Elevated CO₂ not increased temperature has specific effects on soil nematode community either with planting of transgenic Bt rice or non-Bt rice

Open-top chambers(OTCs)

The experiment was conducted in twelve open-top chambers (2.5 m in height and 4.2 m in diameter) at the Innovation Research Platforms for Climate Change, Biodiversity and Pest Management (CCBPM; http://www.ccbpm.org) in 2016 and 2017, which is located in Ningjin County, Dezhou City, Shandong Province of China (37.64°N, 116.8°E). This region has a warm and semi-humid monsoon climate, with an annual mean temperature of 12.9 °C and annual mean precipitation of 547.5 mm. The CO₂ concentrations were continuously monitored and adjusted with an infrared CO₂ analyzer (Ven-tostat 8102, Telaire Company, Goleta, CA, USA) at an interval of 20 min every day, and the temperature was measured three times a day using an automatic temperature analysis system (U23-001, HOBO Pro V2 Temp/RH Data Logger; MicroDAQ Ltd, Contoocook, NH; The accuracy of temperature was defined as ± 0.02 °C from 0 to 50 °C). Two levels of CO₂ concentration, including the ambient (A, 382 ± 4.02 µl/L) and the elevated (E, 754 ± 3.70 µl/L) levels, and two temperature levels, including the normal (25.57 ± 0.41 °C) and the increased (26.14 ± 0.43  °C) levels were applied continuously in the OTCs. Three blocks were used for CO₂ and temperature treatments and each block was split into four OTCs, i.e., one with ambient CO₂ concentration and increased temperature (A+T), one with ambient CO₂ concentration and normal temperature (A-T), one with elevated CO₂ concentration and increased temperature (E+T), and one with elevated CO₂ concentration and normal temperature (E-T).

Rice cultivars and planting

Transgenic Bt rice (cv. HH with fused Cry1Ab/Cry1Ac) and its near-isogenic parent line of non-Bt rice (cv. MH) were both provided by the College of Plant Science and Technology of Huazhong Agricultural University, Wuhan City, Hubei Province of China. These two cultivars of rice were planted in the plastic pots (32 cm in height and 24 cm in diameter; containing 10 kg sifted field soil) located in the OTCs on June 11 in 2016 and 2017 respectively. Thirty pots for each cultivar were placed randomly in each OTC. The potted field soil was collected from sifted field soil. The fields have been planted with conventional crops, such as corn, wheat, cotton, but transgenic crops had never previously been cultivated within 100 km. The chemical properties were as follows: pH 8.5, organic C 6.6 g kg⁻¹, total N 0.4 g kg⁻¹, alkaline hydrolysis N 7.9 g kg⁻¹, available P 14.4 mg kg⁻¹ and available K 96 mg kg⁻¹. The soil belong to fluvo-aquic soil with sandy texture. After growing for 30 days, the rice seedlings were thinning to twelve plants in each pot. To minimize the effect of microclimate, the pots were weekly rotated within the chambers. Normal cultural practices for rice cultivation, such as fertilization and irrigation, were followed except that no insecticides were applied during the entire experimental periods.

Soil sampling

Soil samples were collected in June (seedling stage) and October (harvesting stage) of 2016 and 2017 respectively, and in August (shooting stage) of 2017. Five sampling pots for each cultivar in each block were randomly selected. In each pot, four soil cores (2.5 cm in diameter) from 0 to 10 cm surface soil were randomly collected, and they were mixed together as a composite sample, and then kept in 4 °C for less than a week until identification. Before nematode extraction, plant tissues including root fragments were removed from the soil samples using a two mm mesh sieve.

Nematodes were extracted from 100 g soils using the minor modified Baermann method (Liu et al., 2008). All individuals of sampled nematodes were counted under the stereoscopic microscope (XTL-BM-7B), and about 100–150 specimens per soil sample were randomly selected, then identified to genera and assigned to four trophic groups: (1) herbivore; (2) bacterivore; (3) fungivore; (4) omnivore-predator (Yeates et al., 1993).

Nematode ecological index

The total number of nematodes at each sampling time was used as an index of abundance expressed as individuals per 100 g dry soil (Liu et al., 2018). Besides, the relative abundance of the dominant, common and rare groups were >10%, ≤10%, and <1% of the total nematodes, respectively (Liang, Zhang & Li, 2001). The ecological indices of nematode were evaluated by total maturity index (∑MI; a measure of disturbance), Shannon diversity (H′: an indicator of diversity index), nematode channel ratio (NCR: an indicator of the prevalence of organic matter decomposition), enrichment index (EI: an indicator of soil food web condition) and structure index (SI: a measure of food web length and connectance). The above indices were calculated using the following equations: (1)${\displaystyle \text{Total maturity index}\left(\sum MI\right):\sum MI=\sum v_i{f}_i}$ (2)${\displaystyle \text{Shannon diversity}\left(H^{\prime }\right):H^{\prime }=-\sum P_{i}ln{P}_i}$ (3)${\displaystyle \text{Nematode channel ratio (NCR)}:\mathrm{NCR}=\mathrm{Ba}∕\left(\mathrm{Ba}+\mathrm{Fu}\right)}$ (4)${\displaystyle \text{Enrichment index}\left(EI\right):EI=100\times \left(e∕\left(e+b\right)\right)}$ (5)${\displaystyle \text{Structure index}\left(SI\right):SI=100\times \left(s∕\left(s+b\right)\right)}$

