The fungal community in non-rhizosphere soil of Panax ginseng are driven by different cultivation modes and increased cultivation periods

Soil sample collection and DNA extraction

Soil samples were collected in growth season (July 10 th 2016) from Dadong Village (42.31 N 127.19 E), Fusong, Jilin Province of China. Fusong is the central ginseng production region of China, and the Dadong Village is one of the several places where understory wild ginseng has been grown for over 40 years. Soil samples were collected from fields of farmland cultivated ginseng of 2 years old (C2) and 5 years old (C5), understory wild ginseng of 15 years old (W15), and 35 years old (W35), fallow fields which have been abandoned for 10 (F10) years after five years of ginseng farmland cultivation. Permission to access the field site was obtained from the landowner, Shihua Chen. We chose 2-year-old and 5-year-old farmland cultivated ginseng because after two years of growth, disease occurrence and death rates of farmland cultivated ginseng generally increase (Dong et al., 2018) and farmland cultivated ginseng is usually harvested after five years of farmland cultivation. Thirty-five-year-old understory wild ginseng was the oldest understory wild ginseng we can found. Coupled with the 15-year-old understory wild ginseng, we can investigate the long-term influence of understory wild ginseng cultivation on the non-rhizospheric fungal community.

The understory wild ginseng used in this study was direct planted to the natural forest and grown without any manual intervention. The soil used for farmland cultivation was removed from the same area of natural forest where understory wild ginseng was grown. Fallow fields used in this study have been abandoned for ten years without planting any crops to avoid the influence of crop cultivation practice on soil microbial community.

Non-rhizosphere soil components of ginseng (20 cm away from ginseng plants) were collected as study materials with three biological replications. For each sample, 40 soil cores were collected from about 2,500 m² circular plots and pooled together. Each soil core was collected using a five cm diam PVC tube, which was hammered into the soil and collect the soil from the surface to down to five cm depth (Tedersoo et al., 2014). Power Soil DNA Isolation kit (MoBio, Carlsbad, CA USA) was used to extracted DNA from about 10g soil according to the manufacturer instructions.

Illumina sequencing of ITS rRNA gene amplicons

Fungi intergenic transcribed sequence (ITS) universal primers F2045 (5′–GCATCGATGAAGAACGCAGC-3′) and R2390 (5′- TCCTCCGCTTATTGATATGC-3) were used. Library construction and sequencing were carried out by Realbio Genomics Institute (Shanghai, China). Illumina Hiseq 2500 platform was used to perform amplicon sequencing and generated about 100100,000 250 bp paired-end (PE) reads in each sample. Original data generated in this study have been deposited into the NCBI SRA database, and the accession number was PRJNA523683.

Data analyses

PANDAseq (Masella et al., 2012) was used to assemble the paired-end (PE) reads, quality assessment and filtering were carried out with QIIME1 pipeline with default parameters (Caporaso et al., 2010). Chimeric sequences were identified and filtered by the UCHIME program (Edgar et al., 2011). USEARCH implemented in QIIME1 software (Sachs, Skophammer & Regus, 2011) was utilized to assign Operational taxonomic units (OTU) at the 97% sequence identity level. The most abundant sequence of each OUT was chosen as the representative sequence and then taxonomically classified using the RDP classifier against the Ribosomal Database Project (RDP) database (Wang et al., 2007; Brown et al., 2013). OTUs which belong to mitochondrion or chloroplast and singleton OTUs were removed. Alpha and beta diversities were calculated using the QIIME1 pipeline (Caporaso et al., 2010) after rarefied to 132,764 sequences for each sample. We used linear discriminant analysis effect size (LEfSe) (Segata et al., 2011) to identify the differentiation of fungal communities between different samples at multiple taxonomical levels, Kruskal-Wallis test P < 0.05 and LDA score of >2 were used as thresholds. Principal coordinate analyses (PCoA) based on unweighted UniFrac distance metrics were carried out using the QIIME1 pipeline (Caporaso et al., 2010). ADONIS was carried out using the R package vegan (Dixon, 2003).

