4 DISCUSSION
Herein, we found that desertified land reclamation for viticulture was associated with significant changes in soil microbial communities and soil properties, consistent with our hypotheses. Such land reclamation was associated with significant increases in SOC relative to that observed in DL soil, consistent with the land conversion to vineyards having increased soil organic matter content. Organic matter inputs include winegrape litter, which can additionally contribute to the soil nutrient content as well as fertilizer and manure applied in the context of viticulture, thereby contributing to SOC and nutrient availability. C soil samples exhibited the highest SOC levels (Table 1), likely because the soil is less disturbed in these fields and because the plant leaves contain secondary metabolites such as tannins, terpenes, and polyphenols that slow the rate of litter decomposition. Yanget al. (2016) have also previously reported reclamation-related increases in SOC and nutrient contents. We found that DL reclamation for winegrape cultivation was associated with a significant increase in rhizosphere soil pH (Table 1), in contrast to prior study results (Djukic, Zehetner, Tatzber, & Gerzabek, 2010; Franco-Otero, Soler-Rovira, Hernández, López-de-Sá, & Plaza, 2012), potentially due to study site-specific differences. Indeed, the soil alkalinity in the Hongsibu area of Ningxia in this study was higher than in the previous studies.
Winegrape planting was associated with a reduction in soil volume that may be attributable to the long-term plowing associated with the viticulture process. Similar changes in TN and SOC were also observed, whereas TP and TK did not differ significantly among the five analyzed land-use applications, potentially owing to the high phosphorus and potassium content of desertified soil minerals (Hendrix, 2000; Ma, Sun, Sun, & Wang, 2012; Gong, He, & Liu, 2016). Soil available potassium rose significantly following reclamation, consistent with SOC accumulation. In contrast, soil available nitrogen and available phosphorus content decreased significantly after reclamation, likely due to the absorption of these nutrients by grape plants at a rate greater than that at which fertilized nitrogen and phosphorus were applied (Gong, He, & Liu, 2016). Soil nutrient contents observed in this study were generally lower than those reported in other regions (Domínguez, Panettieri, Madejón, & Madejón, 2020), consistent with the fact that winegrapes are suited to cultivation in regions with harsh ecological conditions and relatively poor soil conditions.
The composition and diversity of bacterial communities in a given ecosystem can change in response to environmental variation (Cheng, Chen, & Zhang, 2018). We therefore hypothesized that bacterial community would differ significantly when comparing DL and reclaimed soils, and confirmed that this was true at the taxa level (Figure 2; Figure 3), through NMDS analyses (Figure 4), and via a hierarchical clustering analysis (Figure 5). We observed distinct differences in the relative abundance of top bacterial phyla among land-use types, indicating that DL reclamation was linked to altered microbial community composition. Actinobacteria were the most abundant bacterial phyla in the present study, and Actinobacteria abundance was significantly higher in CS, M, C, and IR soils relative to DL soils, consistent with prior reports (Gao et al., 2020). While pesticide use in vineyards can impact soil bacterial communities, Actinobacteria , which degrade hydrocarbons and pesticides (Alvarez et al., 2017), are not impacted by pesticides. Proteobacteria were also dominant in the present study, with higher relative Proteobacteria abundance in reclaimed soil relative to DL soil. Relative Proteobacterialevels were significantly higher in M and IR soil samples relative to DL soil samples, potentially due to the higher sulfur levels in vineyards. Sustained livestock organic fertilizer application and the input of organic matter in M and IR vineyards can contribute to these increases in soil sulfur levels. Indeed, shifts in Deltaproteobacteriaabundance have been linked to changes in soil sulfur content (Wasmund, Mußmann, & Loy, 2017). We also observed reclamation-related changes in the abundance of Acidobacteria , which are frequently reported to be regulated by soil pH (Ramirez et al., 2014). In line with the results of a prior study (Lauber, Hamady, Knight, &Fierer 2009), we found that relative Acidobacteria abundance declined with rising soil pH. We additionally found that bacterial abundance (Chao1) and diversity (Shannon) in C and IR soils were significantly increased relative to DL soils (Table 2), owing to differences in soil chemical properties.