Where, v_i was given c-p value based on life history strategies of free life and plant parasitic nematodes in ecological succession, f_i represents one genus proportion in total nematodes (Bongers, 1990); P_i is the proportion of individuals belongs to the ith taxon in the total number of individuals of nematode (Shannon & Weaver, 1949); Ba represents bacterivores and Fu represents fungivores; e (enrichment component), b (basal component) and s (structural component) are calculated using those guilds indicating enrichment (Ba₁, Fu₂), basal (Ba₂, Fu₂) and structure (Ba₃-Ba₅, Fu₃-Fu₅, Om₃-Om₅, Ca₂-Ca₅), respectively (Ferris, Bongers & De Goede, 2001).

Data analysis

All statistical analyses were performed with R software version R 3.0.3. Prior to statistical analysis, the Shapiro–Wilk and Levene’s tests were applied to evaluate data normality and homogeneity respectively. Nematode abundance was log(x+1) transformed and the percentage of nematode were arcsine square-root transformed for further statistical analysis, but untransformed means were presented in figures and tables. Variables were evaluated by use of split-plot analysis of variance for the repeated-measures analysis to measure the effects of rice variety (HH vs. MH), CO₂ (382 µl/L vs. 754 µl/L), temperature (25.57 °C vs. 26.14 °C) and sampling time (Jun and Oct in 2016, and Jun, Aug and Oct in 2017) and their interactions on the nematode community. The CO₂ and temperatue were assigned as main treatments, and variety of rice was assigned as a split-plot in the split-plot design. When there are interactive effects between sampling time and variety, then we compared the main and interactive treatment effects within each sampling time. Moreover, if the main effects and their interactions with different sampling season were significant, then we do one-way ANOVA to test the differences in these above parameters on each sampling date. The LSD test was used to analyze the significant differences between treatments at P <0.05. Non-metric multidimensional scaling (NMDS) was performed to determine which factors (rice variety, CO₂ concentration, temperature and sampling time) were markedly correlated with the NMDS ordination of soil nematode community by returning squared correlation coefficients (i.e., envfit function, Vegan package in R) and the Bray-Curtis distance analysis was used to evaluate the dissimilarity of nematode community composition across rice variety, CO₂ concentration, temperature and sampling time.

Community composition and dissimilarity comparison

A total of 34 genera of nematode were observed in the sampled soil collected during two consecutive years of 2016 and 2017. There was no difference in the nematode variables between transgenic Bt rice (cv. HH) and the parental line of non-Bt rice (cv. MH) in all the CO₂ and temperature conditions (i.e., A+T, A-T, E+T, E-T), except that Heterodera, Diplogaster and Mononchus were just found in the soil of transgenic Bt rice (cv. HH), and Hieschmanniella was only observed in the soil of non-Bt rice (cv. MH) (Fig. 1).

The results of NMDS consistently showed that variety, CO₂ concentration and temperature have no significant effects on nematode communities during 2016–2017 (Fig. 2).

Abundance

The treatments of rice variety, CO₂ concentration or temperature didn’t significantly affect the total nematode abundance (Table 1). Although the absolute abundances of four trophic groups were not significantly influenced by rice variety or temperature, the absolute abundance of fungivores in the condition of elevated CO₂ concentration was remarkably different from that in the condition of ambient CO₂ concentration. One-way ANOVA further showed that the absolute abundance of fungivores was significantly higher in the condition of elevated CO₂ concentration than that in the condition of ambient CO₂ concentration for HH-T treatment (i.e., transgenic Bt rice under the condition of normal temperature) in October of 2016 and 2017 and in August of 2017, and for MH+T treatment (i.e., the parental line of non-Bt rice under the condition of increased temperature) in October of 2016 and 2017 (P < 0.05; Figs. 3B–3C).