Differences of alpha diversity indexes between samples were determined by Tukey’s method (multiple comparisons of means), and normality and homogeneity of variance analysis were respectively performed by using the Shapiro–Wilk normality test and Brarlett test. These analyses were carried out using R software (Version 3.5.2, https://www.r-project.org/). OTUs were parsed against the FUNGuild database to assign putative trophic strategies (Nguyen et al., 2016).

Regression analysis

We used a linear model (LM) regression analysis to test the relationship between OTU richness and cultivation periods. Standardized major axis (SMA) analysis was carried out using (S)MATR package in R (Falster, Warton & Wright, 2006; Warton et al., 2011) to fit a line of best fit and to obtain a slope estimate and R2 value.

Variation in fungi community compositions

Soil samples included four cultivation periods: 2-year-old (C2) and 5-year-old (C5) cultivated ginseng, 15-year-old (W15), and 35-year-old (W35) understory wild ginseng. Intergenic transcribed sequence (ITS) gene amplicon sequencing generated 135,340-189,257 reads and was rarefied to 132,764 clean reads for each sample. The Good’s coverage scores ranged from 0.996792 to 0.999862, indicating that the sequencing depth was sufficient to quantify the fungi communities (Table 1).

According to Principal Co-ordinates Analysis (PCoAs) analyses, samples from farmland cultivated ginseng and understory wild ginseng showed clear separation (Fig. 1A). PCoAs1 and PCoAs2 explained 23.13% and 16.68% of variation, respectively (Fig. 1A). ADONIS analysis showed that fungi communities in non-rhizosphere soil of farmland cultivated ginseng and understory wild ginseng were significantly different (R² = 0.22377, P = 0.003). Only 27.4% of OTUs from farmland cultivated ginseng and understory wild ginseng were overlapped (1,494 out of 5,446) (Fig. 1B). There was 1,310 common OTUs (28.1% of a total of 4,663 OTUs in C2 and W15) between C2 and W15, the number of common OTUs between C5 and W35 decreased to 285 (11.8% of a total of 2414 OTUs in C5 and W35, Fig. 1B). These results indicated that with the extension of cultivation periods, the differences of fungal community between the two cultivation modes were enlarged. In both farmland cultivated ginseng and understory wild ginseng, Ascomycota and Basidiomycota were dominant phyla, which accounted for 75.1%-91.1% of the total OTUs (File S1). LEfSe analysis showed that Agaricomycetes, Pezizomycetes and Glomeromycota were significantly enriched in understory wild ginseng compared to that of cultivated ginseng. In contrast, Sordariomycetes, Dothideomycetes, Ascomycota_unidentified, Eurotiomycetes, Tremellomycetes, Lecanoromycetes, Cystobasidiomycetes, and Wallemiomycetes were significantly enriched in farmland cultivated ginseng compared to that of understory wild ginseng (Figs. 1C, 1D, LEfSe, P <  0.05, LDA>2).

Extension of cultivation periods lead to decreased OTU richness and PD whole tree indices in both cultivation modes

Extension of cultivation periods significantly decreased the OTU richness and PD whole tree indices in both understory wild ginseng and farmland cultivated ginseng (Tukey’s method, P < 0.01; Table 1). After another three years of farmland cultivation, the OTU number decreased from 3261 OTUs in C2 to 1724 in C5, which was 52.9% of that in C2 (5914 OTUs) (Fig. 1B, File S1). Similar results were also found in understory wild ginseng, coupled with the extension of cultivation periods, the OTU number decreased from 2712 in W15 to 975 in W35 (Fig. 1B, File S1). The OTU number and cultivation periods were negatively correlated (SMA, p-value = 0.0020962, R2 = 0.62863, slope = -48.42971, 95% CI’s = −73.50041 to −31.91053, Fig. 2). However, the species evenness was not affected by the extension of cultivation periods in both cultivation modes, as indicated by Simpson indices (Tukey’s method, P >0.05, Table 1).

