resistance (Desneux et al. 2018; Liu et al. 2022).For example, emerging evidences have shown that some of them can protect against parasitic wasps in aphids (Oliver et al. 2010; Monticelli et al. 2019a; Vorburger 2022). It is thus important to understand the role(s) of these symbionts in the efficiency of biological pest control (Käch et al. 2018; Attia et al. 2022).

Aphid is one of the most serious piercing-sucking agricultural pests on wheat plants worldwide (Li et al. 2020, Aradottir and Crespo-Herrera 2021). Most aphids live in obligatory symbiosis with the intracellular bacterium Buchnera aphidicola (Sochard et al. 2021), which synthesizes some essential amino acids and vitamins for the host survival and reproduction (Perreau and Moran 2022). In addition, some aphid individuals can also host one or more facultative symbionts, but their association varies from free association to co-obligation with intermediate stages of dependence (Henry et al. 2013; Guyomar et al. 2018). Until now, at least 9 different facultative symbionts have been described from aphids caught in the fields, the most common being Hamiltonella defensa, Regiella insecticola, and Serratia symbiotica, while Rickettsiella viridis, Candidatus Fukatsuia symbiotica, Rickettsia, Spiroplasma, Wolbachia and Arsenophonus are rare in most aphid populations (Augustinos et al. 2011; Guyomar et al. 2018; Cariou et al. 2018). In general, facultative symbionts are not necessary for the growth and reproduction of host aphids, but they could strongly affect the resistance (i.e., direct effects) to natural enemies (Scarborough et al. 2005; Zytynska et al. 2021; Luo et al. 2022; Frago and Zytynska 2023). Interestingly, emerging evidences suggested that indirect interactions (e.g., natural enemy-mediated effects in insect behavior and physiology by odors) are equally or more significant in the change of host population dynamic than the direct effect (Hermann and Landis 2017; Luo et al. 2022). In addition, the odor-mediated effects of natural enemies often occur in the field and may further involve other trophic organisms. Therefore, deciphering this type of indirect effect is essential for the understanding dynamics of insect ecosystems and communities (Ives and Carpenter 2007; Luo et al. 2022). A new literature has shown that compared to without exposure to the predator Harmonia axyridis, the effect of symbiont Regiella on aphid population increase would be compromised under the indirect odor effect (Luo et al. 2022). However, wasp odor-mediated effects on the behavior and physiology of aphid carrying or not the symbiont have not been well documented.

During previous surveys of grain aphid Sitobion avenae populations from different locations in China, we found that most of these populations were infected by Regiella (Luo et al. 2016; Hu et al. 2020). Some studies have suggested that the presence of Regiella could have a fitness cost for host aphids, but help to resist parasitoids (Von et al. 2008; Luo et al. 2020a; Vorburger and Gouskov 2011). Evidence for this trade-off is limited under the indirect interactions in aphid-parasitoid communities, although a study has shown that symbionts can protect aphids from parasitic wasps by attenuating herbivore-induced plant volatiles (Frago et al. 2017). Besides that, some evidences also suggested that the odor-mediated effect on insects could be mediated by natural enemy density component (Muller and Godfray 1999; Uriel et al. 2021). Here, we studied, under different densities of parasitoids, the effects of Regiella on: (1) aphid population dynamics under the odor-mediated situation of the parasitoid Aphidius gifuensis, and (2) resistance levels against this parasitoid. These results are expected to broaden the ecological roles of Regiella, and heighten the significance of facultative symbionts in biological pest control.

Materials and methods

Biological materials

To explore the biological effects of the Regiella insecticola symbiont, it was important to generate aphid strains that were genetically identical and differed only in Regiella infection. Two clone lines derived from the Sitobion avenae YL₂₇ were used in the current study. The natural YL₂₇-NA clone line, devoid of facultative symbionts, was derived from a single parthenogenetic aphid female, which was collected from a winter wheat field on the university research farm (Northwest A&F University, Yangling, China) in May 2019. The YL₂₇-Ri was obtained by the microinjection of body fluids of Regiella-infected S. avenae clone YP-Ri (Luo et al. 2022) to YL₂₇-NA (Chen and Purcell 1997). This new clone line artificially harboring Regiella was established after confirming the presence of the symbiont at the 10th generations (Additional file 1: Table S1). All these aphid lines were fed on wheat seedings (Triticum aestivum L. cv. ‘Aikang 58’) and maintained in a climatic room (20 ± 1 °C, RH of 75%, LD 16: 8 h). During the study, the two YL₂₇ clone lines were regularly screened by PCR method (Monticelli et al. 2019b) to confirm the presence/absence of Regiella, and have been maintained for over 1 year prior to the present study.

