Arbuscular mycorrhizal fungi alter the food utilization, growth, development and reproduction of armyworm (Mythimna separata) fed on Bacillus thuringiensis maize

Plant materials and AMF inoculation

A two-year study (2017–2018) was conducted in Ningjin County, Shandong Province of China (37°38′30.7″N, 116°51′11.0″E). The Institute of Crop Sciences, Chinese Academy of Agricultural Science provided us with the transgenic Bt maize cultivar (Line IE09S034 with Cry1Ie, Bt) and its non-Bt parental line (cv. Xianyu 335, Xy). Glomus caledonium (strain number 90036, referred to as GC) was provided by the State Key Laboratory of Soil & Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences. The inoculum consisted of spores, mycelium, maize root fragments, and soil. Both genetically modified and non-modified maize were put into plastic buckets (45 cm height, 30 cm diameter) with 20 kg of soil sterilized in an autoclave, 300 g GC inoculum (GC inoculation treatment, ab. GC) and 300 g sterilized strains (control group, ab. CK) were evenly spread four cm below the maize seeds on June 10 in each sampling year. The whole experiment involved four treatments: two maize cultivars (Bt and Xy), and two AMF inoculations (GC and CK). Each bucket served as one replication and replicated 15 times for each treatment. So, there were 15 buckets for each maize cultivar × AMF inoculation treatment, and a total of 60 buckets in this study. In each bucket, three maize seeds were put at a depth of two cm. During the whole experimental period, no pesticides were applied and the manual weeding was performed to keep the maize buckets free from incidence of weeds.

AMF colonization

AMF colonization was determined on July 3 (seeding stage), August 25 (heading stage) and September 23 (harvest stage) in two sampling years. This was determined by the method of trypan blue staining and grid counting (Phillips & Hayman, 1970). The fresh plant roots were washed with distilled water and then blotted dry with absorbent paper. One hundred one cm roots were randomly cut and placed in a 10% KOH solution at 30 °C for 30 min, and then the KOH was discarded and rinsed with distilled water. After acidification in 2% HCl for 60 min, the HCl was discarded, rinsed with distilled water and stained in 5% trypan blue dye solution (w/v, lactic acid: glycerol: water = 1:1:1). Then the dye solution was discarded, and the roots were rinsed with distilled water and transferred to a square with a grid at the bottom. We observed the number of infected and uninfected root segments under the microscope. Colonization (%) = number of infected root segments/total root segments (McGonigle et al., 1990).

Insect rearing

The colony of armyworm M. separata was originated from a population collected in maize fields in Kangbao County, Hebei province of China (41.87°N, 114.6°E) in the summer of 2014. They were reared on the artificial diet (Bi, 1981) for more than 15 generations in climate-controlled growth chambers (GDN-400D-4; Ningbo Southeast Instrument Co., Ltd., Ningbo, China) at 26  ± 1 °C, 65  ± 5% RH, and 14: 10 h L/D photoperiod. The same rearing breeding conditions were maintained kept for the subsequent experiments. The newly-hatched first instar larvae were randomly selected from the above colony of M. separata and fed on the same artificial diet until the second instar larval stage, and then the third instar M. separata larvae were individually fed on excised leaves of the sampled maize plants. A certain amount of experimental maize leaves were randomly chosen from 10 buckets of each maize cultivar × AMF inoculation treatment beginning August 22 (heading stage) for the feeding trials conducted in plastic dish (six cm in diameter and 1.6 cm in height) and replace fresh maize leaves every 24 h until M. separata pupation in 2017 and 2018. Each maize cultivar × AMF inoculation treatment consisted of five replicates (30 larvae for one replicates).

