Soil physicochemical properties
The test results of physicochemical factors of the soil are shown in Table 2. In the soil with 15-year vines, the average contents of TK and SK were highest and the contents of SOM and TN were lowest. In the soil with 5-year vines, the contents of XN and SK were relatively higher, and the soil pH was between 7.86 and 7.98, thus it is alkaline soil.
The analysis of variance showed that grape planting year had significant effect on TK and SP (P < 0.05), and soil depth had significant effect on SOM and SK (P < 0.05). Among the physicochemical indices, organic matter had an extremely significant effect on TN and TK (P < 0.01), while TN had a significant effect on TK and XN (P < 0.05). AN was closely related to XN (P < 0.05), and XN had a significant effect on pH value (P < 0.05).
Basic analysis of sequencing data
From the 45 soil samples, we obtained 3,625,859 original sequences, in which 3,068,591 were valid ones; so, from each sample, 68,199 valid sequences were obtained on average, with the minimum number 59999 and the maximum number 76469. Based on the abundance of OTUs at 97% level, we used QIIME to process the OTUs for the valid sequences, and obtained a total of 31,533 OTUs, among which the highest number of OTUs from single sample were 871. Further, the highest number of OTUs was 11,144, occurring in the middle layer (25–35 cm) and the lowest number of OTUs was 9745, occurring in the lower layer (35–50 cm); and the 12-year vine group (WH) had the highest number of OTUs, which was 7251, while the 5-year vine group (ZN) had the lowest number of OTUs, which was 5382. See Table 3 for the sequence number and OTU number of each soil sample, and see Table 4 for the sequence number and OTU number of soil samples in the terms of soil depth and grape planting year.
QIIME (Version 1.9.1) was used to generate rarefaction curves for the 45 samples (Fig. 1). Rarefaction curves can reveal the representativeness of samples, and they can be used to evaluate whether the current sequencing depth is enough to reflect the microbial diversity contained in the community samples. It can be seen from Fig. 1 that when the sequencing amount exceeded 60,000 reads, there were still new OTUs appearing, but the curve had become flat, indicating that the sampling was basically reasonable, which meant the confidence of the bacterial community structure in the real environment was relatively high, in other words, the current sequencing depth was enough to reflect the diversity of fungal communities contained in the community sample.
Rarefaction curves at the OTUs similarity of 97%. In Fig. 1, the OTUs similarity of the samples is 97%. The abbreviations ZN, XZ, HY, WH, and GM represent the planting years of 5, 8, 10, 12 and 15, respectively. The first digit of sample name indicated the replicate number and the second indicated the soil depth (1, 2 and 3 denoted the soil depth of 5–20, 20–35 and 35–50 cm, respectively).
Analysis of composition and abundance of soil fungi
Species composition analysis reflects the community structure of samples at different taxonomic levels. At the phylum level, as shown in Fig. 2, the main dominant fungi in soil were Ascomycota with an average relative abundance of 34.49%, Basidiomycota with an average relative abundance of 13.85%, Mortierellomycota with an average relative abundance of 2.89%, Glomeromycota with an average relative abundance of 1.39%, and Chytridiomycota with an average relative abundance of 0.75%. The other dominant fungi included Kickxellomycota, Olpidiomycota, Mucoromycota, and Rozellomycota. The undetermined taxa of fungi in soil samples accounted for 45.67%. Analysis of variance (ANOVA) showed that the grape planting year had a significant effect on the abundance of Glomeromycota and Mucoromycota (P < 0.05) and the soil depth had a significant effect on the abundance of Ascomycota (P < 0.05).
Relative abundance of top 10 fungal phyla.
At the class level, as shown in Fig. 3, the main dominant taxa were Sordariomycetes, Dothideomycetes, Tremellomycetes, Agaricomycetes, and Tremellomycetes, and their average relative abundance were 13.23%, 10.60%, 6.07%, 5.22%, and 2.43%, respectively. The undetermined taxa at class level accounted for 54.69%. ANOVA analysis showed that grape planting year had a significant effect on the abundance of Tremellomycetes (P < 0.05), and soil depth had a significant effect on the abundance of Agaricomycetes (P < 0.05).
