Influence of elevated CO₂ on development and food utilization of armyworm Mythimna separata fed on transgenic Bt maize infected by nitrogen-fixing bacteria

Abstract

Background

Bt crops will face a new ecological risk of reduced effectiveness against target-insect pests owing to the general decrease in exogenous-toxin content in Bt crops grown under elevated carbon dioxide (CO₂). The method chosen to deal with this issue may affect the sustainability of transgenic crops as an effective pest management tool, especially under future atmospheric CO₂ level raising.

Methods

In this study, rhizobacterias, as being one potential biological regulator to enhance nitrogen utilization efficiency of crops, was selected and the effects of Bt maize (Line IE09S034 with Cry1Ie vs. its parental line of non-Bt maize Xianyu 335) infected by Azospirillum brasilense (AB) and Azotobacter chroococcum (AC) on the development and food utilization of the target Mythimna separate under ambient and double-ambient CO₂ in open-top chambers from 2016 to 2017.

Results

The results indicated that rhizobacteria infection significantly increased the larval life-span, pupal duration, relative consumption rate and approximate digestibility of M. separata, and significantly decreased the pupation rate, pupal weight, adult longevity, fecundity, relative growth rate, efficiency of conversion of digested food and efficiency of conversion of ingested food of M. separata fed on Bt maize, while here were opposite trends in development and food utilization of M. separata fed on non-Bt maize infected with AB and AC compared with the control buffer in 2016 and 2017 regardless of CO₂ level.

Discussion

Simultaneously, elevated CO₂ and Bt maize both had negative influence on the development and food utilization of M. separata. Presumably, CO₂ concentration arising in future significantly can increase their intake of food and harm to maize crop; however, Bt maize infected with rhizobacterias can reduce the field hazards from M. separata and the application of rhizobacteria infection can enhance the resistance of Bt maize against target lepidoptera pests especially under elevated CO₂.

Introduction

With increased fossil fuel combustion and drastic changes in land utilization, the concentration of atmospheric carbon dioxide (CO₂) has increased by more than 40%, from 280 to 400 ppm, between the industrial revolution and now (Ciais et al., 2013). The recent forecast indicated that atmospheric CO₂ concentration will increase to approximately 900 ppm by 2,100 (Intergovernmental Panel on Climate Change (IPCC) 2014). Increasing atmospheric CO₂ concentration alone can be very significant in crop production because of its direct effect on plant physiology and biochemistry (Cornelissen, 2011), and indirect effect on tri-trophic interactions involving plants, herbivores, and predators or pathogens (Robinson, Ryan & Newman, 2012; Trębicki et al., 2017). Elevated atmospheric CO₂ also affects the crop production via direct or indirect impact on the physiology and feeding behavior of phytophagous insects (Zvereva & Kozlov, 2006; Massad & Dyer, 2010; O’Neill et al., 2010). These changes may then lead to more severe and frequent outbreaks of pest insects in agricultural ecosystems (Percy et al., 2002).

Several studies have shown that the elevated CO₂ increased lepidopteran insect feeding and damage severity in agricultural crops (Ainsworth et al., 2007; Lindroth et al., 2001), because of the increased proportion of C:N in host plant tissue and lower nutritional quality caused by elevated CO₂ (Ainsworth et al., 2007). For example, larvae of Helicoverpa armigera fed on wheat grown in elevated CO₂ showed the extended larval life span and increased consumption with reduced growth rate (Chen, Wu & Ge, 2004). Transgenic maize that expresses insecticidal Cry proteins derived from the soil bacterium Bacillus thuringiensis Berliner (Bt) has been used to control target lepidopteran insects (Carrière, Crowder & Tabashnik, 2010; Huang et al., 2014; Walters et al., 2010), e.g., European corn borer Ostrinia nubilalis (Hübner), Asian corn borer O. furnacalis (Guenée) (Lepidoptera: Crambidae) and corn armyworm Mythimna separata (Lepidoptera: Noctuidae) (Guo et al., 2016; Zhang et al., 2013; Jia et al., 2016). Transgenic Bt maize has widely been adopted worldwide (Cattaneo et al., 2006; Huang et al., 2005; Hutchison et al., 2010; Lu et al., 2012). It was anticipated that the primary effect of elevated CO₂ on Bt toxin production would be due to differences in N concentration in plant tissues (Coviella, Stipanovic & Trumble, 2002). Biologically relevant changes in plant defensive chemistry of Bt maize are expected to have measurable effects on the target lepidopteran pests under climate change.

Additionally, many researchers found that the nitrogen metabolism of transgenic Bt crops could affect the expression of Bt toxin protein, and stimulating plant N uptake to increase in biomass N relative to C to increase the nitrate reductase activity and Bt toxin production of Bt crops. (Stitt & Krapp, 1999; Pang et al., 2005; Gao et al., 2009). Nitrogen plays the most important role for plant growth, and it is an important complement of enzymes catalyzing and controlling reactions in plants for normal physiological processes (Richardson et al., 2009). While most of nitrogen in the environment is found in a form of nitrogen gas (N₂) which approximately amounts to 78% in the atmosphere, plant available nitrogen found in soil is generally derived from fertilizer augmentation. As plants cannot use N₂ directly, soil-inhabiting microbes play a significant role in nitrogen uptake by plants as they change the N₂ into ammonia (Yamprai, Mala & Sinma, 2014). Azospirillum sp. and Azotobacter sp. are the two major free-living soil microbes (Biari, Gholami & Rahmani, 2008), that are economically important nitrogen-fixing bacteria in maize crop production system (Yamprai, Mala & Sinma, 2014). Thus, optimization of soil-nitrogen management offers significant potential in the utilization of soil rhizobacterias to increase Bt-crop nitrogen utilization to affect the expression of Bt toxin under elevated CO₂.

Materials and Methods

Setup of CO₂ levels

A two-year study (2016–2017) was conducted in six open-top chambers (i.e., OTCs; Granted Patent: ZL201120042889.1; 2.5 m in height × 3.2 m in diameter) (Chen et al., 2011) at the Innovation Research Platforms for Climate Change, Biodiversity and Pest Management (CCBPM; http://www.ccbpm.org) field laboratory in Ningjin County, Shandong Province of China (37°38′ 30.7″ N, 116°51′ 11.0″ E). A total of two CO₂ levels, ambient (375 μl/L, hereafter referred to as aCO₂) and elevated (750 μl/L or double-ambient, hereafter referred to as eCO₂) were applied continuously from 10 June to 7 October in both years. A total of three OTCs were used for each CO₂ treatment, and the CO₂ concentrations in each OTC were monitored continuously and adjusted using an infrared CO₂ analyzer (Ventostat 8102; Telaire Company, Goleta, CA, USA). The OTCs of elevated CO₂ treatments were inflated with canned CO₂ gas with 95% purity and automatically controlled by the same type of infrared CO₂ analyzer (Chen, Wu & Ge, 2004). Actual mean CO₂ concentrations and temperature throughout the entire experiment for both 2016 and 2017 are provided in Table 1.

Table 1. Actual mean (±SE) CO₂ concentration and temperature in the open-top chambers (OTC) from seedling emergence to harvest of transgenic Bt maize and its parental line of non-Bt maize in 2016 and 2017.

