AMF colonization and community of a temperate invader and co-occurring natives grown under different CO₂ concentrations for 3 years

Abstract

Global changes such as atmospheric CO₂ enrichment often facilitate exotic plant invasions and alter soil arbuscular mycorrhizal fungi (AMF) community. However, it is still unclear whether the effects of CO₂ enrichment on exotic plant invasions are associated with its effects on root-AMF symbiosis of invasive and native plants. To address this issue, the annual invasive plant Xanthium strumarium and two phylogenetically related annual natives were compared under ambient and elevated CO₂ concentrations for three consecutive years. Atmospheric CO₂ enrichment increased AMF colonization rates for the species only in few cases, and the invader did not benefit more from CO₂ enrichment in terms of AMF colonization. Under ambient CO₂ concentration, however, the invader had a higher AMF colonization rate than the natives in the first year of the study, which disappeared in the second and third year of the study due to the increase of AMF colonization rates in the natives but not in the invader. The influences of species, CO₂ concentrations and planting year on AMF colonization were associated with their effects on both soil nutrient and AMF community, and the former may be more important as it also influenced the latter. Our results indicate that the invader could more quickly form symbiosis with soil AMF, contributing to adaptation and occupation of new habitats, and that it is necessary to consider the roles of AMF and the effects of time when determining the effects of global changes such as atmospheric CO₂ enrichment on exotic plant invasions.

摘要

连续3年不同CO₂浓度下一种温带外来入侵植物和两种共存本地植物丛枝菌根真菌群落及侵染率

大气CO₂浓度升高等全球变化过程不仅能促进外来植物入侵，也能改变土壤丛枝菌根真菌(AMF)的群落结构，但我们并不清楚大气CO₂浓度升高促进外来植物入侵是否与其对外来入侵植物和 本地植物AMF共生的影响有关。为回答这一问题，我们在环境和倍增CO₂浓度下连续3年栽培一年生外来入侵植物瘤突苍耳(Xanthium strumarium)与两种共存的一年生本地近缘植物，比较了AMF侵染率、土 壤养分和土壤AMF群落组成的差异。研究结果表明，大气CO₂浓度升高只在少数情况下提高根系AMF侵染率，并且瘤突苍耳AMF侵染率的提高并不比本地种多。在环境CO₂浓度下，栽培第一年瘤突苍耳的AMF侵染侵染率显著高于两种本地植物；而栽培第二年和第三年与两种本地植物的差异不显著，因为两种本地植物的AMF侵染率随种植时间的增加而增加，而瘤突苍耳AMF侵染率受种植时间的影响较小。物种、CO₂浓度和种植时间对AMF侵染率的影响与它们对土壤养分和AMF群落的影响有关，土壤养分对AMF侵染率的影响可能比AMF群落组成的影响更大，因为后者也受土壤养分的影响。上述结果表明，与本地植物相比，入侵植物能更快地与AMF形成共生关系，有利于其适应和入侵新生境；在探究全球变化如大气CO₂浓度升高等对外来植物入侵的影响时，需要考虑AMF的影响和时间效应。

Open in new tabDownload slide

INTRODUCTION

Exotic plant invasions are results of the interactions between exotic plants and biotic and/or abiotic environments in the recipient habitats (Zhao et al. 2020). It has been documented that arbuscular mycorrhizal fungi (AMF) play important roles in the invasions of many exotic plant species (Bunn et al. 2015; de Souza et al. 2019; Dierks et al. 2019; Kong et al. 2022). AMF could facilitate plants to absorb soil nutrients and water, and improve their adaptation to stressful conditions such as barren and heavy metal polluted environments (Brundrett and Tedersoo 2018; Thirkell et al. 2020). As a major component of global changes, exotic plant invasions are also influenced by other components of global changes such as atmospheric carbon dioxide enrichment (eCO₂) (Liu et al. 2017; Ren et al. 2021). It has been recognized that eCO₂ could aggravate invasions of exotic plants (Lei et al. 2012; Liu et al. 2017). However, it is still unclear whether the effects of eCO₂ on exotic plant invasions are associated with its effects on symbiotic relationships between exotic plants and AMF.