The parasitoid Aphidius gifuensis, maintained on the symbiont-free S. avenae clone Linyi-NA (Luo et al. 2017a) under the laboratory setting, was collected from a wheat field of the university experimental farm (Northwest A&F University, Yangling, China) in June 2014. Adult wasps were maintained in separate cages (side length = 25 cm) with cotton balls wetted with a honey solution (honey/water = 1/10). Four-day-old female wasps assumed to be mated were used in the experiment. Thirty minutes before the parasitism assays, the female wasp was exposed to uninfected 2nd-instar aphids in a Petri dish and only the wasp that oviposited within 5 min were used in the experiments (Luo et al. 2020b).

Effect of female wasp parasitoid odor on aphid fitness

To obtain synchronized aphid individuals for each aphid line, 10 adult wingless aphids (11 days old) were transferred to a new plant to reproduce for 12 h. Then, 30 aphid offspring were collected and transferred to an aerated transparent container (∅ 6 cm and 25 cm height) with a 2-leaf stage wheat. After 6 h of adaptation, a 50 ml of conical centrifugal tube, with a 100-micron nylon filter mesh at the top for ventilation, containing parasitoids (1, 3 or 5 individuals; 0 = empty tube used as the control; in order to avoid the odor interplay of different parasitoids, the treatment groups and the control group were kept in the transparent Perspex cages and over five meters apart from each other) was introduced for 6 h to expose aphids to the wasp odor. Thereafter, the aphids were individually transferred into a new two-leaf stage wheat seedling in aerated transparent cages. Developmental times of the 4 nymphal stages (DT1 ~ DT4), adult pre-reproductive period (APRP; the duration from adult emergence to the first reproduction) and total pre-reproductive period (TPRP; the duration from birth to the first reproduction; only reproducing aphids were considered in the following analysis) were recorded. In addition, the offspring number and mortality event of each aphid were also recorded daily until all died. During this assay, all offspring were removed every day to maintain a uniform reproduction environment. Each treatment was performed using four groups for each clone lines, with 30 aphids in each group. Finally, based on the aphid fitness parameters, a TIMING-MS Chart was used to predict the population growth of the S. avenae lines. For this prediction, we assumed unlimited growth with an initial population of 10 pairs of newly emerged S. avenae nymphs (Chi et al. 2022a).

Effect of symbiont on aphid parasitism

For each clonal line, thirty 4-day-old synchronized nymphs (2nd-instar) were transferred on a two-leaf stage wheat seedling in a transparent ventilated container. After 18 h, female wasps (1, 3 or 5) were introduced into the container and left 6 h for parasitism. Four days later, the 30 aphids were individually placed into Petri dishes (∅ 9 cm), containing a fresh wheat leaf blade that was replaced every 3 days (Luo et al. 2020b). During the following ~ 20 days, the numbers of successfully parasitized individuals (mummies), emerged adult wasps and wasp development time in aphids (from the day of parasitoid removal until the emergence of wasp offspring) were recorded. The emerged wasps were also weighed within 24 h (Mettler Toledo XS3DU; Mettler Toledo GmbH, Switzerland). Twelve biological replicates were performed for this study.