Climate factors	OTC	2016	2017	Two-way ANOVAs (F/P values)
CO2 (μl/L)	eCO2 OTC	744.4 ± 3.3a, A	748.8 ± 4.5a, A	FCO2 = 399.32, P = 0.000
				FYear = 2.85, P = 0.13
	aCO2 OTC	372.6 ± 4.7b, A	374.5 ± 3.8b, A	FInteraction = 0.45, P = 0.52
Temperature (°C)	eCO2 OTC	26.01 ± 0.5a, A	26.12 ± 0.3a, A	FCO2 = 0.006, P = 0.94
				FYear = 3.38, P = 0.10
	aCO2 OTC	25.99 ± 0.4a, A	26.11 ± 0.4a, A	FInteraction = 0.057, P = 0.82

Notes:

OTC: ambient-CO₂ OTC (aCO₂ OTC) and elevated-CO₂ OTC (eCO₂ OTC). Different lowercase letters indicate significantly different between the eCO₂ OTC and aCO₂ OTC in same year by the Duncan test at P < 0.05, respectively.

^ANot significantly different between 2016 and 2017 at same CO₂ level or temperature by the Duncan test at P > 0.05, respectively.

Plant materials

The Bt maize cultivar (Line IE09S034, hereafter referred to as Bt) and its non-Bt parental line (cv. Xianyu 335, referred to as Xy) were both obtained from the Institute of Crop Sciences, Chinese Academy of Agricultural Sciences. Both Bt and non-Bt lines used in this study had the similar maturity (approximately 102 d: from 10 June to 20 September) and were well adapted to the growing conditions of northern China (Guo et al., 2016; Zhang et al., 2013; Jia et al., 2016; Ling, 2010). Both maize accessions were planted in plastic buckets (diameter × height = 30 × 45 cm) filled with 20 kg autoclaved soil and 10 g compound fertilizer (N:P:K = 18:15:12), then placed them into chambers on 10 June each year.

Soil nitrogen-fixing bacteria and infection of maize seeds

Lyophilized Azospirillum brasilense (strain number ACCC 10103) and Azotobacter chroococcum (strain number ACCC 10006) were provided by Agriculture Culture Collection of China (ACCC) in plastic tubes (3 cm in diameter and 15 cm in height) with bacterial growth medium. Both species of rhizobacterias were grown in liquid medium at 28 °C under continuous shaking (200 rpm) until they reached an absorbance of 1.008 (A. brasilense) and 1.005 (A. chroococcum) at a wavelength of 600 nm. Before inoculation, the culture was centrifuged, and the supernatant was discarded, and the pellet of cells was re-suspended in the liquid medium to a density of 10⁸ copies per milliliter. The seeds of both Bt and non-Bt maize were infected with A. brasilense and A. chroococcum cultures each, and the inoculation doses were all adjusted to a final volume of 10 ml for each seed. After inoculation, all the treated seeds were maintained under sterile laminar air flow for 2 h at 28 °C (Cassán et al., 2009). Bacteria inoculation treatments consisted of three types of rhizobacteria infection, including (1) seeds infected with A. brasilense (referred to as AB); (2) seeds infected with A. chroococcum (referred to as AC); and (3) non-infected seeds (control) treated with a final volume of buffer solution (referred to as CK). The entire experiment, thus, consisted of 12 treatments, including two CO₂ levels (aCO₂ and eCO₂), two maize cultivars (Bt and Xy), and three rhizobacteria infections (AB, AC, and CK), replicate six time. Each pot serves as one replication. Specifically, six buckets for each maize cultivar (Bt and Xy) and three rhizobacteria inoculations (6 buckets per transgenic treatment × 2 transgenic treatments × 3 inoculation treatments = 36 buckets) were placed randomly in each CO₂ chamber (ambient and double-ambient CO₂), and three maize seeds were sown in each bucket at 2 cm soil depth. No pesticides were applied during the entire experimental period and the manual weeding keep the maize buckets weed-free during the experiment. The rhizosphere soil was sampled from each bucket one-day before planting, 14 days after planting, and at harvest and measured the relative density of A. brasilense and A. chroococcum using RT-PCR (Tables 2 and 3) (Jiang et al., 2017).

Table 2. Sequence specific primers of rhizobacterias, Azospirillum brasilense (AB) and Azotobacter chroococcum (AC) for qRT-PCR.

Primer	Sequence (5′-3′)	GenBank accession	Description
AB-4	Forward: CAAGGGCACCATCCCGAC	X51500.1	A. brasilense NifH gene
	Reverse: CTGCTGCTCCTCCGACT
AC-2	Forward: GTGACCCGAAAGCTGACTCC	EU693338.1	A. chroococcum nifH gene
	Reverse: CCACCTTCAGCACGTCTTCC

Table 3. The rhizosphere soil densities of rhizobacterias inoculated in the potted soil of transgenic Bt maize and its parental line of non-Bt maize grown under ambient and elevated CO₂ in 2016 and 2017.

Measure matters	Rhizobacteria infections		2016 (AB; AC copies/g)	2017 (AB; AC copies/g)
Sampled soil before maize planting			5.53 ± 0.24 105; 4.47 ± 0.12 105	5.61 ± 0.11 105; 4.33 ± 0.17 105
Sampled soil at the maize seedling after 14 days	AB	aCO2-Bt	8.46 ± 0.24 1011; 4.48 ± 0.26 105	8.40 ± 0.28 1011; 4.44 ± 0.11 105
		aCO2-Xy	8.25 ± 0.26 1011; 4.21 ± 0.08 105	8.69 ± 0.23 1011; 4.56 ± 0.22 105
		eCO2-Bt	8.36 ± 0.19 1011; 4.43 ± 0.15 105	8.59 ± 0.21 1011; 4.47 ± 0.17 105
		eCO2-Xy	8.70 ± 0.27 1011; 4.58 ± 0.29 105	8.24 ± 0.12 1011; 4.34 ± 0.27 105
	AC	aCO2-Bt	5.54 ± 0.25 105; 7.37 ± 0.29 1011	5.70 ± 0.28 105; 7.40 ± 0.26 1011
		aCO2-Xy	5.73 ± 0.24 105; 7.29 ± 0.17 1011	5.36 ± 0.22 105; 7.66 ± 0.25 1011
		eCO2-Bt	5.62 ± 0.30 105; 7.71 ± 0.15 1011	5.13 ± 0.04 105; 7.32 ± 0.13 1011
		eCO2-Xy	5.46 ± 0.28 105; 7.59 ± 0.17 1011	5.42 ± 0.13 105; 7.57 ± 0.22 1011
	CK	aCO2-Bt	5.71 ± 0.20 105; 4.52 ± 0.21 105	5.92 ± 0.08 105; 4.67 ± 0.17 105
		aCO2-Xy	5.50 ± 0.29 105; 4.24 ± 0.15 105	5.33 ± 0.18 105; 4.31 ± 0.13 105
		eCO2-Bt	5.46 ± 0.08 105; 4.26 ± 0.18 105	5.62 ± 0.31 105; 4.48 ± 0.21 105
		eCO2-Xy	5.46 ± 0.18 105; 4.76 ± 0.23 105	5.47 ± 0.17 105; 4.21 ± 0.09 105
Sampled soil at the maize harvest	AB	aCO2-Bt	8.50 ± 0.19 1011b; 4.65 ± 0.21 105	8.39 ± 0.26 1011b; 4.01 ± 0.26 105
		aCO2-Xy	8.44 ± 0.15 1011b; 4.11 ± 0.23 105	8.65 ± 0.19 1011b; 4.30 ± 0.18 105
		eCO2-Bt	9.81 ± 0.23 1011a; 4.13 ± 0.17 105	1.09 ± 0.04 1012a; 4.67 ± 0.20 105
		eCO2-Xy	9.98 ± 0.25 1011a; 4.49 ± 0.22 105	1.02 ± 0.03 1012a; 4.89 ± 0.23 105
	AC	aCO2-Bt	5.57 ± 0.31 105; 7.27 ± 0.26 1011b	5.40 ± 0.08 105; 7.30 ± 0.14 1011b
		aCO2-Xy	5.99 ± 0.25 105; 7.49 ± 0.19 1011b	4.97 ± 0.15 105; 7.24 ± 0.19 1011b
		eCO2-Bt	4.89 ± 0.27 105; 8.98 ± 0.15 1011a	5.94 ± 0.14 105; 9.07 ± 0.12 1011a
		eCO2-Xy	5.33 ± 0.10 105; 8.96 ± 0.21 1011a	5.77 ± 0.12 105; 9.03 ± 0.18 1011a
	CK	aCO2-Bt	5.15 ± 0.35 105; 4.65 ± 0.23 105	5.39 ± 0.08 105; 4.56 ± 0.22 105
		aCO2-Xy	5.49 ± 0.19 105; 4.37 ± 0.33 105	4.97 ± 0.16 105; 4.66 ± 0.15 105
		eCO2-Bt	5.59 ± 0.14 105; 4.76 ± 0.11 105	4.88 ± 0.25 105; 4.64 ± 0.17 105
		eCO2-Xy	5.12 ± 0.14 105; 4.69 ± 0.05 105	5.13 ± 0.13 105; 4.24 ± 0.10 105