Invasive plants can alter structures and functioning of soil AMF communities, and therefore influence symbiotic relationships between co-occurring native plants and AMF, suppressing soil resource uptake of native plants and contributing to exotic plant invasions (de Souza et al. 2019; Dierks et al. 2019; Grove et al. 2017). Dierks et al. (2019) found that Bromus tectorum invasion declines growth of the native Artemisia tridentate, which is associated with its effects on soil AMF community. de Souza et al. (2019) found that Prosopis juliflora invasion alters soil AMF community composition, and this alteration is beneficial to the invader while decreases AMF colonization in roots of a co-occurring native plant. Exotic plants can influence soil AMF by releasing allelochemicals. For example, the invasive plant Impatiens gladulifera inhibits the colonization of native plants by both ectomycorrhizal and arbuscular mycorrhizal fungi via releasing napthoquinone (Ruckli et al. 2014). In addition, exotic plant invasions often increase soil nutrient availabilities (Castro-Díez et al. 2014), which may decrease soil AMF richness and abundance (Grove et al. 2017; Johnson et al. 2010).

Many studies showed that eCO₂ is beneficial to AMF proliferation, increasing its abundance, colonization and hyphal density and length in roots (Alberton et al. 2005; Saleh et al. 2020; Sun et al. 2017; Wei et al. 2019). Sun et al. (2017) found that the mean AMF abundance of Quercus mongolica is increased by 11.5% after 10 years of eCO₂ treatment in growing season. Saleh et al. (2020) found that the eCO₂ increases AMF colonization rate of Oreganum vulgare, promoting growth of hyphae and plant nutrient absorption. Panneerselvam et al. (2019) revealed that the eCO₂ increases AMF colonization of rice, improving grain yield by 25.08%. Meta-analyses also showed that the eCO₂ increases AMF hypha length by 23% and colonization rate by 17% (Alberton et al. 2005), enhancing biomass and nutrient contents for AM plants (Dong et al. 2018).

Invasive relative to native plants may respond to the eCO₂ more quickly and strongly in terms of their symbiotic relationships with AMF. Firstly, the eCO₂ may promote growth more greatly in invasive relative to native plants (Lei et al. 2012; Liu et al. 2017), which allows the invasive plants to allocate more carbohydrates to roots (de Graaff et al. 2011; Wei et al. 2019), promoting root growth, secretory function (Kao-Kniffin and Balser 2007) and therefore symbioses with AMF. In addition, the greater increase in growth caused by the eCO₂ in invasive relative to native plants may more greatly increase their acquisition of soil nutrients and water, decreasing soil resource availabilities more greatly and therefore more greatly enhancing symbioses between the invasive plants and AMF (Lei et al. 2012). Secondly, invasive relative to native plants often exhibit acquisitive growth strategy in prolific environments (Leishman et al. 2010), which may allow them to respond more quickly to the eCO₂ in terms of AMF colonization.

In order to know whether the facilitation of the eCO₂ on exotic plant invasions is associated with its effects on AMF colonization, we grew the invasive plant Xanthium strumarium (synonym X. italicum) and the phylogenetically related native plants X. sibiricum and Bidens biternata under two CO₂ concentrations for three consecutive years in open-top chambers (OTCs). All these species have AMF mycorrhizas based on our previous observation. We hypothesize that (i) elevated CO₂ concentration may increase AMF colonization rates for all three species; and (ii) the invasive relative to the native plants may have a higher AMF colonization rate, especially under the eCO₂ (respond more strongly) and in the first year of the study (respond more quickly).