Statistical analyses

The data of the aphid fitness were analyzed by using the computer program TWOSEX-MS Chart (Chi et al. 2022a; Chi et al. 2022b) to obtain aphid population parameters (i.e., intrinsic rate of increase, finite rate of increase, net reproductive rate and mean generation time) and aphid development times (DT1 ~ DT4, APRP and TPRP). In brief, the intrinsic rate of increase (r) was calculated by the Euler-Lotka equation (${\sum }_{{\text{x}}=0}^{\infty }{e}^{-r\left(x+1\right)}{l}_{x}{m}_{x}=1$) based on the iterative bisection method, showing a potential of population growth, where l_x is the age-specific survival rate, m_x is the age-specific fecundity, x is age interval; the finite rate of increase (λ) was equal to e^r; the net reproductive rate (${R}_{0}={\sum }_{{\text{x}}=0}^{\infty }{l}_{x}{m}_{x}$) illustrated a mean number of aphid offspring produced during one life cycle; mean generation time ($T=\frac{{\text{ln}}{R}_{0}}{r}$) showed a time that aphid population reach to R₀-times under a stable age-stage distribution condition. Parasitism rate in the study was measured by the number of mummified aphids divided by the number of whole treated aphids (30 aphids), and emergence rate was measured by the number of emerging parasitoids divided by the number of mummified aphids. Containers that had more than 5 missing aphids or had suffered no parasitism have been discarded during the statistical analysis (i.e., two repetitions were removed under the three parasitoid densities).

Factorial two-way analyses of variance were done to test the effects of parasitoid density, symbiont presence or absence and their interaction on the fitness and parasitism parameters calculated in the study. No interaction was found between (Parasitoid density × Symbiont) for all these traits (see Additional file 1: Table S2). Then, based on the analysis results above, the population parameters (i.e., r, λ, R₀, T), aphid development times, weight of emerged wasps and development time of the parasitoid offspring were analyzed using general linear models after validation of the normal distribution of the dependent variable. In addition, effects of the presence of the symbiont on the parasitism rate and emergence rate were analyzed by generalized linear modeling with a quasi-binomial error distribution (link = logit), which considers any over-dispersion in the data. The package “multcomp” was used to perform multiple comparisons using Tukey’s contrasts in the study. All statistics were done with R version 3.4.3.

Results

Effect of presence of symbiont on aphid fitness parameters under the indirect odor effect

The fitness traits of symbiotic aphids were compared with those of the symbiont-free aphids without (0) or after the exposure to the parasitoid odors using different parasitoid densities (1, 3 or 5 individuals). For the population parameters (Fig. 1, Table 1), in the absence of parasitoid exposition, compared to symbiont-free aphids, symbiont-infected aphids showed a reduction in the intrinsic rate of increase (r) and the finite rate of increase (λ), but had an increase on the mean generation time (T). Under the 1, 3, or 5 wasps odor exposition, we found a similar result between infected and uninfected aphids on the three population traits. The net reproductive rate (R₀) was significantly reduced under the 1 or 5 wasps density in the infected aphids. For the development times (Tables 1, 2), in the absence of parasitoid exposition, compared with symbiont-free aphids, symbiont-infected aphids had a significant increase on the development time of 1st and 2nd-instars (DT1, DT2) as well as the adult pre-reproductive period (APRP). A similar result on the three traits between infected and uninfected aphids was also found under the 1, 3, or 5 wasps density. A significant increase of the DT3, DT4 and TPRP (total pre-reproductive period) of the infected aphids was observed with the 3 wasps density, and a clear increase on DT3 and TPRP was examined with the 5 wasp density and the without-exposition to the parasitoid odors, respectively. Overall, the presence of the symbiont had an adverse effect on aphid population increase.

Fig. 1

Population parameters of the Regiella-infected and the symbiont-free S. avenae lines submitted to A. gifuensis odor. The different letters indicated significant difference (P < 0.05) between symbiont-free (S-; white column) or symbiont-infected (S + ; black column) aphids facing odor of different density of A. gifuensis. *P < 0.05, **P < 0.01, ***P < 0.001 (significant difference between aphid lines for the same densities of A. gifuensis, ns: no significant difference). Each value is the Mean ± SEM

Table 1 Statistical analyses of the effects of Regiella presence on aphid populational traits under different densities of A. gifuensis. Table indicates the intrinsic rate of increase (r), finite rate of increase (λ), net reproductive rate (R₀), mean generation time (T), and the developmental times of 1st to 4th instar nymphs (DT1 ~ DT4), adult pre-oviposition period (APRP), total pre-oviposition period (TPRP) of host aphids across different A. gifuensis densities (0, 1, 3, 5 parasitoids)

Table 2 The effect of Regiella on the development times of host aphids under different densities of A. gifuensis (i.e., 0, 1, 3, 5 parasitoids). Days; Mean ± SEM

