Study area and study system
The study was conducted in an area of Xujiaba Reserve on Mt. Ailao (23°32′N, 101°01′ E) in central-southern Yunnan, China, at an altitude of 2045 m. The mean annual temperature and precipitation are 11.7 ℃, and 1923.1 mm, respectively and the natural vegetation of the area mainly consists of natural evergreen, broad-leaved forest (Feng et al. 2021).
Ageratina adenophora was mainly distributed in forest gaps at the nature reserve. We divided the gaps into two types: lightly disturbed gaps due to snow, wind damage, cutting, and forest edges, where topsoil was not broken; and heavily disturbed gaps caused by landslides, building firebreaks, and soil erosion, in which the topsoil was broken and removed. We measured soil chemical traits of the two newly formed disturbed gaps. Soil pH, water content, available nitrogen (AN, mg per kg dry soil), phosphorus (AP, mg per kg dry soil), and potassium (AK, mg per kg dry soil) were analyzed at Institutional Center for Shared Technologies and Facilities of Xishuangbanna Tropical Botanical Garden (XTBG), Chinese Academy of Sciences (CAS). Then, we selected 12 sites in each disturbance gap, and four sites were located for each of the three invasion periods: long (first record of A. adenophora at least 25 y ago), short (1–5 y ago), and non-invaded sites (see Table S1, Fig. S1).
Field data and soil samples collection
In August 2018, according the shape of forest gaps, we set two or three transects separated by a distance of 6 m, with three 1 × 1 m plots per transect to determine plant species composition, aboveground biomass and cover of A. adenophora. In the sites with A. adenophora populations present, we located the grid in the middle of the invader stands. The data collected in all plots at a particular site were pooled to calculate the richness and Shannon diversity index of native species.
We selected four 1 × 1 m plots to collect soil for each site. Ageratina sp. have a well-developed, fibrous root system. Thus, roots of A. adenophora were excavated 15 cm deep in each plot, and rhizosphere soil was collected from three to eight individuals by shaking the roots. For non-invaded sites, four 1 × 1 m plots were selected at each site and soil depths of 0–15 cm were collected at 6 kg per plot. Roots, residues, and small stones were removed from the soil sample using a 2-mm sieve. Four samples per site were mixed for fungal community analyses.
Pot experiment
The soil in each plot was divided into two parts. One part was steam-pasteurized at 121 ℃ for 1h twice at 24 h intervals and sterilized as a live inoculum, and then maintained at 25°C for 2 wk to allow for mineralization; the other sample was not sterilized as a live inoculum. Simultaneously, we sterilized sand as described above.
To evaluate the strength of PSFs of soil from different sites, we established two soil sterilized treatments and four replicates for each of the 24 sites, with a total of 192 pots (1 L). Each pot contained 500 mL of sterile or live soil from one plot, and 500 mL of sterile sand. Seeds of A. adenophora (collected from Kunming, Yunnan Province, China) were surface disinfected with 0.1% KMnO4 for 10 min, rinsed thrice in sterile water, and sown in a tray with sterile sand in a phytotron (25/20 ℃ [day/night] and light intensity of 1000 µmol m− 2 s− 1 for a 12 h photo history) in XTBG. Two weeks after germination, seedlings were transplanted into pots in the same phytotron. To avoid contamination, materials, tools, and surfaces used in the experiment were sterilized by soaking in 0.3% aqueous NaClO. All plants were harvested after 2 months, dried at 60 ℃ for 48 h, and weighed.
Responses of plant growth to their own soil biota relative to sterile soil (total PSFs) reflect net effects of all soil biota including generalist and specialized biota; growth responses to their own soil biota relative to soil biota of all plant species (specific PSFs) reveal effects of more specialized soil organisms (Cortois et al. 2016). The strength of the total PSF was calculated as follows:
(variables in live inocula - mean of variables in sterile inocula) / mean of variables in sterile inocula,
where > 0, < 0, and = 0 mean a positive, negative, and neutral feedback, respectively.
Soil fungal community analysis
DNA of soil samples from 24 sites was extracted using the E.Z.N.A Soil DNA Kit (Omega Bio-tek, Norcross, GA, US), according to the manufacturer’s protocols. DNA was sent to Shanghai Majorbio Biopharm Technology Co., Ltd. for sequencing on the Illumina MiSeq platform. The rDNA internal transcribed spacer (ITS) I region of the fungi in all soil samples was amplified using the primers ITS1 CTTGGTCATTTAGAGGAAGTAA and ITS2-Rev GCTGCGTTCTTCATCGATGC. PCR reactions were performed with 15 µL of Phusion® High-Fidelity PCR Master Mix (New England Biolabs), 2 µM of forward and reverse primers, and approximately 10 ng of template DNA. The PCR program was as follows:98 ℃ for 1 min, followed by 30 cycles of 98 ℃ for 10 s, annealing at 50 ℃ for 30 s, elongation at 72 ℃ for 30 s, and a final extension at 72 ℃ for 5 min. The PCR product was extracted from a 2% agarose gel and purified using the Qiagen Gel Extraction Kit (Qiagen, Germany) according to the manufacturer’s instructions.
The PCR products were sequenced using the paired-end method with Illumina MiSeq. The data were processed using PIPITS 2.3.0 pipeline (Gweon et al. 2015) according to the following criteria: both short (< 100 bp) and singleton sequences were removed, and the operational taxonomic units (OTUs) were defined at a 97% sequence similarity threshold using VSEARCH (Rognes et al. 2016); taxonomy was assigned using the RDP classifier trained on the UNITE fungal ITS reference database (Abarenkov et al. 2010). Functional information based on OTUs was obtained by aligning the FUNGuild database with the confidence ranking of probable (http://www.stbates.org/guilds/app.php) (Nguyen et al. 2016).
Data analyses
We used one-way ANOVA to test the effect of gap type on soil chemical traits. A general linear model was used to test the effects of gap type and invasive period on the aboveground biomass, covers of A. adenophora, richness, and Shannon index of native species in the field, and LSD post hoc tests were run. We used a linear mixed model to test biomass in live soil and the strength of total PSFs, with invasive period, gap type, soil treatment (only for biomass) and their interaction as fixed effects and sites as random effects. Furthermore, multiple comparisons between invasion periods and gaps were made.
We fitted generalized linear mixed models with invasion period to assess changes in the relative abundance (proportion of reads, Binomial response) and richness (number of OTU, Poisson response) of plant pathogens and AM fungi in pre-planned contrast invasion periods for the two gap types, respectively. Contrast 1: uninvaded sites were compared with those with a short invasion history; Contrast 2: uninvaded sites were compared with those with a long invasion history; Contrast 3: sites with a short invasion history were compared with those with long invasion history. These analyses were performed using IBM SPSS Statistics for Windows 26.0 (Armonk, NY, IBM Corp.). We calculated non-metric multidimensional scaling (NMDS) ordination with all OTUs using the VEGAN package in R (Oksanen et al. 2008).