MATERIALS AND METHODS

Study site and species

This study was conducted in four OTCs (QiuShi Environment, Hangzhou, China; at least 14 m apart from one another, Supplementary Fig. S1), which were located at the experimental base of Shenyang Agricultural University (41°50′9″ N, 123°34′33″ E; 53 m a.s.l.), in Shenyang, Liaoning Province, Northeast China. This site is in the warm-temperate continental monsoon climate. The mean annual temperature is 8.5 °C, mean annual precipitation is 698.5 mm (mainly in June to September), mean annual sunshine is 2628.3 h and annual frost-free season is 171 days (Zhou et al. 2017). The use area is 9 m² per OTC, with 2.5 m in height. The illuminance in the center of each OTC is 75% of that outside at a sunny noon. The temperature in each OTC was automatically controlled using computers, air conditioners and spray systems, and the temperature and humidity in each OTC were monitored using sensors (HF332-WB1XXXXX, Rotronic, Switzerland). The CO₂ concentration in each OTC was automatically monitored and controlled using high purity CO₂ and gas detection transmitter (BMG-CO₂-NDIR, Bluemoom, China).

Xanthium strumarium, an annual herb in Compositae, is a native to North America. It was first found in Beijing, China in 1991, and now has widely spread in Liaoning, Hebei, Beijing, Inner Mongolia and Xinjiang (Feng 2020). The invader has a strong competitive ability, suppressing growth of neighboring native plants and reducing biodiversity seriously in invaded habitats (Iqbal et al. 2020). The native congeneric annual X. sibiricum and confamilial annual B. biternata are widespread in China, and often co-occur with X. strumarium in the field.

Experimental design

This study was conducted annually for three consecutive years (2015–2017). The seeds were collected from more than 20 individuals (>30 m apart from one another) for each species in Shenyang in the autumn before each planting year. In the middle of April, the seeds were stratified in wet sand at 4 °C for 2 weeks, sterilized with 0.5% KMnO₄ for 30 min, washed repeatedly with distilled water, and then sowed in sterilized substrates. In the middle of May, when the seedlings were ≈10 cm in height, similar-sized seedlings were transplanted into pots (15 cm × 30 cm × 18 cm), one per pot.

The potted seedlings were grown in an open area for 1 week, and then moved into the OTCs. Before being used in each year, the OTCs were fumigated with mixture of purified KMnO₄ and 37% methanal for 12 h, and then ventilated using fans for 1 week. Two randomly selected OTCs were used as ambient CO₂ concentration treatment (aCO₂; control), and the other two as elevated CO₂ concentration treatment (eCO₂; treatment). The CO₂ concentrations in the control and treatment OTCs fluctuated around 400 and 800 µmol mol⁻¹, respectively (Supplementary Fig. S2). The seedlings were watered and weeded as needed, and their positions in each OTC were randomly changed weekly.

In the first year of the study, the growth substrate was a mixture of 70% forest topsoil and 30% river sand (the studied plant species were not found in the soil and the sand in recent years), in which the contents of organic matter, total carbon, total nitrogen, total phosphorus, available nitrogen and available phosphorus were 5.89, 4.1, 0.6, 0.42 mg g⁻¹, and 32.15, 11.22 mg kg⁻¹, respectively. In the second and third year, the seedlings were grown in the pots, in which the same species had been grown under the same CO₂ concentration in the previous year, i.e. in soil trained by the same species under the same CO₂ concentration. The naturally dead roots of the previous year remained intact in the pots, and leaf and stem litters were collected, and put on the surface of their source pots. For each of the three species under each of the two CO₂ treatments, 24, 16 and 8 seedlings were grown in the first, second and third year, respectively. The number of the seedlings was reduced for each species and treatment in the second and third year, as some of seedlings had been sampled in the previous year.

Root and soil sampling

In order to measure AMF colonization rates, soil characteristics and soil AMF communities, four randomly selected pots were harvested for each species and CO₂ treatment in August (plants at the vigorous growth stage) of each year (four replicates; two pots from each of the two OTCs with the same CO₂ treatment). In order to collect roots easily, soil (with 30% river sand) moisture was controlled before harvest. Aboveground parts of the plant and the topsoil (0.5 cm) were removed for each pot when sampling, some intact fine roots (<2 mm) were collected carefully, immersed in FAA fixed liquid (70% ethanol 90 mL, 100% glacial acetic acid 5 mL, 37% methanal 5 mL) and stored at 4 °C for measuring AMF colonization rates. The soil in the center of each pot was collected, sieved through a 10 mesh sieve and mixed well. Approximate 10 g soil was collected, frozen with liquid nitrogen and stored at −80 °C for measuring AMF community structure; ≈100 g soil was collected, and dried at room temperature for measuring soil nutrients.