Note:

Rhizobacteria infections: A. brasilense (AB) and A. chroococcum (AC) vs. the control buffer solution (CK). CO₂ levels: ambient CO₂ (aCO₂) and elevated CO₂ (eCO₂). Transgenic treatment: Bt maize (Bt) and non-Bt maize (Xy). Different lowercase letters indicate significantly different between ambient CO₂ and elevated CO₂ for same maize cultivar in same year by the Duncan test at P < 0.05, respectively.

Insect source and 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, and fed on artificial diet and maintained for more than 10 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 conditions were maintained for the following experiments. Newly-hatched larvae were randomly selected from the above colony of M. separata and fed on artificial diet (Bi, 1981) until the second instar larvae, and then the third instar larvae were individually fed on excised leaves of the experimental plants growing in CO₂ chambers. Feeding trials were conducted in plastic dish (6 cm in diameter and 1.6 cm in height) and the experimental leaves were randomly selected from six buckets for each of the 12 experimental treatment combinations (2 transgenic treatments × 2 CO₂ treatments × 3 bacteria inoculations) during the tasseling stage until pupation. Sample size for the M. separata larval feeding trial consisted of 20 larvae (sample unit size) with five replicates for each of the 12 treatment combinations (i.e., 1,200 larvae evaluated for the entire study). Because of the cannibalism among the late instar larvae of M. separata (Jiang et al., 2016; Ali et al., 2016; Liu et al., 2017), the sampled larvae were reared separately in the Petri dish until pupation.

Development and reproduction of M. separata

Larval development was evaluated from third instar to pupation by way of observing each individual petri dish every 8 h and recording the timing of larval ecdysis, pupation, and emergence of M. separata moths. After eclosion, the newly emerged moths were paired (female: male = 1:1) for mating in a metal frame screen cage (length × width × height = 35 × 35 × 40 cm), and the paired moths were fed with a 10% honey solution provided on a large cotton wick in a single plastic cup (diameter × height = 8 × 20 cm) covered with cotton net yarn butter paper for oviposition. The cotton net yarn and butter paper were replaced every day. Moth survivorship and oviposition were recorded daily until both moths from each pair died.

Food utilization of the larvae of M. separata

Each third instar test larvae of M. separate was weighed at the initiation of the feeding trial by using an electronic balance (AL104; METTLER-TOLEDO, Greifensee, Switzerland). Total accumulated feces from third instar until the larva entered pupal stage (sixth instar), sixth instar larval weight, and the remaining leaves were also weighed. The food utilization indices of M. separata included the relative growth rate (RGR), relative consumption rate (RCR), approximate digestibility (AD), efficiency of conversion of ingested food (ECI) and efficiency of conversion of digested food (ECD) (Chen, Ge & Parajulee, 2005a; Chen et al., 2005b). Formulas for calculation of the measured indices were adapted from Chen et al. (2005b).

Data analysis

All data were analyzed using the statistical software SPSS 19.0 (2015; SPSS Institute, Chicago, IL, USA). Four-way analysis of variance was used to analyze the effects of CO₂ levels (elevated vs. ambient), transgenic treatment (Bt maize vs. non-Bt maize), rhizobacteria infection (AB and AC vs. CK), sampling years (2016 vs. 2017), and the interactions on the measured indices of growth, development, and reproduction, including larval life-span, pupation rate, pupal weight, pupal duration, adult longevity and fecundity of M. separata. The measured food utilization indices were analyzed by using an analysis of covariance with initial weight of M. separata (i.e., third instar larva) as a covariate for RCR and RGR, while food consumption was a covariate for ECI and AD to correct the effect of variation in the growth and food assimilation of M. separata (Raubenheimer & Simpson, 1992); food assimilated was also used as a covariate to analyze the ECD parameter (Hägele & Rowell-Rahier, 1999). The assumption of a parallel slope between covariate and dependent variable was satisfied for each analysis. Treatment means were separated by using the Duncan-test to examine significant difference at P < 0.05.

Results

Effects of CO₂ level, transgenic treatment, and rhizobacteria infection on the rhizosphere soil densities of A. brasilense and A. chroococcum in different sampling period

Significant effects of rhizobacteria infection (P < 0.001) were observed on the measured rhizosphere soil densities of A. brasilense (AB) and A. chroococcum (AC) 14 days after maize planting. Compared with ambient CO₂, elevated CO₂ significantly increased the rhizosphere soil densities of both A. brasilense and A. chroococcum; compared with the control buffer solution (CK), rhizobacteria infection significantly increased the rhizosphere soil densities of A. brasilense and A. chroococcum (P < 0.001; Table 3). CO₂ level and rhizobacteria infection both significantly affected the densities of A. brasilense and A. chroococcum in rhizosphere soil at maize harvest (Table 4).

Table 4. Four-way ANOVA on the rhizosphere soil densities of rhizobacterias inoculated in the potted soil of Bt maize and its parental line of non-Bt maize grown under ambient and elevated CO₂ in 2016 and 2017.