With the extension of cultivation periods, non-rhizosphere soil of more extended cultivation periods lost relatively the same amount of OTUs in the two cultivation modes (2000 OTU loss in W35 and 2241 loss in C5), while non-rhizosphere soil of farmland cultivated ginseng acquired more OTUs than that in understory wild ginseng (704 OUT gain in C5 compared with 263 in W35, Fig. 1B). Only a small proportion of the lost or gained OTUs in more extended cultivation periods of the two cultivation modes were overlapped (Fig. 1B). Specifically, there was 531 common OUT loss (26.6% in W35 and 23.7% in C5) in samples of more extended cultivation periods, and the number of common OUT gain in samples of more extended cultivation period was 19 (7.2% in W35 and 2.7% in C5, Fig. 1B).

Under farmland cultivation mode, with the extension of cultivation periods the abundance of Leotiomycetes, Ascomycota_unidentified, Microbotryomycetes, and Agaricostilbomycetes were significantly increased, while the abundance of Agaricomycetes, Incertae_sedis_10, Pezizomycetes, Chytridiomycetes, Orbiliomycetes, Incertae_sedis_14, Glomeromycetes, and Saccharomycetes were significantly decreased (Fig. 3, LEfSe, P <  0.05, LDA>2). Under understory wild ginseng cultivation mode, with the extension of cultivation periods, the abundance of Incertae_sedis_14 was significantly increased, and the abundance of Chytridiomycetes_unidentified was significantly decreased in non-rhizosphere soil compartment (Fig. 4, LefSe, P <  0.05, LDA>2).

Influence of fallowing on non-rhizospheric fungal community

Ginseng plants were harvested after five years of farmland cultivation, and the resulted fields were abandoned for ten years without growing any crops. According to PCoAs analyses, samples from fallow soil and farmland cultivated ginseng showed clear separation (Fig. 1A). ADONIS analysis showed that non-rhizospheric fungi communities in the fallow field and farmland cultivated ginseng were significantly different (R² = 0.35168, P = 0.009). After fallowing for ten years, the OTU richness, PD whole tree indices, and species evenness in non-rhizosphere soils were significantly increased (Tukey’s method, P <  0.05; Table 1). The OTU richness in F10 recovered to 2969 OTUs, but still less than that of C2 (Fig. 5A, File S1). Fallowing recovered 31.3% (702 out of 2,241) of the lost OTUs in C5 during the extension of cultivation periods, and for the gained OTUs, 38.2% (274 out of 704) of which were conserved in F10 (Fig. 5A). The other 1293 OTUs were gained OTUs in F10 (Fig. 5A). These results demonstrated that an increase of OTU richness in fallowed soil was achieved mainly by the gain of new OTUs other than by recovering the lost OTUs during the extension of cultivation periods. Compared with C5, the abundance of Pezizomycetes, Chytridiomycetes, Pucciniomycetes, Chytridiomycetes_unidentified, Orbiliomycetes, Incertae_sedis_14, and Glomeromycetes were significantly increased, and the abundance of Eurotiomycetes, Ascomycota_unidentified, Lecanoromycetes, Cystobasidiomycetes, and Agaricostilbomycetes were significantly decreased in F10 (Fig. 5B, LefSe, P <0.05, LDA>2).

Trophic mode prediction

FUNGuild software was used to assign a putative life strategy. Compared with understory wild ginseng, the non-rhizosphere soil of farmland cultivated ginseng had a lower abundance of symbiotrophs (Fig. 6). What is more, with the extension of the cultivation period, the abundance of symbiotrophs became even lower (Fig. 6). After fallowing for ten years, the abundance of symbiotrophs in F10 recovered and became slightly higher than C2 (31.02% in F10 vs. 29.37% in C2). In contrast, the non-rhizosphere soil of farmland cultivated ginseng had a higher abundance of saprotrophs than that of understory wild ginseng (Fig. 6). Moreover, with the extension of the cultivation period, the abundance of saprotrophs became even higher (Fig. 6). Similar to the result of symbiotrophs, the abundance of saprotrophs in F10 became even lower than C2 (42.27% in F10 vs. 56.49% in C2). Extension of cultivation period leads to the increased abundance of pathotrophs in both understory wild ginseng and farmland cultivated ginseng. F10 has the highest abundance of pathotrophs.