Soil characteristic measurements

Total soil carbon and nitrogen contents were measured using elemental analyzer (EA 3000, Euro Vector, Italy), soil available nitrogen content using the alkaline hydrolysis diffusion method (acidic titration), soil organic matter content using the potassium dichromate titration method (acidic-digestion) and soil available and total phosphorus contents were measured using the molybdenum-antimony anti-colorimetric method (alkaline-digestion) (UV-9000S, METASH, Shanghai, China). For detailed methods please see Bao (2000), and all these measurements were conducted by Analysis and Testing Center, Shenyang Agricultural University.

AMF root colonization measurements

AMF colonization rates were measured using an improved ink–acetic acid staining method (Vierheilig et al. 2005). The fine roots were taken out from the FAA fixed liquid, washed with distilled water, and then cut into 1 cm pieces. For each species and CO₂ treatment, 50 pieces of fine roots were put into 10% (w/v) KOH solution at 60 °C (40 min for X. strumarium, 20 min for B. biternata and 50 min for X. sibiricum; based on our preliminary experiment), rinsed with distilled water, acidified with 5% (v/v) acetic acid for 20 min, stained with 5% ink–acetic acid (5 mL Parker black ink and 95 mL 5% glacial acetic acid) for 40 min, rinsed several times with distilled water, and finally destained in distilled water for 2 h. AMF colonization was observed using a microscope (Eclipse Ni-U, Nikon, Japan), and the colonization rates were assessed using the method of Biermann and Linderman (1981) as follows:

where n1, n2, n3, n4, n5 and n6 are the number of the root segments with 0, 20%, 40%, 60%, 80% and 100% of colonization rates, respectively.

AMF community measurements

AMF community were measured using high-throughput sequencing method (Xu et al. 2018). Total DNA was extracted for each soil sample using Omega soil DNA kit (D5625-01, Omega Bio-Tek, USA), and the quality and quantity were determined using 1% agarose gel electrophoresis and NanoDrop ND-1000 spectrophotometer (ND-1000, Thermo Fisher Scientific, USA), respectively. The fungal ITS1 region was amplified using the forward primer ITS5F (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and the reverse primer ITS1R (5′-GCTGCGTTCTTCATCGATGC-3′). Sample-specific 7-bp barcodes were incorporated into the primers for multiplex sequencing. PCR was performed in 25 µL reaction mixture containing 5 µL of Q5 reaction buffer, 5 µL of Q5 high-fidelity GC buffer, 0.25 µL of Q5 high-fidelity DNA polymerase, 2 µL of dNTPs, 1 µL of each primer, 2 µL of DNA template and 8.75 µL of ddH₂O with the following cycle conditions: 95 °C for 30 s, 25 cycles of 98 °C for 15 s, 55 °C for 30 s, 72 °C for 30 s; followed by extension at 72 °C for 5 min. The PCR product of each sample was detected using 2% agarose gel electrophoresis, quantified using the PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, USA) and microplate reader (FLx800, BioTek, USA), and then pooled together with equal proportion. DNA library was constructed using TruSeq Nano DNA LT Library Prep Kit (FC-121-4001, Illumina, USA), and pair-end 2 × 300 bp sequencing was performed using MiSeq Reagent Kit v3 (MS-102-3003, Illumina, USA) at the Illumina MiSeq platform. The aforementioned works were performed by Shanghai Personal Biotechnology Co., Ltd (Shanghai, China).

Low-quality sequences were removed (Magoč and Salzberg 2011) and 2 035 277 high-quality sequences (200–300 bp per sequence) were obtained for all 48 samples. An averaged, rounded rarefied OTU table was generated by averaging 100 evenly resampled OTU subsets under the 90% of the minimum sequencing depth in order to minimize the confounding effects of the differences in sequencing depth among samples. The high-quality sequences were clustered at the level of 97% similarity, and annotated using the UNITE database (Kõljalg et al. 2013). All AMF were selected, and classified into three guilds: edaphophilic, rhizophilic and ancestral (Weber et al. 2019).