Impact factors	Sampled soil at the maize seedling after 14 days		Sampled soil at the maize harvest
	AB	AC	AB	AC
Ya	0.00/0.99	0.002/0.96	0.97/0.33	1.13/0.29
Cv.b	0.32/0.574	0.36/0.55	0.070/0.79	0.080/0.78
CO2c	0.19/0.89	0.80/0.38	26.01/<0.001***	331.16/<0.001***
Rhizobacteriad	2555.00/<0.001***	1380.37/<0.001***	1311.83/<0.001***	2080.71/<0.001***
Y × Cv.	0.62/0.44	1.96/0.17	0.18/0.673	0.53/0.47
Y × CO2	1.21/0.28	2.50/0.12	0.74/0.40	0.032/0.86
Y × Rhizobacteria	0.00/1.00	0.02/0.99	0.97/0.39	1.13/0.33
Cv. × CO2	0.35/0.55	0.010/0.92	0.32/0.57	0.36/0.55
Cv. × Rhizobacteria	0.32/0.73	0.36/0.70	0.070/0.93	0.80/0.93
CO2 × Rhizobacteria	0.019/0.98	0.80/0.45	26.01/<0.001***	331.16/<0.001***
Y × Cv. × CO2	5.99/0.018*	0.06/0.94	0.82/0.37	0.63/0.43
Y × Cv. × Rhizobacteria	0.62/0.54	1.96/0.15	0.18/0.84	0.53/0.59
Y × CO2 × Rhizobacteria	1.21/0.31	2.50/0.093	0.74/0.49	0.032/0.97
Cv. × CO2 × Rhizobacteria	0.35/0.70	0.10/0.99	0.32/0.73	0.36/0.70
Y × Cv. × CO2 × Rhizobacteria	5.99/0.05	0.006/0.99	0.82/0.45	0.63/0.54

Notes:

^*P < 0.05;

^**P < 0.01;

^***P < 0.001.

^aYear (2016 vs. 2017).

^bTransgenic treatment (Bt maize vs. non-Bt maize).

^cCO₂ levels (elevated CO₂ vs. ambient CO₂).

^dRhizobacteria infection (A. brasilense and A. chroococcum vs. the control buffer), the same as in Tables 5 and 7.

Effects of CO₂ level, transgenic treatment, and rhizobacteria infection on the development and reproduction of M. separata

Carbon dioxide level and transgenic treatment both significantly affected the larval life-span, pupation rate, pupal weight and duration, adult longevity, and fecundity in M. separata fed on both Bt and non-Bt maize infected with A. brasilense and A. chroococcum (P < 0.001). However, the rhizobacteria infection significantly affected the larval life-span, pupal duration (P < 0.05) and fecundity (P < 0.001) of M. separata fed on both transgenic treatments and at both CO₂ levels (Table 5).

Table 5. Four-way ANOVA on the development and reproduction of Mythimna separata fed on Bt and non-Bt maize infected with A. brasilense and A. chroococcum under ambient and elevated CO₂ in 2016 and 2017.

Impact factors	Larval life-span (day (n = 828))	Pupation rate (% (n = 828))	Pupal weight (g (n = 663))	Pupal duration (day (n = 663))	Adult longevity (day (n = 576))	Fecundity (eggs per female (n = 198))
Ya	1.62/0.32	0.94/0.71	1.21/0.62	0.11/0.90	1.65/0.33	2.99/0.10
Cv.b	1662.03/<0.001***	329.16/<0.001***	275.04/<0.001***	229.78/<0.001***	450.27/<0.001***	2032.99/<0.001***
CO2c	62.13/<0.001***	55.53/<0.001***	12.44/0.001**	19.02/<0.001***	41.62/<0.001***	278.70/<0.001***
Rhizobacteriad	3.64/0.034*	2.53/0.077	1.20/0.63	7.04/0.011*	0.35/0.70	27.54/<0.001***
Y × Cv.	9.22/0.004**	2.22/0.14	2.22/0.14	1.98/0.32	3.10/0.054	0.102/0.75
Y × CO2	5.33/0.025*	2.16/0.20	0.067/0.80	0.17/0.85	13.53/<0.001***	2.74/0.10
Y × Rhizobacteria	3.63/0.034*	1.69/0.058	0.63/0.57	0.48/0.71	0.59/0.64	1.87/0.053
Cv. × CO2	19.11/<0.001***	6.62/<0.001***	2.68/0.008**	14.11/<0.001***	8.26/0.006**	5.86/0.049*
Cv. × Rhizobacteria	224.53/<0.001***	73.67/<0.001***	30.41/<0.001***	42.06/<0.001**	26.38/<0.001***	195.08/<0.001***
CO2 × Rhizobacteria	12.81/<0.001***	12.22/0.004**	11.63/0.005**	14.98/<0.001***	4.24/0.02*	22.02/<0.001***
Y × Cv. × CO2	0.01/0.92	1.07/0.55	0.10/0.75	1.16/0.52	0.006/0.94	4.25/0.085
Y × Cv. × Rhizobacteria	1.55/0.33	1.53/0.17	0.064/0.94	2.02/0.23	0.46/0.68	8.87/0.071
Y × CO2 × Rhizobacteria	0.063/0.69	0.51/0.76	0.48/0.62	1.72/0.45	1.78/0.18	14.61/0.045*
Cv. × CO2 × Rhizobacteria	8.88/0.006**	12.61/0.003**	7.41/0.011*	5.66/0.005**	13.55/0.002**	24.04/0.000***
Y × Cv. × CO2 × Rhizobacteria	0.19/0.83	0.33/0.71	0.14/0.87	0.32/0.72	1.73/0.30	0.36/0.70

Notes:

^*P < 0.05;

^**P < 0.01;

^***P < 0.001.

^aYear (2016 vs. 2017).

^bTransgenic treatment (Bt maize vs. non-Bt maize).

^cCO₂ levels (elevated CO₂ vs. ambient CO₂).

^dRhizobacteria infection (A. brasilense and A. chroococcum vs. the control buffer), the same as in Tables 4 and 7.

Compared with ambient CO₂, elevated CO₂ significantly prolonged the larval life-span (+6.21%), pupal duration (+5.56%), and significantly decreased the pupation rate (−18.08%), pupal weight (−8.12%), adult longevity (−6.06%), and fecundity (−22.58%) of M. separata (P < 0.05; Table 6). Also, compared with the CK, rhizobacteria infection with A. brasilense and A. chroococcum both significantly shortened the larval life-span (−5.20% and −5.70%), pupal duration (−3.68% and −3.81%) and fecundity (−10.20% and −9.53%) of M. separata (P < 0.05; Table 6). Moreover, Bt maize significantly prolonged the larval life-span (+13.67%) and pupal duration (+7.54%), shortened the adult longevity (−10.41%), and decreased the pupation rate (−75.55%), pupal weight (−13.54%) and fecundity (−75.46%) of M. separata compared to that for non-Bt maize (P < 0.05; Table 6).

Table 6. The development and reproduction of M. separata larvae fed on Bt maize and non-Bt maize during the heading stage, infected with rhizobacterias under ambient and elevated CO₂ in 2016 and 2017.

Impact factors	Factor levels	Larval life-span (day)	Pupation rate (%)	Pupal weight (g)	Pupal duration (day)	Adult longevity (day)	Fecundity (eggs per female)
Cv.	Bt	24.19 ± 0.16a	40.98 ± 2.45b	0.192 ± 0.007b	10.98 ± 0.18a	6.63 ± 0.14b	171.50 ± 13.27b
	Xy	21.28 ± 0.17b	71.94 ± 2.38a	0.218 ± 0.006a	10.21 ± 0.19b	7.32 ± 0.09a	300.92 ± 28.64a
CO2	Elevated	23.42 ± 0.20a	51.78 ± 2.69b	0.197 ± 0.007b	10.78 ± 0.18a	6.77 ± 0.11b	212.25 ± 19.13b
	Ambient	22.05 ± 0.17b	61.14 ± 2.68a	0.213 ± 0.005a	10.41 ± 0.19b	7.18 ± 0.13a	260.17 ± 21.87a
Rhizobacteria infection	AB	22.53 ± 0.16b	56.06 ± 2.24	0.205 ± 0.004	10.54 ± 0.09b	6.97 ± 0.24	228.00 ± 15.27b
	AC	22.42 ± 0.15b	55.53 ± 2.01	0.204 ± 0.004	10.53 ± 0.08b	6.96 ± 0.24	229.38 ± 17.46b
	CK	23.25 ± 0.16a	57.79 ± 2.58	0.206 ± 0.005	10.72 ± 0.13a	6.99 ± 0.18	251.25 ± 23.21a

Note:

Data in table are average ± SE. Different lowercase letters indicate significant difference between treatments by the Duncan’s test at P < 0.05; the same as in Table 8.