Relative abundance of top 10 fungal classes. Figure 2 shows the relative abundance of top 10 fungal phyla in grape rhizosphere soil, and Fig. 3 shows the relative abundance of top ten fungal classes in grape rhizosphere soil. The abbreviations ZN, XZ, HY, WH, GM represent the grape planting years of 5, 8, 10, 12 and 15, respectively; and the numbers 1, 2 and 3 represent the rhizosphere soil depths of 5–20, 20–35 and 35–50 cm, respectively.
The effect of soil physicochemical properties on fungal community
Redundancy analysis (RDA) at phylum level was carried out to reveal the relationship between soil physicochemical factors and fungal communities (Fig. 4). The first two axes of RDA analysis explained the 70.44% variations of fungal communities, in which the first axis explained 50.7% and the second axis explained 19.74%. SOM had the most significant effect on fungal communities, followed by TK, and the effects of soil physicochemical properties on fungal communities were in the order of SOM > TK > TN > TP > SK > pH > XN > SP > AN.
Redundancy analysis of soil physicochemical indices and fungal phyla. The abbreviations ZN, XZ, HY, WH, GM represent the grape planting years of 5, 8, 10, 12 and 15, respectively; and the numbers 1, 2 and 3 represent the rhizosphere soil depths of 5–20, 20–35 and 35–50 cm, respectively.
In this study, Spearman analysis was used to evaluate the correlation between the abundance of top 13 dominant fungal taxa at the phylum level and the soil physicochemical properties, as shown in Fig. 5. The results showed that the abundance of Basidiomycota was negatively significantly correlated with TP (P < 0.05) and extremely negatively correlated with SOM (P < 0.01); the abundance of Glomeromycota was negatively significantly correlated with AN (P < 0.05) and extremely positively correlated with XN (P < 0.01); the abundance of Kickxellomycota was positively significantly correlated to TK, SP and pH (P < 0.05); the abundance of Zoopagomycota was positively significantly correlated with SP (P < 0.05) and extremely positively correlated to AN (P < 0.01); the abundance of Chytridiomycota was positively significantly correlated to XN (P < 0.05); the abundance of Mortierellomycota was positively significantly correlated to pH (P < 0.05); and the abundance of Blastocladiomycota was negatively significantly correlated with TN (P < 0.05). For other dominant fungal phyla, the correlation between their abundance and soil physicochemical factors were not significant.
Spearman analysis between soil physicochemical factors and fungal phyla. SOM soil organic matter content, TN total nitrogen content in soil, TP total phosphorus content in soil, TK total potassium content in soil, XN nitrate nitrogen content in soil, AN ammonium nitrogen content in soil, SP available phosphorus content in soil, SK available potassium content in soil, PH soil pH value.
Analysis of soil fungi diversity
Alpha diversity represents the diversity of species in a specific ecosystem, and it is mainly related to two factors, one is the number of species, i.e., richness, and the other is diversity, i.e., the uniformity of individual distribution in the community. The community richness indices mainly include Chao index and Ace index. The larger Chao and Ace, the more OTUs in the community, and the higher the community richness. Community diversity indices mainly include Shannon index and Simpson index. The greater Shannon index, the higher community diversity; on the contrary, the higher Simpson index, the lower community diversity16,17.
The community richness indices Chao1 and ACE and diversity indices Shannon and Simpson of each sample are shown in Tables 5 and 6.
According to the analysis on Chao1 and ACE indices in the terms of grape planting years, the sample groups were in the order of WH (12-year vines) > XZ (8-year vines) > GM (15-year vines) > HY (10-year vines) > Zn (5-year vines), indicating the richness of fungi in the soil with 12-year vines was the highest and that of the 5-year vines was the lowest. The Shannon indices of the soil with vines of different ages showed no significant difference; comparatively, the Shannon index of WH (12-year vines) was slightly higher than that of other groups, and the Shannon index of Zn (5-year vines) was the lowest, indicating that the fungal diversity of soil with 12-year vines was slightly higher than that of other groups, and the fungal diversity of the soil with 5-year vines was the lowest, In general, the fungal diversity and richness of soil with 12-year vines were higher than those of other groups, and the fungal diversity and richness of soil with 5-year vines were lower than those of other groups.
According to the analysis on Chao1 and ACE indices in the terms of soil root depth, the sample groups were in the order of 35 cm deep group > 20 cm deep group > 50 cm deep group, indicating that the fungal richness was highest in the middle layer of the soil, followed by the top layer, and lowest in the lower layer. Shannon indices of different groups showed no significant difference. Comparatively, the Shannon index of the middle layer of the soil was slightly higher than that of other layers, and the Shannon index of the lower layer was the lowest. Fungal diversity also showed the trend of highest in the middle layer, followed by the upper layer and lowest in the lower layer.