Food utilization of M. separata larvae

The initial weights of the tested third instar larvae of M. separata were individually determined with an electronic balance (AL104; METTLER-TOLEDO, Greifensee). The weights of the total feces from the third instar until pupation (sixth instar), pupal weight, and the residual leaves were also carefully measured. At the same time, the moisture content of the third instar larvae, the sixth instar larvae and maize leaves replaced each time were determined to calculate the dry weight of the tested larvae and the maize leaves during the experiment. Several food utilization indexes of M. separata larvae fed on the excised leaves of Bt and non-Bt maize inoculated with AMF, G. caledonium and without G. caledonium, were determined. The indexes included RCR (relative consumption rate), RGR (relative growth rate), AD (approximate digestibility), ECD (efficiency of conversion of digested food) and ECI (efficiency of conversion of ingested food) (Li, Parajulee & Chen, 2018). The indexes calculations were done with formulas adapted from Chen et al. (2005): ${\displaystyle \mathrm{RCR}=I∕\left(B\ast T\right);\mathrm{RGR}=G∕\left(B\ast T\right);\mathrm{AD}\left(\text{\%}\right)=\left(I-F\right)∕I\ast 100\text{\%};}$ ${\displaystyle \mathrm{ECD}\left(\text{\%}\right)=G∕\left(I-F\right)\ast 100\text{\%};\mathrm{ECI}\left(\text{\%}\right)=G∕I\ast 100\text{\%}.}$ Where I is the feeding amount (the dry weight of maize leaves before feeding minus the dry weight of maize leaves before feeding after feeding); B is the average larval weight during the experiment (the average larval dry weight before feeding and after feeding); T is experiment time (d); G is the added larval weight (the larval dry weight after feeding minus the larval dry weight before feeding); F is the dry weight of total feces.

Growth & development and reproduction of M. separata

Larval growth and development were evaluated from the third instar to pupation by observing each petri dish every 8 h and recording the timing of larval ecdysis, pupation, and emergence of M. separata moths that fed on the excised leaves of Bt and non-Bt maize inoculated with G. caledonium and without G. caledonium. After the eclosion, novel moths were paired by maintaining the female: male ratio of 1: 1 in a metal screen cage and were fed with a 10% honey cotton ball, covered with cotton net yarn and butter paper for oviposition which were replaced every day. Survivorship and oviposition were recorded on a daily basis until death.

Yield of Bt and non-Bt maize

On September 25, 2017 and 2018, eight maize plants were randomly taken from five pots of each maize cultivar × AMF inoculation treatment at the harvest stage to measure the grain weight per plant (g) and 1,000-grain weight (g) with an electronic balance (AL104; METTLER-TOLEDO, Greifensee, Switzerland), in order to ascertain the effects of AMF inoculation on the yield of Bt and non-Bt maize inoculated with and without GC.

Data analysis

All experimental data were analysed with the software IBM-SPSSv.20.0 (IBM, Armonk, NY, USA). Three-way repeated-measures ANOVA was used to study the impacts of treatment (Bt maize vs. non-Bt maize), AMF inoculation (GC vs. CK), sampling years (2017 vs. 2018), and their bi- and tri-interaction on the AMF colonization. Moreover, three-way ANOVA was used to analyze the effects of treatment (Bt maize vs. non-Bt maize), AMF inoculation (GC vs. CK), sampling years (2017 vs. 2018), and their bi- and tri-interactions on the measured indexes of growth, development, reproduction and food utilization of M. separata, and the yield of Bt and non-Bt maize inoculated with and without GC in 2017 and 2018. Finally, the means were separated by using the Turkey test to examine significant difference between/among treatments at P < 0.05.

AMF colonization of Bt and non-Bt maize inoculated with and without G. caledonium

Colonization represents the infestation status of inoculated AMF, proving whether the access of AMF in maize is effective. Three-way repeated-measures ANOVAs indicated that GC inoculation (P < 0.001) and sampling year (P < 0.001) both significantly affected the AMF colonization, and there were significant interactions between GC inoculation with sampling year (P < 0.001), and between transgenic treatment with sampling year (P = 0.013 < 0.05; Table 1). Compared with the non-GC inoculation, the GC inoculation significantly enhanced the AMF colonization of Bt and non-Bt maize in 2017 and 2018 respectively, with significant increases for the Bt maize during the seedling (2017: +653.9%; 2018: +284.1%), heading (2017: +589.6%; 2018: +491.0%) and harvest (2017: +457.6%; 2018: +409.6%) stages in 2017 and 2018, and for the non-Bt maize during the seedling (2017: +613.6%; 2018: +432.2%), heading (2017: +472.8%; 2018: +425.7%) and harvest (2017: +437.1%; 2018: +448.7%) stages in 2017 and 2018 (P < 0.05; Fig. 1).