Impacts of CO₂ level, transgenic treatment, and rhizobacteria infection on the food utilization of M. separata

There were significant effects of CO₂ level, transgenic treatment, and rhizobacteria infection (P < 0.01 or P < 0.001) on food utilization of M. separata fed on both Bt and non-Bt maize infected with A. brasilense and A. chroococcum at both CO₂ levels in both years of the study (Table 7).

Table 7. Four-way ANCOVA on the food utilization indices of Mythimna separata fed on Bt maize and non-Bt maize infected with A. brasilense and A. chroococcum under ambient and elevated CO₂ in 2016 and 2017.

Impact factors	The 3rd to 6th instar larvae (n = 828)
	RGR (mg g−1day−1)	RCR (mg g−1day−1)	AD (%)	ECD (%)	ECI (%)
Covariatee	1.81/0.11	3.83/0.067	0.781/0.23	0.580/0.41	0.87/0.32
Ya	1.19/0.31	3.96/0.12	4.56/0.061	4.82/0.059	5.80/0.053
Cv.b	1545.53/<0.001***	302.67/0.<0.001***	185.62/0.<0.001***	716.17/0.<0.001***	1038.95/0.<0.001***
CO2c	67.09/<0.001***	27.98/<0.001***	9.69/0.003**	35.90/<0.001***	57.98/<0.001***
Rhizobacteriad	12.26/<0.001***	26.84/<0.001***	35.28/<0.001***	13.64/<0.001***	7.22/0.002**
Y × Cv.	4.27/0.049*	5.69/0.021*	1.86/0.18	19.21/<0.001***	15.64/<0.001***
Y × CO2	6.90/0.012*	4.82/0.033*	6.73/0.013*	6.22/0.016*	3.41/0.071
Y × Rhizobacteria	5.04/0.010*	0.17/0.84	0.27/0.77	0.436/0.65	0.25/0.78
Cv. × CO2	30.44/<0.001***	7.39/0.009**	1.40/0.043*	1.80/0.017*	1.61/0.011*
Cv. × Rhizobacteria	213.46/<0.001***	48.31/<0.001***	73.42/<0.001***	132.82/<0.001***	144.91/<0.001***
CO2 × Rhizobacteria	13.25/<0.001***	6.70/0.003**	9.78/<0.001***	13.01/<0.001***	12.63/<0.001***
Y × Cv. × CO2	0.220/0.64	1.83/0.18	4.16/0.047*	0.743/0.39	0.56/0.46
Y × Cv. × Rhizobacteria	3.27/0.047*	0.55/0.58	2.23/0.12	4.01/0.025*	1.41/0.25
Y × CO2 × Rhizobacteria	0.72/0.49	1.88/0.16	1.33/0.28	2.66/0.080	2.47/0.095
Cv. × CO2 × Rhizobacteria	0.62/0.043*	13.24/<0.001***	9.84/<0.001***	9.59/<0.001***	9.92/<0.001***
Y × Cv. × CO2 × Rhizobacteria	0.48/0.62	0.087/0.92	0.98/0.38	0.59/0.56	0.06/0.94

Notes:

^*P < 0.05;

^**P < 0.01;

^***P < 0.001.

^aYear (2016 vs. 2017).

^bTransgenic treatment (Bt maize vs. non-Bt maize).

^cCO₂ levels (elevated CO₂ vs. ambient CO₂).

^dRhizobacteria infection (A. brasilense and A. chroococcum vs. the control buffer), the same as in Tables 4 and 5.

^eInitial weight as a covariate for RGR and RCR, and food consumption as a covariate for AD and ECI, and food assimilated as a covariate for ECD.

Compared with ambient CO₂, elevated CO₂ significantly reduced the RGR (−9.95%), ECD (−16.05%), and ECI (−17.95%), and significantly enhanced the RCR (+10.44%) and AD (+5.59%) of M. separata (P < 0.05; Table 8). Compared with the CK, rhizobacteria infection with A. brasilense and A. chroococcum both significantly decreased the ECD (−9.28% and −7.48%) and ECI (−9.22% and −7.91%), and significantly increased the RGR (+4.75% and +5.56%), RCR (+6.78% and +7.53%) and AD (+5.28% and +4.93%) in M. separata (P < 0.01; Table 8). Moreover, significant decreases in RGR (−13.85%), ECD (−41.25%) and ECI (−31.97%), and significant increases in RCR (+16.60%) and AD (+7.88%) were found when M. separata fed on Bt maize compared to that on non-Bt maize (P < 0.05; Table 8).

Table 8. The food utilization of M. separata larvae fed on Bt maize and non-Bt maize during the heading stage, infected with rhizobacterias under ambient and elevated CO₂ in 2016 and 2017.

Impact factors	Factor levels	RGR (mg g−1day−1)	RCR (mg g−1day−1)	AD (%)	ECD (%)	ECI (%)
Cv.	Bt	82.79 ± 6.43b	1636.31 ± 13.23a	56.13 ± 0.72a	9.26 ± 0.98b	5.13 ± 0.57b
	Xy	94.26 ± 7.16a	1403.36 ± 14.80b	52.03 ± 0.69b	13.08 ± 1.23a	6.77 ± 0.66a
CO2	Elevated	84.33 ± 8.67b	1595.24 ± 16.14a	55.55 ± 0.87a	10.34 ± 1.52b	5.46 ± 0.76b
	Ambient	92.72 ± 6.59a	1444.43 ± 15.01b	52.61 ± 0.81b	12.00 ± 1.21a	6.44 ± 0.63a
Rhizobacteria infection	AB	89.64 ± 5.52a	1549.07 ± 9.32a	55.06 ± 0.85a	10.78 ± 1.35b	5.75 ± 0.67b
	AC	90.34 ± 5.52a	1559.90 ± 9.32a	54.88 ± 0.85a	10.96 ± 1.35b	5.82 ± 0.67b
	CK	85.58 ± 6.14b	1450.73 ± 9.54b	52.30 ± 0.62b	11.78 ± 1.13a	6.28 ± 0.66a

Interactive influence of CO₂ level, transgenic treatment, and rhizobacteria infection on growth, development and reproduction of M. separata

In addition to the significant main effects of CO₂ level, transgenic treatment, and rhizobacteria infection, there were significant two-way and three-way interaction of these three main effects on larval life-span, pupation rate, pupal weight and duration, adult longevity, and fecundity of M. separata fed on Bt and non-Bt maize infected with A. brasilense and A. chroococcum under both CO₂ levels in both years of the study (P < 0.05, P < 0.01 or P < 0.001; Table 5).