Soil AMF composition was not measured for the samples from the third year because of the lack of funds.

Statistical analyses

To test the effects of species, CO₂ concentrations, planting years and their interactions on AMF colonization rates and soil traits, linear mixed effects models were fitted using the lmer function in R package lmerTest (Kuznetsova et al. 2020). Species (invasive and two natives), CO₂ concentrations (aCO₂ and eCO₂), planting years and their interactions were treated as fixed factors. Linear mixed effects models were also used to test the effects of species (invasive vs. native) in the same planting year and the same CO₂ concentration, planting years for the same species under the same CO₂ concentration, and CO₂ concentrations for the same species in the same planting year on AMF colonization rates and soil characteristics. The OTC number was treated as a random factor (two of the four replicates from one of the two OTCs with the same CO₂ treatment) for all the linear mixed effects models.

The difference in AMF community structure among different planting year for the same species under the same CO₂ concentration were analyzed using principal co-ordinates analysis (PCoA) and PERMANOVA with adonis function (both based on Bray–Curtis distance) in R VEGAN package (v 2.5.7) (Oksanen et al. 2020). Multivariate relationships between the relative abundance of each AMF genus under different treatments were analyzed using principal component analysis (PCA). The differences in scores of the relative abundance of each AMF genus on PC1 and PC2 between the invasive and each native species in the same CO₂ concentration and the same year, and between different CO₂ concentrations for the same species in the same year were analyzed using independent sample t-test. The relationships between AMF colonization rates and soil characteristics were analyzed using Pearson’s correlations, and the relationships between different guilds of AMF, i.e. edaphophilic, rhizophilic and ancestral, and soil characteristics were analyzed using Mantel test with mantel function in R DPLYR package (Wickham et al. 2020). The proportional explanation for variations of AMF colonization rates by soil characteristics and AMF communities were analyzed using variation partitioning analysis with rda function in the R VEGAN package. All these analyses were performed in R 3.5.2 (R Core Team 2018).

Accession numbers

The double-ended sequences were spliced into single-end sequences, and deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) under accession number PRJNA656668.

RESULTS

AMF colonization rates

Effects of species, CO₂ concentrations, planting year and the interactions between species and planting year and among species, CO₂ concentrations and planting year on AMF colonization rates were significant (Table 1). Effects of the OTC number were not significant. When the data from each CO₂ treatment were analyzed separately, the effects of species, planting year and their interaction (marginally under eCO₂) on AMF colonization rates were also significant (Fig. 1). However, the magnitude of the effects of planting year was lower for the invasive relative to the two native species.

AMF colonization rate was marginally higher for the invasive plant X. strumarium under the eCO₂ than under the aCO₂ in the third year of the study (χ² =5.84, P = 0.052), but similar in the first and second year (Fig. 1). The eCO₂ significantly increased AMF colonization rates for X. sibiricum (χ² = 30.63, P < 0.001) in the first year and B. biternata (χ² = 33.85, P < 0.001) in the third year (Fig. 1).

Under the aCO₂, the invasive plant X. strumarium had a significantly higher AMF colonization rate than the native B. biternata and X. sibiricum in the first planting year (Fig. 1a). In the second year, the invader also tended to have a higher AMF colonization rate, although the difference was not significant. In the third year, the AMF colonization rate of the invader was similar with that of the native B. biternata, but significantly lower than that of the native X. sibiricum (Fig. 1a). Under the eCO₂, the invader also tended to have higher AMF colonization rates than the two native species in the first and second year, while the differences were not significant except that between the invader and the native B. biternata in the second year (Fig. 1b). In the third year, the difference between the invader and each of the native species was not significant.