Food utilization of M. Separata larvae fed on Bt and non-Bt maize inoculated with and without G. caledonium

Food utilization indexes can reflect the preference and adaptability of insects to food materials to a certain extent. Transgenic treatment significantly affected all the measured indexes of feeding of M. separata larvae (P < 0.001), AMF inoculation (P < 0.05) and the interactions between transgenic treatment and AMF inoculation (P < 0.001) had important effects on the RGR, RCR, ECI and AD of M. separata larvae, and there were significant differences in the RGR, ECD, ECI and AD of M. separata larvae between the two sampling years (P < 0.01; Table 2). Moreover, there were significant interactions between transgenic treatment and sampling year on the RGR, RCR and AD of M. separata larvae (P < 0.05; Table 2). Furthermore, there were significant interactions between AMF inoculation and sampling year, and among transgenic treatment, AMF inoculation and sampling year on the RGR of M. separata larvae fed on the detached leaves of Bt maize and its parental line of non-Bt maize during the heading stage in 2017 and 2018.

When considering the case of M. separata larvae fed on Bt maize and non-Bt maize inoculated with GC in comparison with the CK in both sampling years, differing trends in the food utilization indexes were observed (Fig. 2). In relation to the CK (i.e., non-GC inoculation), GC inoculation significantly reduced the RGR (−24.2% and −23.3%), RCR (−10.5% and −6.1%), ECI (−15.3% and −18.2%) and AD (−9.0% and −16.4%) of M. separata larvae fed on the detached leaves of Bt maize during the heading stage in 2017 and 2018 (P < 0.05; Fig. 2). However, GC inoculation significantly enhanced the RGR (+36.9% and +56.7%), RCR (+10.8% and +15.0%), ECI (+19.9% and +26.4%) and AD (+17.2% and +19.3%) of M. separata larvae fed on the detached leaves of non-Bt maize during the heading stage in 2017 and 2018 (P < 0.05; Fig. 2).

Moreover, significant decreases in the RGR (2017: −64.2% and −35.4%; 2018: −68.6% and −35.9%), RCR (2017: −25.7% and −8.1%; 2018: −30.6% and −15.1%), ECD (2017: −32.2% and −25.4%; 2018: −22.1% and −15.5%), ECI (2017: −51.9% and −31.8%; 2018: −54.7% and −30.0%), and AD (2017: −29.2% and −8.7%; 2018: −41.8% and −17.1%) were found when M. separata larvae fed on the detached leaves of Bt maize in contrast to non-Bt maize as inoculated with and without GC in 2017 and 2018 (P < 0.05; Fig. 2). Additionally, the decreased percentages in the RGR, RCR, ECD, ECI and AD of M. separata larvae fed on the detached leaves of Bt maize were all obviously higher under GC inoculation in contrast to CK when compared with non-Bt maize.

Reproduction, growth and development of M. Separata fed on Bt or non-Bt maize inoculated with and without G. caledonium

The effects of a Bt and non-Bt maize diet on the growth, development and reproduction of M. Separata demonstrate the indirect effects of AMF on the suitability of M. Separata through Bt and non-Bt maize. Transgenic treatment (P < 0.05) and AMF inoculation (P < 0.05) significantly affected all the calculated indexes of M. separata in two sampling years, and there were significant difference in pupal weight (P < 0.001) of M. separata between two sampling years (Table 3). Moreover, there were significant interactions between transgenic treatment and sampling year on larval life-span (P = 0.008 < 0.01), between AMF inoculation and sampling year on larval life-span (P = 0.011 < 0.05), adult longevity (P = 0.044 < 0.05) and fecundity (P = 0.004 < 0.01), between transgenic treatment and AMF inoculation on all the measured indexes except larval life-span (P < 0.01), and among transgenic treatment, AMF inoculation and sampling year on larval life-span (P = 0.013 < 0.05) and fecundity (P = 0.009 < 0.01) for M. separata fed on the detached leaves of Bt maize and non-Bt maize inoculated with and without GC during the heading stage in 2017 and 2018 (Table 3).