Transgenic treatment × CO₂

Similar trends were found in the measured growth, development and reproduction indexes of M. separata fed on both Bt and non-Bt maize cultivars grown under elevated CO₂ in contrast to ambient CO₂, infected with A. brasilense (AB) and A. chroococcum (AC) as well as the CK in 2016 and 2017 (Figs. 1A–1F). Compared with ambient CO₂, elevated CO₂ significantly prolonged the larval life-span (Bt maize: +6.44%; non-Bt maize: +8.39%) and pupal duration (non-Bt maize: +7.27%) and shortened the adult longevity (non-Bt maize: −6.19%), and significantly decreased the pupation rate (non-Bt maize: −20.81%), pupal weight (Bt maize: −7.03%; non-Bt maize: −13.73%) and fecundity (Bt maize: −29.43%; non-Bt maize: −18.85%) when M. separata fed on Bt maize and non-Bt maize (P < 0.05; Figs. 1A–1F).

Figure 1. Effects of bi-interactions between transgenic treatment and CO₂, between transgenic treatment and rhizobacteria and between CO₂ and rhizobacteria on development and reproduction of Mythimna separata.

Larval life-span–(A), (G), (M); Pupation rate–(B), (H), (N); Pupal weight–(C), (I), (O); Pupal duration–(D), (J), (P); Adult longevity–(E), (K), (Q); Fecundity–(F), (L), (R); Each value represents the average (±SE). Different lowercase letters indicate significant differences treatments by the Duncan test at P < 0.05.

Transgenic treatment × Rhizobacteria

An inverse trend was found in the measured growth, development and reproduction indexes of M. separata fed on Bt maize and non-Bt maize, which were infected with A. brasilense (AB) and A. chroococcum (AC) under ambient and elevated CO₂ in 2016 and 2017 (Figs. 1G–1L). Compared with the CK, rhizobacteria infection significantly prolonged the larval life-span (AB: +7.63%; AC: +8.45%), pupal duration (AB: +4.53%; AC: +5.08%) and shortened the adult longevity (AB: −4.88%; AC: −6.94%), and decreased pupation rate (AB: −20.83%; AC: −30.81%), pupal weight (AB: −7.24%; AC: −10.65%) and fecundity (AB: −48.36%; AC: −64.60%) when M. separata larvae fed on Bt maize (P < 0.01; Figs. 1G–1L); and rhizobacteria infection significantly shortened the larval life-span (AB: −7.97%; AC: −10.15%), pupal duration (AB: −5.36%; AC: −6.08%) and prolonged the adult longevity (AB: +3.95%; AC: +5.62%), and increased pupation rate (AC: +9.73%), pupal weight (AB: +6.27%; AC: +8.16%) and fecundity (AB: +9.75%; AC: +16.16%) when M. separata larvae fed on non-Bt maize (P < 0.01; Figs. 1G–1L).

CO₂ × Rhizobacteria

Similar trends were found in the larval life-span, pupation rate, pupal weight, and pupal duration, while inverse trends were observed in adult longevity and fecundity of M. separata under ambient and elevated CO₂, which fed on Bt maize vs. non-Bt maize infected with A. brasilense (AB) and A. chroococcum (AC) as well as the CK in 2016 and 2017 (Figs. 1M–1R). Compared with the CK, rhizobacteria infection significantly shortened the larval life-span (AB: −6.92%; AC: −7.84%) and pupal duration (AB: −5.01%; AC: −5.01%) of M. separata under ambient CO₂, and significantly shortened the larval life-span (AB: −4.98%; AC: −5.11%) and decreased the pupal weight (AC: −4.77%) under elevated CO₂ (P < 0.05; Figs. 1M–1P); and rhizobacteria infection significantly decreased the adult longevity (AB: −4.63%; AC: −5.09%) and fecundity (AB: −22.90%; AC: −22.58%) of M. separata under elevated CO₂ (P < 0.05; Figs. 1Q and 1R).

Transgenic treatment × CO₂ × Rhizobacteria

There were opposite trends in the measured growth, development and reproduction indexes of M. separata fed on Bt maize and non-Bt maize infected with A. brasilense (AB) and A. chroococcum (AC) compared with the CK in 2016 and 2017 regardless of CO₂ level (Fig. 2). In comparison with the CK, rhizobacteria infection with A. brasilense and A. chroococcum both significantly prolonged the larval life-span and pupal duration of M. separata fed on Bt maize, and significantly shortened the larval life-span and pupal duration of M. separata fed on non-Bt maize under the same CO₂ level; and rhizobacteria infection with A. brasilense and A. chroococcum both significantly reduced the pupation rate, pupal weight, adult longevity and fecundity of M. separata fed on Bt maize, and significantly enhanced the pupation rate, pupal weight, adult longevity and fecundity of M. separata fed on non-Bt maize under the same CO₂ level. Moreover, compared with ambient CO₂, there were opposite trends in the larval life-span, pupal weight, pupal duration, adult longevity and fecundity of M. separata fed on Bt maize infected with A. brasilense and A. chroococcum compared with the CK under elevated CO₂ in both years; compared with ambient CO₂, elevated CO₂ significantly decreased the pupation rate of M. separata fed on Bt maize, and decreased the pupation rate, pupal weight, adult longevity and fecundity of M. separata fed on non-Bt maize, and prolonged the larval life-span and pupal duration of M. separata fed on non-Bt maize infected with A. brasilense and A. chroococcum compared with the CK in both years.

Figure 2. Impacts of the tri-interactions among CO₂ level, transgenic treatment, and rhizobacteria infection on the growth, development and reproduction of M. separata in 2016 (A–F) and 2017 (G–L).

Each value represents the average (+SE). Different lowercase and uppercase letters, and *indicated significant difference among three types of rhizobacteria infection for same type of maize under same CO₂ level, between Bt maize and non-Bt maize for same type of rhizobacteria infection under same CO₂ level, and between ambient and elevated CO₂ for same type of maize and rhizobacteria infection by the Duncan test at P < 0.05 respectively.

Interactive effects of CO₂ level, transgenic treatment, and rhizobacteria infection on food utilization of M. separata

In addition to significant main effects of CO₂ level, transgenic treatment, and rhizobacteria infection, two- and three-way interactions of these factors influenced the RGR, RCR, AD, ECD, and ECI of M. separata larvae fed on Bt maize and non-Bt maize infected with A. brasilense and A. chroococcum under ambient and elevated CO₂ in both years (P < 0.05, P < 0.01 or P < 0.001; Table 7).

Transgenic treatment × CO₂

Similar trends were found in the measured food utilization indexes of M. separata fed on Bt maize (Bt) and non-Bt maize (Xy) grown under elevated CO₂ in contrast to ambient CO₂, infected with A. brasilense (AB) and A. chroococcum (AC) as well as the CK in 2016 and 2017 (Figs. 3A–3E). Compared with ambient CO₂, elevated CO₂ significantly decreased the RGR (non-Bt maize: −7.34%), ECD (Bt maize: −9.67%; non-Bt maize: −10.25%) and ECI (Bt maize: −8.53%; non-Bt maize: −8.89%), and significantly increased the RCR (Bt maize: +9.69%; non-Bt maize: +6.37%) when M. separata larvae fed on Bt maize and non-Bt maize (P < 0.05; Figs. 3A–3E).

Figure 3. Effects of bi-interactions between transgenic treatment and CO₂, between transgenic treatment and rhizobacteria and between CO₂ and rhizobacteria on food utilization of Mythimna separata larvae.

RGR–(A), (F), (K); RCR–(B), (G), (L); AD–(C), (H), (M); ECD–(D), (I), (N); ECI–(E), (J), (O); Each value represents the average (±SE). Different lowercase letters indicate significant differences treatments by the Duncan test at P < 0.05.

