In this study, we identified four Orders of AMF. The relative diversity of these four Orders varied significantly, which were reflected in the differences of the four genuses, with Glomus, Diversispora, and Ambispora genus accounting for 83.46%, 14.73%, and 1.21% of the fungi in analyzed samples, respectively (Fig. 3a). In addition, some AMF could not be identified, and Paraglomus was detected in few samples. Some molecular studies of AMF communities have reported the difficulty of detecting Paraglomerales and Archaeosporales [24, 25], which are found in some soils. Alternatively, that suggests potential PCR primer-related issues, given that 18S rRNA primers are unable to identify all AMF species, with particular limitations in the amplification of certain Paraglomerales species [26]. Owing to its adaptability, Glomus species are abundant in many ecosystems, which is the same as our study. As Glomus and Diversispora were the two primary genera detected in soil samples in the present study, this suggests that these fungi are better adapted to the desert environmental at this study site.
Our petal chart analysis revealed 12 core OTUs shared among our different soil sample groups (Fig. 2), including 10 Glomus OTUs and 2 Diversispora OTUs that were present at all soil depths and slope positions. Given their universality among collected samples, we hypothesize that these fungi are closely associated with Ferula sinkiangensis growth, potentially suggesting that further study of these fungi may offer key insights into soil microbiology that can support artificial Ferula sinkiangensis cultivation.
In LEFse analyses, we found that biomarkers [27–29] differed significantly as a function of soil depth and slope position. With decreasing biomarker levels as soil depth increased, suggesting that certain AMF species are sensitive to soil depth (Fig. 8b). We also found that most of these soil depth-sensitive AMF biomarkers were located in the bottom position. In combination with Fig. 8a, it is showed that most AMF biomarkers were enriched at a soil depth of 0–20 cm in samples collected from the bottom of the slope, which may be a consequence of the fact that plant residues typically accumulate on the soil surface [28, 30, 31], particularly on relatively flat bottom slope. Such residues created high soil, nutrient content, good ventilation and hydrothermal conditions, which is very conducive to the growth of soil microorganisms. Moreover, partial microorganism can work synergistically with other AMF species [32, 33] to promote Ferula sinkiangensis growth.
Spearman correlation analyses revealed that soil physicochemical properties were significantly associated with AMF alpha diversity indices, with TP and pH being positively correlated with Shannon and Chao1 index values (p < 0.05). Soil phosphorus levels are one of the most important factors regulating AMF community diversity [33, 34], with certain studies having found that the diversity of AMF is significantly negatively correlated with AP [34, 35]. Herein, we found AMF diversity to be significantly positively correlated with soil TP (p < 0.05), whereas it was not significantly related to levels of available phosphorus. This may be related to the low levels of available phosphorus in these soil samples, with levels ranging from 1.67–6.85 mg/kg. It has been shown that the AMF function of providing P to their host plants is phylogenetically conserved [36], and that different AMF phylogenetic groups would prefer to colonize roots under different P availability [35]. Consequently, it is possible that low P availability would like to select (environmental filtering) functionally similar AMF (e.g. high P-uptake efficien cy) from the local species pool. In contrast, TP contents varied from 0.49–0.85 g/kg, suggesting high potential phosphorus abundance in these soil samples. Ferula sinkiangensis growth is dependent upon the absorption of available soil phosphorus, and AMF species can facilitate such phosphorus uptake [35, 36]. This thus explains the increase of TP content consistent with the substantial enrichment of AMF species which adapted to low AP content within the rhizosphere.
Soil pH is another key parameter influencing AMF community diversity, with AMF diversity often being found to be significantly negatively correlated with pH [37]. In contrast, in the present study we found that AMF community diversity was significantly positively correlated with soil pH, potentially due to unique local environmental factors. Some studies have shown that tolerance to different AMFs varies greatly [37, 38]. The soil pH range in the present study was from 7.80–8.81, with only certain AMF species including Glomus and Diversispora being able to survive in this pH range. As pH values rose, we found that the richness and diversity values corresponding to these AMFs also increased.
In this analysis, we detected significant differences in AMF diversity and richness as a function of slope position but not as a function of soil depth. These differences may be related to differences in soil composition at different slope positions. Cluster analyses clearly separated soil samples into three categories, which indicated that AMF community composition at a given slope level was similar, whereas this composition varied significantly as a function of slope level. We additionally observed no significant differences in soil properties as a function of soil depth, whereas these properties did differ at different slope positions, with significant differences in OM, TP, TK, AN, PH, SM (p < 0.05). AMF diversity and richness were closely associated with environmental factors, and CCA analyses revealed that organic matter, total nitrogen, total potassium, nitrate nitrogen, ammonia nitrogen, available potassium, total salt, pH, water content, and slope all had a significant impact on AMF community composition. AE was found to be positively correlated with TDS and DE, and to be negatively correlated with other environmental factors. These factors were also correlated with AMF community composition, with OM, TN, TP, TK, AN, NN, TDS, PH, and SM all being positively correlated with many OTUs, whereas AE and TDS were negatively correlated with many OTUs. Soil composition thus differed significantly as a function of slope position, in turn affecting AMF community diversity and richness.