Under the aCO₂, AMF colonization rates in the third relative to the first year were higher by 85% and 104% for the native plants B. biternata and X. sibiricum, respectively, while similar for the invader (Fig. 1a). Under the eCO₂, the magnitude of the increase in AMF colonization rates was also much lower for the invader (by 28%) than for B. biternata (by 123%) and X. sibiricum (by 39%) (Fig. 1b).

Composition and diversity of AMF communities

In total, five genera of AMF (in three families, two orders and one class) were identified through high-throughput sequencing (Supplementary Fig. S3). Diversispora was the dominant AMF genus for all three species in the first planting year, while Redeckera became into the dominant AMF genus and no Funneliformis was detected in the second year (Supplementary Fig. S3). In the first planting year, CO₂ concentrations did not significantly influence AMF communities for all three species (Fig. 2a; Supplementary Table S1). In the second year, CO₂ concentrations significantly influenced AMF communities of rhizosphere soils for B. biternata, which was mainly caused by the changes in genera Diversispora, Rhizophagus and Pacispora (Fig. 2b; Supplementary Table S1), but not for the invader and its native congener (Supplementary Table S1). PCA showed that there were no significant differences in AMF communities between the invasive and native plants in the first planting year under both CO₂ concentration treatments (Fig. 2a; Supplementary Table S1). In the second year, however, the differences in AMF communities between X. strumarium and its native congener were significant under both aCO₂ and eCO₂ treatments (Fig. 2b; Supplementary Table S1).

In general, planting year significantly influenced AMF communities of the invasive and native species (Fig. 3a). When pooling the data from both CO₂ concentration treatments together, planting year significantly influenced the AMF communities for X. strumarium (Fig. 3b) and the native B. biternata (Fig. 3c), but not for the native X. sibiricum (Fig. 3d). When analyzing the data from each CO₂ treatment separately, planting year also influenced the AMF communities for X. strumarium and B. biternata (marginally, 0.05 < P < 0.1), but not for X. sibiricum under the eCO₂ (Table 2). Under aCO₂, however, the effects of planting year on the AMF communities were significant for all three species.

Correlations among AMF colonization rates, soil characteristic and AMF communities

AMF of rhizophilic guild (Rhizophagus and Funneliformis genus) and edaphophilic guild (Redeckera and Diversispora genus) were strongly correlated with soil total carbon (P ≤ 0.01, Fig. 4). AMF of edaphophilic guild were also correlated with soil organic matter and available nitrogen, respectively (P ≤ 0.05). AMF of ancestral guild (Pacispora genus) were not associated with the soil characteristics (Fig. 4).

Variation partitioning analyses showed that in the first planting year, soil characteristics and AMF communities explained the variations of AMF colonization rates by 22.3% and 15.9%, respectively (Fig. 5a). In the second year, the magnitude explained by soil characteristics and AMF communities decreased to 17.4% and 3.8%, respectively (Fig. 5b).

DISCUSSION

Effects of CO₂ concentrations, species and planting year on AMF colonization

Inconsistent with our hypothesis, the eCO₂ significantly increased AMF colonization rates for the invasive and native species only in few cases. Many studies have demonstrated that plants could allocate more carbon to AMF under elevated relative to aCO₂ concentration (Drigo et al. 2013; Panneerselvam et al. 2019; Wei et al. 2019), which could facilitate growth of AMF hyphae, and in turn promote soil nutrient uptake and growth of the plants (Saleh et al. 2020; Sun et al. 2017). The effects of CO₂ concentrations on AMF root colonization were influenced by other factors such as plant species and planting year (Fig. 1; Table 1). In addition, the invasive plant X. strumarium did not benefit more from the eCO₂ in terms of AMF colonization than the native species. CO₂ enrichment did not influence AMF colonization for the invader in the first 2 years of the study; in the third planting year, however, the eCO₂ increased AMF colonization for the invader as well as for the native plant B. biternata. Our results indicate that the increased growth advantage of the invader over the two native species under the eCO₂ (data not shown) may not depend on AMF colonization. Invasive species can increase nutrient uptake by increasing biomass allocation to roots (Barros et al. 2020; Lei et al. 2012) besides increasing AMF colonization, or increase photosynthetic nutrients-use efficiencies to improve their competition under the eCO₂ (Feng et al. 2009; He et al. 2018).