Effect of presence of symbiont on aphid parasitism under the different parasitoid density

The parasitism traits of infected aphids were compared with those of the uninfected aphids with different parasitoid density (1, 3 or 5 individuals) (Fig. 2). When parasitized by 1 female wasp, compared to symbiont-free aphids, infected aphids showed a lower parasitism rate (number of mummies; χ² = 57.622, df = 1, P < 0.001), an increase on the parasitoid offspring development time (F = 44.698, df = 1, P < 0.001) and a lower weight of emerged parasitoid wasps (F = 55.495, df = 1, P < 0.001). A similar result on these parameters between infected and uninfected aphids was also found when 3 or 5 parasitoids were used (parasitism rate: χ² = 68.579, df = 1, P < 0.001; χ² = 75.061, df = 1, P < 0.001; the weight: F = 43.561, df = 1, P < 0.001; F = 94.097, df = 1, P < 0.001; the development time: F = 26.793, df = 1, P < 0.001; F = 18.792, df = 1, P < 0.001, respectively). However, for the emergence rate, measured by the ratio of emerged parasitoids from the mummies, no difference was observed under the one density (χ² = 0.098, df = 1, P = 0.754), while there was a significant decrease in the presence of Regiella with 3 or 5 parasitoids (χ² = 20.667, df = 1, P < 0.001; χ² = 4.808, df = 1, P < 0.001, respectively).

Fig. 2

Effects of Regiella infection on the parasitism rate (A), emergence rate (B), weight of emerged wasp (C) and development times of A. gifuensis (D) under different densities of A. gifuensis. The different letters indicated significant difference (P < 0.05) for symbiont-free (S-; white column) or symbiont-infected (S + ; black column) aphids among different densities of A. gifuensis. *P < 0.05, **P < 0.01, ***P < 0.001 (significant difference between the two aphid lines for the same densities of A. gifuensis, ns: no significant difference). Each value is the Mean ± SEM

Effects of parasitoid wasp density on fitness and parasitism of the uninfected or infected aphids

The traits of the uninfected or infected aphids were compared among different parasitoid wasp density. For the fitness traits (see capital letter in Table 2), in the uninfected aphids (YL₂₇-NA), no difference on the population parameters (r: P = 0.244; λ: P = 0.243; R₀: P = 0.371; T: P = 0.894) and the development time traits (DT1: P = 0.294; DT2: P = 0.462; DT3: P = 0.161; DT4: P = 0.715; APRP: P = 0.149; TPRP: P = 0.073) was observed among the different parasitoid densities (0, 1, 3 and 5 individuals). Similar results were obtained for these traits (r: P = 0.487; λ: P = 0.485; R₀: P = 0.396; T: P = 0.635; DT1: P = 0.402; DT2: P = 0.606; DT3: P = 0.215; DT4: P = 0.688; APRP: P = 0.175; TPRP: P = 0.334) in the infected aphids (YL₂₇-Ri) among the different parasitoid densities. For the parasitism traits (see letters in Fig. 2), there also was no difference for the uninfected or infected aphids on the parasitism rate (P = 0.445, 0.752), the weight of emerged parasitoid wasp (P = 0.115, 0.165) and the development time of parasitoid offspring (P = 0.180, 0.958) among different parasitoid densities (1, 3 and 5 individuals), except a reduction on the emergence rate of infected aphids at the medium density (P < 0.01).

Discussion

Symbiont-mediated trade-offs between the fitness cost and resistance have been widely reported in insects, but most focused on the model pea aphids (Zytynska et al. 2021; Vorburger 2022). In addition, the effects of the density of parasitoid on aphid fitness and resistance were little known, although the density effect is common in nature and is important to understand community dynamics (Muller and Godfray 1999). Since the grain aphid Sitobion avenae is the most devastating pest for winter wheat in China (Xu et al. 2011; Jiang et al. 2019) and Regiella is widely present in S. avenae individuals in this country (Li et al. 2014; Luo et al. 2016; Hu et al. 2020), we investigated whether the presence of this symbiont could induce a trade-off between S. avenae fitness cost and resistance to A. gifuensi, a biological control agent widely used in China, across different parasitoid densities.