Transgenic treatment × Rhizobacteria

Inverse trend was found in the measured food utilization indexes of M. separata fed on Bt maize and non-Bt maize, which were infected with A. brasilense (AB) and A. chroococcum (AC) under ambient and elevated CO₂ in 2016 and 2017 (Figs. 3F–3J). Compared with the CK, rhizobacteria infection significantly enhanced the RGR (AB: +9.53%; AC: +11.78%), ECD (AB: +11.61%; AC: +19.79%) and ECI (AB: +10.08%; AC: +15.79%), and significantly decreased the RCR (AC: −6.52%) and AD (AC: −6.19%) when M. separata larvae fed on non-Bt maize (P < 0.001; Figs. 3F–3J); and rhizobacteria infection significantly decreased the RGR (AB: −9.62%; AC: −10.41%), ECD (AB: −34.32%; AC: −41.55%) and ECI (AB: −20.16%; AC: −25.28%), and significantly increased the RCR (AB: +14.99%; AC: +19.06%) and AD (AB: +9.60%; AC: +10.79%) when M. separata larvae fed on Bt maize (P < 0.001; Figs. 3F–3J).

CO₂ × Rhizobacteria

Similar trends were observed in RGR, RCR, and AD, while inverse trends were shown in ECD and ECI of M. separata under ambient and elevated CO₂, which fed on Bt maize and non-Bt maize infected with A. brasilense (AB) and A. chroococcum (AC) vs. CK (Figs. 3K–3O). Compared with the CK, rhizobacteria infection significantly decreased ECD (AB: −20.71%; AC: −22.07%) and ECI (AB: −12.77%; AC: −12.89%) of M. separata larvae under elevated CO₂, and significantly increased ECD (AB: +5.35%; AC: +8.04%) and ECI (AB: +7.43%; AC: +9.85%) of M. separata larvae under ambient CO₂ (P < 0.05; Fig. 3); and rhizobacteria infection significantly enhanced RGR (AB: +3.32% and +7.40%; AC: +5.14% and +8.67%), RCR (AB: +9.78% and +5.29%,; AC: +11.32% and +5.93%) and AD (AB: +7.34% and +4.18%; AC: +7.92% and +4.66%) under elevated and ambient CO₂, respectively (P < 0.01; Figs. 3K–3O).

Transgenic treatment × CO₂ × Rhizobacteria

There were opposite trends in the measured food utilization indexes of M. separata larvae fed on Bt maize (Bt) and non-Bt maize infected with A. brasilense (AB) and A. chroococcum (AC) compared with the CK in both years regardless of CO₂ level (Fig. 4). In comparison with the CK, rhizobacteria infection with A. brasilense and A. chroococcum both significantly decreased RGR, ECD, and ECI of M. separata fed on Bt maize, and significantly increased RGR, ECD, and ECI of M. separata fed on non-Bt maize under the same CO₂ level; and rhizobacteria infection with A. brasilense and A. chroococcum both significantly enhanced RCR and AD of M. separata fed on Bt maize, and significantly reduced RCR and AD of M. separata larvae fed on non-Bt maize under the same CO₂ level. Moreover, compared with ambient CO₂, elevated CO₂ significantly increased RCR and AD, and significantly decreased RGR, ECD, and ECI of M. separata larvae fed on same type of maize cultivar infected with A. brasilense and A. chroococcum in both years (P < 0.05; Fig. 4). Furthermore, there were significant decreases in RGR, ECD, and ECI, and significant increases in RCR and AD of M. separata larvae fed on Bt maize in contrast to non-Bt maize infected with same type of rhizobacteria species within the same CO₂ level in both years.

Figure 4. Impacts of the tri-interactions among CO₂, transgenic treatment, and rhizobacteria infection on the food utilization of M. separata from the third to the sixth instar larvae in 2016 (A–E) and 2017 (F–J).

Each value represents the average (+SE). Different lowercase and uppercase letters, and *indicated significant difference among three types of rhizobacteria infection for same type of maize under same CO₂ level, between Bt maize and non-Bt maize for same type of rhizobacteria infection under same CO₂ level, and between ambient and elevated CO₂ for same type of maize and rhizobacteria infection by the Duncan test at P < 0.05 respectively.

Discussion

Insects are sensitive to environmental variations, and environmental stresses can cause changes on their growth, development, fecundity, food utilization and the occurrence and distribution of populations as a result of metabolic rate fluctuation (Bloom et al., 2010). In this study, elevated CO₂ significantly prolonged larval and pupal duration and decreased pupation rate and pupal weight of M. separata compared to ambient CO₂. Elevated CO₂ negatively affected the larval survival, weight, duration, pupation, and adult emergence of cotton bollworm, H. armigera (Akbar et al., 2016), and reduced the egg laying by Cactus moth Cactoblastis cactorum (Stange, 1997) and Achaea Janata (Rao et al., 2013). In this study, elevated CO₂ significantly increased the RCR (+10.44%) and the AD (+5.59%) (i.e., AD), and significantly reduced the RGR (−9.95%), ECD (−16.05%) and ECI (−17.95%) of M. separata larvae compared with ambient CO₂. RGRs of Gypsy moth (Lymantria dispar) were reported to be reduced by 30% in larvae fed on Quercus petraea exposed to elevated CO₂ (Hattenschwiler & Schafellner, 2004). RCR was significantly higher for H. armigera larva fed maize grown at 375 and 750 ppm CO₂ in contrast to ambient CO₂ condition, and elevated CO₂ significantly decreased the ECI food, the ECD food, and the RGR of H. armigera larvae compared with ambient CO₂ (Yin et al., 2010). According to the “Nutrition compensation hypothesis,” elevated CO₂ can affect the development fitness of herbivores by changing the nutritional components, above and below-ground biomass, and photosynthetic rate of host plants indirectly (Ainsworth & Rogers, 2007; Jackson et al., 2009; Zavala, Nabity & Delucia, 2013), including increased C/N ratio and decreased nitrogen content etc. Declined growth rate, reproduction, and survival rate were found in the chewing mouthparts insects (e.g., H. armigera, Spodoptera exigua, M. separata), and the food consumption of which increased so that they could obtain necessary nutrition to survive (Bottomley, Rogers & Prior, 1993; Rogers et al., 2006). Yin et al. (2010) reported that elevated CO₂ increased the food consumption and prolonged the development time of H. armigera, which due to the reduced nutritional quality of maize leaves, as a result of reduced nitrogen content and increased C/N ratio. Elevated CO₂ significantly reduced the food conversion rate and enhanced the food ingestion of H. armigera, which attribute to reduced nitrogen content of the cotton, Simian-3 (Chen, Ge & Parajulee, 2005a; Chen et al., 2005b). Thus, Chen, Ge & Parajulee (2005a) and Chen et al. (2005b) inferred that elevated CO₂ might be unfavorable to H. armigera. Our results in maize system appear to be similar to the study by Chen, Ge & Parajulee (2005a) and Chen et al. (2005b) in a cotton system.