Opposite trends were also seen in the calculated indexes for reproduction, growth, and development of larvae fed on the detached leaves of Bt and non-Bt maize inoculated with GC in contrast to the CK in both sampling years (Fig. 3). In comparison with the CK, GC inoculation significantly extended the larval life cycle (+7.6% and +10.4%) and shortened the adult longevity (−14.7% and −15.2%), and decreased the pupal weight (−9.1% and −14.1%) and fecundity (−19.2% and −19.9%) of larvae fed on the detached leaves of Bt maize in 2017 and 2018 (P < 0.05; Fig. 3). At the same time, GC inoculation significantly shortened the larval life-span (−12.3% and −10.3%) and prolonged the adult longevity (+24.6% and +17.1%), and significantly increased the pupation rate (+8.4% and +11.9%), pupal weight (+10.5% and +11.9%) and fecundity (+35.7% and +14.1%) of larvae fed on the detached leaves of non-Bt maize in 2017 and 2018, and significantly increased the eclosion rate (+16.3%) of larvae fed on the detached leaves of non-Bt maize in 2018 (P < 0.05; Fig. 3).

In comparison with the non-Bt maize, Bt maize significantly prolonged the larval life cycle (2017: +31.6% and +7.3%; 2018: +31.1% and +6.6%) and shortened the adult longevity (2017: −44.2% and −17.7%; 2018: −32.6% and −13.4%), and significantly decreased the pupation rate (2017: −32.6% and −27.7%; 2018: −31.1% and −22.0%), pupal weight (2017: −26.9% and −11.2%; 2018: −28.5% and −6.9%), eclosion rate (2017: −33.3% and −28.2%; 2018: −36.0% and −24.6%) and fecundity (2017: −48.1% and −12.8%; 2018: −43.0% and −18.7%) of larvae fed on the detached maize leaves inoculated with and without GC in 2017 and 2018 (P < 0.05; Fig. 3). Additionally, the larval life-span percentage increased, and percentages decreased in the pupation rate, pupal weight, eclosion rate, adult longevity and fecundity of larvae fed on the detached leaves of Bt maize were all obviously higher under GC inoculation in contrast to CK when compared with non-Bt maize.

Yields of Bt and non-Bt maize inoculated with and without G. caledonium

The increase in yields is our ultimate goal; that is, the inoculation of AMF has a corresponding indirect effects on the M. Separata feeding on Bt and non-Bt maize, and in this section, we evaluated the change of the final economic maize output after inoculation. AMF inoculation and sampling year significantly affected the grain weight per plant (P < 0.001), and 1,000-grain weight were significantly affected by AMF inoculation and transgenic treatment (P < 0.05). There were significant interactions between sampling year and transgenic treatment on the grain weight per plant (P = 0.041 < 0.05) and 1,000-grain weight (P = 0.001 < 0.01), and between sampling year and AMF inoculation on the grain weight per plant (P = 0.012 < 0.05) of Bt and non-Bt maize inoculated with and without GC inoculation in 2017 and 2018 (Table 1).

Compared with the CK, GC inoculation significantly increased the grain weight per plant (Bt maize: +39.6% and +24.1%; non-Bt maize: +33.1% and +30.6%) and 1,000-grain weight (Bt maize: +8.7% and +7.4%; non-Bt maize: +8.5% and +8.5%) of Bt and non-Bt maize in 2017 and 2018 (P < 0.05; Fig. 4).