Partially consistent with our hypothesis, the invasive plant interacted more quickly with AMF than the native plants, but this occurred only under the aCO₂. Under the aCO₂, the AMF colonization rate was significantly higher for the invasive relative to the native species in the first planting year, while the difference disappeared in the second year, or even lower for the invader than for the native X. sibiricum in the third year. Under the aCO₂, the AMF colonization rates increased with increasing time of the study for the two native species but not for the invader (Fig. 1a). The faster AMF colonization may promote nutrient absorption and utilization for the invader, increasing its carbon assimilation ability and then facilitating invasion success in new habitat under aCO₂ (Awaydul et al. 2019; Bunn et al. 2015; Harner et al. 2010). Through ¹⁵N and ³³P isotope labeling, Awaydul et al. (2019) found that intact common mycorrhizal networks preferentially transferred mineral nutrients to the invasive plant Solidago canadensis, and decreased N and P uptakes for the co-occurring native plant Kummerowia striata.

The effects of planting year on the difference in AMF colonization rates between the invasive and native species may be associated with the interannual differences in soil characteristics and climates. Our results showed that soil characteristics were more powerful in explaining the invasive–native difference in AMF colonization rates than soil AMF communities, which were also influenced by soil characteristics. The contents of soil available and total phosphorus and organic matter changed significantly among the 3 years of the study for each of the species, which all significantly influenced AMF colonization rates (Supplementary Fig. S4). For example, in the third relative to the first planting year, the contents of total and available soil phosphorus were lower, and the content of soil organic matter was higher for the native B. biternata under both aCO₂ and eCO₂ (Supplementary Table S2 and Fig. S5). Similarly, Harner et al. (2010) found that AMF colonization rates were positively correlated with soil N:P; and Chen et al. (2017) found that nitrogen addition decreased (while increasing precipitation increased) soil AMF abundance. The difference in AMF colonization rates between invasive and native species may also change with invasion processes of exotic plants, as which may alter soil nutrient contents (Supplementary Fig. S5), and therefore influence AMF colonization on both the invasive and native species (Castro-Díez et al. 2014). Our results indicate that the environmental and temporal factors should be taken into account when comparing the difference in AMF colonization rates between invasive and native species.

Effects of CO₂ concentrations, plant species and planting year on soil AMF communities

Species, CO₂ concentrations and planting year all significantly influenced soil AMF communities. The dominant AMF genus was Diversispora for all three species in the first year of the study, while Redeckera in the second year. Planting year significantly influenced soil AMF communities for all species under both aCO₂ and eCO₂ (marginally for B.biternata and except X. sibiricum under eCO₂, Table 2), which may be associated with its effects on soil nutrient availabilities. Many studies showed that soil nutrients influenced AMF communities significantly (Grove et al. 2017; Johnson et al. 2010). Our study showed that soil nutrients such as total and available phosphorus, total and available nitrogen and organic matter changed among the planting years for the invasive and native species under both CO₂ concentrations (Supplementary Table S2 and Fig. S5), and that soil nutrients indeed influenced AMF in rhizophilic and edaphophilic guilds (Fig. 4). However, the invasive and native species may also influence soil AMF communities by exuding allelochemicals into soils (Dieng et al. 2015). Tai et al. (2016) found that water extract of X. strumarium increased soil fungi amount, and the contents of available nitrogen and potassium. Pei et al. (2020) found that invasive populations of Triadica sebifera released more flavonoids into soils than its native populations, contributing to higher AMF colonization. Li et al. (2017) found that soil allelochemicals changed significantly with the changes of invasion intensity or resident time by exotic plants, affecting allelopathic effects.