For symbiont-infected or symbiont-free aphids, the presence of different wasp density nearly had no impact on traits of aphid fitness and resistance. Earlier studies also suggested that the parasitoid density had no effect on offspring numbers and the timing of peak reproduction of aphids across four experimental generations (Uriel et al. 2021). These results may suggest that aphid fitness/phenotypes were independent of the parasitoid in the range of density tested, but the conclusion should be viewed with caution for several reasons. For example, the setup of parasitoid density gradient in these studies may hardly induce a significant difference on the quantity of volatiles, although a previous study suggested that there was a significant ladybeetle-odor effect under the low density on aphid population dynamic (Luo et al. 2022). In addition, some studies also showed that in the presence of parasitoids, different aphid species would have various behavioral responses. For example, apterous pea aphids Acyrthosiphon pisum would produce alate offspring (Sloggett et al. 2002), which generally had lower reproductive output, whereas soybean aphids Aphis glycines produced offspring with greater reproductive output (Kaiser 2017). All these findings may suggest that the odor quantity and aphid species should be considered when exploring the indirect effect of parasitoids on insect fitness.

The results of aphid population parameters indicated that Regiella infection results in a significant decrease on the intrinsic rate of increase (r) across different parasitoid densities (1, 3 and 5 individuals), the most important parameter for evaluating the population growth potential of insects (Qayyum et al. 2018). Therefore, the adverse effect on the intrinsic rate of increase suggested that Regiella infection was significantly disadvantageous for the increase of aphid population number and was not wasp density-dependent, since a small change of this parameter could cause a significant change in insect population numbers (Qayyum et al. 2018; Chi et al. 2022a; Chi et al. 2022b). This negative impact on aphid population growth rate was confirmed by the result of the population projection model in the study (Additional file 1: Fig. S1). In addition, this notion was also verified by the results of the net reproductive rate (R₀) and the finite rate of increase (λ) of the infected aphids, which were lower than that of their corresponding symbiont-free aphid line. It should be noted that the λ value of either the symbiont-infected aphid or the symbiont-free line was greater than one, suggesting that the population studied may have increased continuously (Liang et al. 2021). How could the symbiont influence population dynamic of host aphid? One possible explanation could be an increase of the development time of 1st and 2nd-instars (DT1, DT2) as well as adult pre-reproductive period (APRP) of symbiont-infected S. avenae observed in the study, which is consistent with previous reports (Laughton et al. 2014; Luo et al. 2017b; Luo et al. 2020a). The increased development time may reduce or delay aphid reproduction but this could be counterbalanced by the benefit of an increase parasitoid resistance, suggesting a trade-off between resistance and fitness cost of aphids. In the field, symbiont-infected and symbiont-free aphids often occur side-by-side and are typical for real aphid populations, which are generally attributed to balancing selection (Oliver et al. 2014; Gimmi et al. 2023).

Indeed, the presence of symbiont reduced the parasitism success of wasps across different parasitoid densities. The conferred resistance of Regiella was confirmed by the clear reduction on the weight of emerged wasps and the emergence rate from symbiont-infected aphids. This result was expected, since previous reports have suggested that some Regiella strains should be regarded as a defensive symbiont in aphids that can adversely influence the normal development of parasitoid eggs/larvae by encoding pathogenicity factors such as RTX toxins (Vorburger et al. 2010; Hansen et al. 2012; Luo et al. 2022; Vorburger 2022). In addition, since previous studies have suggested that the presence of endosymbionts could impact the cellular and humoral immunity of host aphids (Schmitz et al. 2012; Luo et al. 2021), we here propose that parasitism success may also be related with host immune responses generated by the presence of Regiella.

Conclusions

Overall, this study documented that Regiella insecticola infection would lower aphid population growth and increase the resistance to the parasitoid Aphidius gifuensis and that these parameters were not highly affected by the parasitoid densities. The results suggest that there could be a possible compromise between the fitness cost of aphids and their resistance, which could offer a new outlook on the utilization of parasitoids in biological pest control by incorporating the effects of symbionts. This, in turn, could aid in the enhancement of integrated pest management (IPM) by reducing the reliance on pesticides and minimizing the negative impact on non-target insects. (Desneux et al. 2018).