Although the transgenic corn, Zea mays L., hybrids expressing the Cry insecticidal protein from Bacillus thuringiensis (Bt) were developed to control H. zea, O. nubilalis, S. frugiperda, and M. separata (Koziel et al., 1993; Armstrong et al., 1995; Jouanin et al., 1998; Lynch, Plaisted & Warnick, 1999), few studies focused on the defense responses of transgenic cry1Ie maize to corn armyworm under elevated CO₂, especially on the growth, development and food utilization of the pest insects. Prutz & Dettner (2005) reported that the transgenic Bacillus thuringiensis-maize could result in decreased growth rate and increased mortality, which might attribute to the termination of larval metamorphosis. Most studies showed that adverse effects on life-table parameters of different herbivores were direct by the Cry protein (Lawo, Wäckers & Romeis, 2010), which might be due to the interaction of feeding inhibitors and growth inhibitors (e.g., secondary plant substances) (Smith & Fischer, 1983). Effects of elevated CO₂ on the plant nutrition, metabolism and secondary defense metabolism might adverse for the growth, development and nutrition utilization of herbivores (Akbar et al., 2016). The insects possessed more nutrients to meet their growth needs and prolong the food digestion time in the midgut so that the RCR and AD increased (Reynolds, Nottingham & Stephens, 1985). In this study, we found that some negative effects of transgenic cry1Ie maize (Bt) and Xianyu 335 (Xy) grown in elevated CO₂ on the food utilization indices (including RGR, ECD, and ECI) of M. separata larvae and some positive effects on the RCR and AD, which indicated that the resistance responses of Bt maize might persist under elevated CO₂, and M. separata might ingest more food to get enough nutrition for surviving in limited developmental time under elevated CO₂. Meanwhile the Bt maize and its parental line (Xianyu 335) prolonged their larval life-span and pupal duration, decreased growth rate and increased mortality that might result in lowering of pests’ occurrence. According to the “carbon nutrition balance hypothesis” (Gebauer, Strain & Reynolds, 1997), elevated CO₂ would increase the fixed organic matter in plant while increase C-based secondary metabolites and decrease N-based secondary metabolites, thus affecting the insects resistance of plants. Robinson, Ryan & Newman (2012) indicated that elevated CO₂ increased 19% phenols, 22% condensed tannins, and 27% flavonoids, while the terpenoids and NBSC decreased by 13% and 16% respectively. Coviella, Stipanovic & Trumble (2002) anticipated that the primary CO₂ effect on Bt toxin production would be due to differences in N concentration within the plant. In a meta-analytical review of 33 studies that simultaneously increased CO₂ conditions compared to ambient conditions, Zvereva & Kozlov (2006) showed that nitrogen concentration in plants was reduced under elevated CO₂, and this decrease was stronger for woody compared to herbaceous plants. If conditions of increased carbon (e.g., elevated CO₂) allow plants to allocate significantly more resources to condensed tannins and gossypol, then the enzyme composition in the insect herbivore is expected to also change. Similarly, if Bt toxin production changes due to elevated CO₂, then the insect herbivore’s body enzymes should also be changed in this circumstance.

Most of the nitrogen, however, is found in the form of N₂ which approximately amounts to 78% in the atmosphere. As plants cannot use this form of nitrogen directly, some microbes can change the N₂ into ammonia. Most free living microbes in soil which can fix nitrogen and whose activities in enhancing the growth of plants are bacteria namely Azotobacter sp. and Azospirillum sp. These two bacteria are particularly important in maize production system due to their greater nitrogen fixing ability. Azospirillum acquires carbohydrate directly from sieve tube as a resource of carbon which promotes its growth (Olivera et al., 2004). Azospirillum can be used to promote the growth of sprouts under normal and arid conditions (Alejandra et al., 2009). Azospirillum also provides more flexibility to cell wall which enhances the growth (Pereyra et al., 2010) and increases products of wheat in waterless plot of land (Martin, 2009). Furthermore, azospirillum had the highest efficiency in nitrogen fixation at the root of sweet corn and it would reach the highest point of nitrogen fixation in the week 4 amounting to 0.20 mgNhr⁻¹m⁻² (Toopakuntho, 2010). Azospirillum can also create auxin, a substance promoting growth of maize, of 53.57 mg/ml (Phookkasem, 2011). Therefore, we used techniques of rhizobacteria (A. brasilense and A. chroococcum) inoculation of maize seeds to stimulate plant N uptake to increase in biomass N relative to C under elevated CO₂, increase Bt toxin production for transgenic cry1Ie maize and create a substance promoting maize plant growth. In this study, we found that elevated CO₂ significantly enhanced the rhizosphere soil densities both A. brasilense and A. chroococcum at the maize harvest, but there was no significant difference of the rhizosphere soil densities both A. brasilense and A. chroococcum between elevated and ambient CO₂ at the maize seedling after 14 days. We hypothesize that the elevated CO₂ increased the maize root bifurcation and soil nutrition (e.g., carbohydrates, amino acids and multi-trace elements) for rhizobacteria to provide the living space and nutrition with a long-time environmental effect. Other researchers have also shown positive effects of elevated CO₂ on the bacterial community in the rhizosphere of maize (Chen et al., 2012). Moreover, significant adverse effects on the growth, development, reproduction, and food utilization of M. separata were observed when the host substrate maize was exposed to rhizobacteria treatments, which might be attributed to rhizobacteria stimulating plant N uptake to increase Bt toxin production for transgenic cry1Ie maize and promoting growth of its parental line (Xianyu 335) (Olivera et al., 2004; Stitt & Krapp, 1999).

There was no significant year-to-year variation in our field research data. Therefore, the overall results clearly indicate that increasing CO₂ had negative effects on M. separata. Resistance performance of transgenic cry1Ie maize decreased under elevated CO₂ as shown by decreased RGR, ECD, and ECI. The rhizobacteria treatments (A. brasilense and A. chroococcum) had positive effects on improving the effectiveness of Bt maize on target Lepidoptera pest management via decreased RGR, ECD, and ECI of M. separata that fed on transgenic cry1Ie maize and promoting growth of Xianyu 335 via increased RGR, ECD, and ECI of M. separata. Under future predicted climate changes (e.g., elevated CO₂), it is particularly important to understand the field insect resistance traits of resistant crops to target pests. In an environment of accelerated greenhouse effect, Bt maize may have decreased resistance performance in the field with inhibiting effect on the development and food utilization of insects. Therefore, we used techniques of rhizobacteria (A. brasilense & A. chroococcum) inoculation of maize seeds to stimulate plant N uptake to increase in biomass N relative to C under elevated CO₂, increase Bt toxin production for transgenic cry1Ie maize, and create a substance promoting maize growth.

Conclusion

Overall, our results indicated that elevated CO₂ and Bt maize were negative against development and food utilization of M. separata. Rhizobacteria infection significantly increased the larval life-span, pupal duration, RCR and AD of M. separata, and significantly decreased RGR, ECD and ECI of M. separata fed on Bt maize; there were opposite trends in development and food utilization of M. separata fed on non-Bt maize infected with rhizobacterias compared with the CK in 2016 and 2017 regardless of CO₂ level. This study demonstrates that the use of rhizobacteria (e.g., A. brasilense and A. chroococcum) as pest control enhancer especially under elevated CO₂ is significantly more beneficial in transgenic Bt maize system compared to that in non-transgenic system. Rhizobacteria (A. brasilense & A. chroococcum), as being one potential biological regulator to enhance nitrogen utilization efficiency of crops, could make the Bt maize facing lower field hazards from the target pest of M. separate, and finally improve the sustainability and resistance of Bt maize against target lepidoptera pests, especially under future CO₂ raising.

Funding Statement

This research was funded by the National Nature Science Foundations of China (NSFC) (31272051), the National Key Research and Development Program of China (2017YFD0200400), the Special Program for New Transgenic Variety Breeding of the Ministry of Science and Technology, China (2016ZX08012005), and the Research Grant from the Innovation Project for Graduate Student of Jiangsu Province (KYLX16_1063). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.