The interactive effects of planting year and species or CO₂ concentrations on soil AMF communities were significant. In the first planting year, both species and CO₂ concentrations did not significantly influence AMF communities based on our PCA. In the second year, however, the differences in AMF communities between the invader and its native congener were significant under both the aCO₂ and eCO₂ treatments, and CO₂ concentrations significantly influenced AMF communities of B. biternata (Supplementary Table S1). These results further indicate that the effects of species and CO₂ concentrations on soil AMF communities were influenced by planting year (temporal effects). Effects of the eCO₂ on soil AMF communities may be associated with its influences on litter characteristics and/or root exudates of invasive plants (Lei et al. 2012; Zhang et al. 2019). The increase of CO₂ concentration often increases the contents of lignin, cellulose, tannin and phenolics, and decreases the contents of nitrogen and phosphorus in plant litters, reducing decomposition rates (de Graaff et al. 2011), and thus altering soil AMF communities (Johnson et al. 2010).

Both soil AMF communities and nutrient availabilities may influence AMF root colonization rates (Fig. 5; Barto et al. 2011; Thirkell et al. 2020; Wang et al. 2017). Wang et al. (2017) found that phosphorus addition reduced AMF colonization rate of maize. Our study showed that AMF colonization rates were correlated negatively with soil total and available phosphorus, and positively with soil organic matter (Supplementary Fig. S4). Barto et al. (2011) found that invasion of Alliaria petiolata influenced root AMF composition and decreased AMF colonization in the native plant Acer saccharum. However, the differences in AMF colonization rates and in soil AMF communities between X. strumarium and the two native species were not consistent in our study (Figs 1 and 2; Supplementary Table S1). In the first planting year, e.g. X. strumarium had significantly higher AMF colonization rate under the aCO₂ than the two native species, while soil AMF communities were similar for the invasive and native species. Our results could not deny the effects of soil AMF communities on AMF colonization rates of plants (Fig. 5). AMF colonization rate would be higher for X. strumarium than for the native species if soil AMF or dominant AMF were more prone to form symbioses with the invader, although the three species had similar soil AMF communities. Different AMF may have favorite host plants although their host specificity is relatively low (Martínez-garcía and Pugnaire 2011).

CONCLUSIONS

Atmospheric CO₂ enrichment increased AMF colonization rates for X. strumarium and its two co-occurring native species only in few cases, and the invader did not benefit more from CO₂ enrichment in terms of AMF colonization than the natives. Under aCO₂ concentration, however, the invasive relative to the native species had a higher AMF colonization rate in the first year of the study, and in the second and third year of the study this difference disappeared due to the increase in AMF colonization rates of the natives. Both soil nutrient and AMF community contributed to the effects of species, CO₂ concentrations and planting year on mycorrhizal fungi colonization. Our results indicate that the increased growth advantage of the invader over the natives under elevated CO₂ may not be associated with AMF colonization, and that under aCO₂ the invader could more quickly form symbiosis with soil AMF, contributing to adaptation and occupation of new habitats. It is necessary to consider the roles of AMF and the effects of time when determining the effects of global changes such as atmospheric CO₂ enrichment on exotic plant invasions.

Supplementary Material

Supplementary material is available at Journal of Plant Ecology online.

Table S1: Results of PCA showing effects of the species and CO₂ concentrations on AMF communities for the invasive and native species.

Table S2: Results of the liner mixed effects models showing effects of the species, CO₂ concentrations, planting year and their interactions on soil characteristics.

Figure S1: Positions of the four open-top chambers and the control room.

Figure S2: Diurnal variations in temperature, moisture and CO₂ concentrations in chambers with different CO₂ concentrations.

Figure S3: Relative abundance (genus level) of soil AMF for the invasive and native species.

Figure S4: Pearson’s correlations between AMF colonization rates and soil characteristics.

Figure S5: Results of linear mixed effects models showing effects of the species, planting years and their interactions on soil characteristics under aCO₂ and eCO₂.

Acknowledgements

We are grateful to Chun-Feng Hu and Dan Li for their assistance during plant cultivation and harvesting, and the editor and the reviewers for their helpful comments and suggestions on an early version of this paper.

Funding

This work was supported by the National Natural Science Foundation of China (31971557, 31670545 and 31470575) and the National Key R&D Program of China (2017YFC1200101).

Conflict of interest statement. The authors declare that they have no conflict of interest.

REFERENCES