Ecological indices

There were no pronounced changes in the measured ecological indices of soil nematodes between rice variety treatments or temperature treatments (Table 1). However, elevated CO₂ concentration significantly influenced the ecological indices of NCR and EI. One-way ANOVA also showed that the ecological index of NCR was significantly lower in the condition of elevated CO₂ concentration than that in the condition of ambient CO₂ concentration for MH+T treatment in October of 2017, and for MH-T treatment (i.e., the parental line of non-Bt rice under the condition of normal temperature) in June and October of 2017 (P < 0.05; Figs. 4C–4D). The ecological index of EI was significantly lower in the condition of elevated CO₂ concentration than that in the condition of ambient CO₂ concentration for HH+T treatment (i.e., transgenic Bt rice under the condition of increased temperature) in October of 2016, and for HH-T treatment in June of 2017 (P < 0.05; Figs. 4E–4F). Besides, significant interaction between rice variety and sampling time on H′ and NCR were observed. Nevertheless, the treatment of rice variety or CO₂ or temperature or their interaction did not significantly influence the ecological indices of H′ and NCR within each sampling time (Table 2). One-way ANOVA further showed that the ecological indices of H′ and NCR were not significantly different between rice variety treatments within each sampling time except that in October of 2016 (Fig. 4). The split-plot analysis of variance and the nematode faunal analysis also showed markedly fluctuations among sampling time (Table 1; Fig. 5).

Rice cultivar effects on soil nematode

Our findings confirmed that transgenic Bt rice (cv. HH) didn’t remarkably change nematode community abundance and ecological indices under ambient CO₂ and temperature conditions. These results lead us to support the hypothesis (i) and these were consistent with previous studies that transgenic Bt plants have no detrimental effects on non-target soil fauna under ambient CO₂ concentration and temperature conditions (Romeis et al., 2019; Liu et al., 2018; Zhao et al., 2017). The Bt protein content was relatively low in soil (Chen et al., 2017b; Liu et al., 2017) and degraded rapidly in soils (Liu et al., 2018; Valldor et al., 2015), these findings may be likely responsible for above results. To our knowledge, few studies have reported the influences of transgenic Bt crops on non-target soil organisms under elevated CO₂ concentration and temperature, which may indirectly change the structure, activities and ecosystem functions of the soil biota, as well as the soil food webs (García-Palacios et al., 2015; Blankinship, Niklaus & Hungate, 2011). In the current study, we found that there were no significant differences between rice variety on nematode community composition under elevated CO₂ concentration or elevated temperature condition, which lead us to support the hypothesis (ii). We believed that the effects of transgenic Bt rice on soil nematode would not be influenced by global change.

Global change effects on soil nematode

Although elevated CO₂ concentration did not change the absolute abundance of herbivores, this condition remarkably increased the absolute abundance of fungivores. Thus, we can reject the hypothesis (iii). Additionally, the rising atmospheric CO₂ concentration often increases the rice root biomass (Hu et al., 2017; Yang et al., 2010) by enhancing carbon allocation belowground (Ainsworth & Long, 2005). Moreover, García-Palacios et al. (2015) reported that soil fungal abundance responding to elevated CO₂ concentration was positively correlated with plant biomass. Therefore, we speculated that a series of above variations may in turn indirectly affect the abundance of soil fungivores under elevated CO ₂condition. Hu et al. (2017) found that elevated CO₂ concentration remarkably increased the abundance of herbivores for the rice cultivar of IIYou084 and significantly reduced the the abundance of herbivores for the rice cultivar of Wuyuniing at the ripening stage. Thus, we assumed that the effect of elevated CO₂ concentration on the abundance of trophic groups was cultivar-specific.

The ecological indices were essential measures of indicating soil health conditions (Chen et al., 2017a; Ferris et al., 2012; Li et al., 2005). High values of NCR (>0.5) were found in all treatments of A+T, A-T, E+T and E-T, suggesting that bacterivores were dominant in organic matter decomposition for both transgenic Bt rice (cv. HH) and its parental line of non-Bt rice (cv. MH). Interestingly, regardless of rice variety or temperature, the ecological index of NCR was pronouncedly declined under the elevated CO₂ concentration, indicating that fungivores became increasingly important in degrading organic matter with the increase of CO₂ concentration. In the present study, we also found that the elevated CO₂ concentration remarkably decreased the ecological index of EI, and nematode faunal distributions under ambient CO₂ concentration were the largest in quadrant B, while those under elevated CO₂ concentration were primarily in quadrant C, irrespective of rice variety. Ferris, Bongers & De Goede (2001) inferred that the value of C/N ratio of the organic material in soil was higher in quadrant C than that in quadrant B according to nematode faunal analysis, and Sun et al. (2010) found that elevated CO₂ concentration could increase C/N ratio in plant tissues. Thus, we speculated that the impact of CO₂ concentration on nematode faunal distribution may be related to increased C/N ratio in plant tissues.

Dynamics of nematodes in controlled environment

The current study was conducted in open-top chambers, where the CO₂ concentration and temperature were tightly controlled. The experiment eliminated the influence of environmental variations present in realistic environment. However, sampling time still significantly changed the nematode composition, abundance and ecological indices. There were several reasons could interpret the fact of temporal dynamics in planted rice soil. First, the wetting-drying cycles caused by irrigation may alter the biological and biochemical activity (Chen et al., 2017b), and indirectly changed the nematode assemblage. Second, crop phenology may affect the population of soil nematode, and this finding agreed with data from previous reports on rhizospheric methanotroph community (Vishwakarma et al., 2009) and fungus-nematode (Bloomberg & Sutherland, 2010).