Through correlation analysis of soil physicochemical indicators on the Alpha diversity indices of fungal communities, the results showed that SP had a positively significant correlation with Shannon index, and SP had an extremely positive correlation with fungal community diversity, while other physicochemical indicators had no significant correlation with the Alpha diversity indices.
Analysis on the correlation of Alpha diversity index of fungal communities and soil physicochemical indices (Table 6) showed that available phosphorus (SP) had an extremely significant positive correlation with both Shannon index and Alpha diversity index, while other physicochemical indices had no significant correlation with Alpha diversity index.
Nonmetric Multidimensional Scaling (NMDS) is often used to compare the differences between sample groups. In this study, NMDS analysis was applied to reflect the species information contained in samples into multi-dimensional space in the form of points. The degree of differences between different samples is reflected by the distance between points, which can reflect both the differences between and within sample groups. From Fig. 6, it can be seen that samples ZN1 (the upper layer of soil with 5-year vines), ZN3 (the lower layer of soil with 5-year vines) and HY3 (the lower layer of soil with 10-year vines) are far away from other samples, indicating that the fungal community structure of those samples was quite different from that of other samples. The close distance between other samples indicates that their similarity of fungal community structure was higher.
NMDS analysis of soil fungi. Note: Each point in the figure represents a sample, the distance between points indicates the degree of difference, and the samples in the same group are represented by the same color. When Stress is less than 0.2, NMDS can accurately reflect the difference between samples. The abbreviations ZN, XZ, HY, WH, GM represent the soil sample groups with the grape planting years of 5, 8, 10, 12 and 15, respectively, and the numbers 1, 2 and 3 represent the soil depths of 5–20 cm, 20–35 cm and 35–50 cm, respectively.
LEfSe (LDA Effect Size) analysis
LEfSe was used to identify fungal species with significant differences in the terms of grape planting years and soil depths. Here, only the cladogram is showed. Firstly, the LEfSe analysis was carried out for different samples from the same soil depth, and the LDA score of 4 was used to identify statistically significant difference between fungal taxa. In the middle layer (20–35 cm), there were 14 taxa with LDA scores higher than 4, including 5 taxa of GM2 (15-year vine), 7 taxa of HY2 (10-year vine), and 2 taxa of ZN2 (5-year vine) (Fig. 7a). In the upper layer (5–20 cm) and the lower layer (35–50 layer), no taxa with significant difference were observed.
(a) Cladogram of LEfSe analysis. (b) Cladogram of LEfSe analysis.
Then, the LEfSe analysis was carried out for samples from different soil depths, the results found there were 7 fungal taxa whose LDA scores were higher than 4, indicating they had statistically significant difference. The 7 fungal taxa included 1 taxa in the lower layer (35–50 cm) of GM3 (15-year vine), 5 taxa in the middle layer (15–25 cm) of HY2 (10-year vine), and 1 taxon in the middle layer (20–35 cm) of ZN2 (10-year vine), as shown in Fig. 7b. Generally, HY (15-year vine) sample group had more biomarkers than the groups of other planting years, especially in the middle layer (20–35 cm) there were more fungal communities with significant difference.
In the cladograms, circles radiated from inside to outside represent the taxonomic ranks from phylum to genus (or species). Each circle at different taxonomic ranks represents a classification at that level, and the diameter of the circle is proportional to the relative abundance. The coloring principle is that the species with no significant difference are uniformly colored in yellow, and the Biomarkers of different species follow the group for coloring. The red nodes indicate the fungal taxa that play an important role in the red group, while the green nodes indicate the fungal taxa that play an important role in the green group. If a certain group in the picture is missing, it means that there is no species with significant difference in that group. The names of species represented by English letters in the figure are displayed in the legend on the right. The abbreviations ZN, XZ, HY, WH, GM represent the grape planting years of 5, 8, 10, 12 and 15, respectively; and the numbers 1, 2 and 3 represent the rhizosphere soil depths of 5–20, 20–35 and 35–50 cm, respectively. Statistically significant difference is defined by LDA > 4.
Source: Ecology - nature.com