With respect to soil fungal community composition and diversity, we found that Ascomycota was the dominant phylum among the analyzed land-use patterns, followed by Basidiomycota , in line with prior reports (Maestre et al., 2015; Porras-Alfaro, Herrera, Natvig, Lipinski, & Sinsabaugh, 2011; Prober et al., 2015). We found that winegrape cultivation was associated with increased fungal diversity, although there were no significant differences in fungal diversity indices associated with different land use types (Table 2). This may be because Bordeaux and lime sulfur mixtures are applied to these vineyards to prevent fungal diseases, thereby reducing fungal diversity, whereasAscomycota species can tolerate harsher conditions (Chen et al., 2017). NMDS and hierarchical clustering results indicated that microbial community structures differed significantly in cultivated soils relative to DL soils, but the bacterial and fungal community structure was similar in CS and M (Figure 4; Figure 5). This may be a result of changes in land-use that led to similar root exudate and litter properties in CS and M soils, thus inducing similar rhizosphere bacterial and fungal community structures. However, the bacterial and fungal community structures differed significantly in M and C samples, both of which were 3 years old, indicating that the microbial community structure differences in the context of short-term cultivation (for less than 5 years) were more closely related to winegrape cultivar type. In addition, LEfSe analyses directly assess differences at all taxonomic levels at once, and can identify differences in species between groups as biomarkers of a particular sample type (Figure 6). These biomarkers may contribute to long-term sustainable reclaimed land use, although further study of this topic is required.
Soil microbial diversity is crucial to ensuring long-term soil ecosystem sustainability (Cheng et al., 2017; Maron et al., 2018). Assessing soil microbial diversity in a reclamation ecosystem can thus offer perspectives on long-term sustainable development strategies for reclaimed desert land. Herein, we found that bacterial and fungal ASVs/OTUs (Figure 2) and diversity index (Table 3) were higher in reclaimed vineyard soils relative to DL soil samples. These changes were attributed to increased organic matter input in the reclaimed soil owing to cultivation-related fertilizer application and associated crop residues. However, we did find that when analyzing fungal ASVs/OTUs, the Chao1, and Observed species index values in C soils were lower than in DL soils, suggesting a decrease in mycorrhizal dependence associated with Chardonnay grape cultivation. Winegrape cultivation is thus generally associated with increased fungal and bacterial diversity owing to the increased diversity of plants in the vineyard rows (Kowalchuk, Buma, de Boer, Klinkhamer, & Van Veen, 2002). Plant diversity and associated rhizosphere carbon input can help drive increased soil microbial diversity (Lange et al., 2015; Meng et al., 2019). Higher microbial diversity is often associated commonly linked to increased ecosystem stability (Delgado-Baquerizo et al., 2016). Therefore, a viticulture model of interrow grass planting with higher plant diversity should be adopted to facilitate the sustainable development of land use in desertification areas.
Several prior reports have demonstrated the important link between soil chemistry and microbial community composition in a variety of ecosystems (Kerfahi, Tripathi, Dong, Go, & Adams, 2016; Nanganoa et al., 2019; Tian et al., 2017). SOC, pH, and nutrient contents have been reported to be the key determinants of bacterial community composition and diversity in vineyards (Li, Chi, Li, Wu, & Yan, 2019). Our correlation analyses revealed SOC, ammonium, nitrate, and bacterial diversity indices to be positively correlated (Table S). As such, the increase in SOC levels in reclaimed soil may contribute to elevated bacterial diversity in reclaimed soils in this particular ecosystem, given that SOC is a key resource necessary for the majority of terrestrial microbial communities (Cai et al., 2016; Zumsteg et al., 2012). RDA analysis further revealed that the relative abundance of the dominant soil bacteriaProteobacteria was positively correlated with SOC. This is consistent with prior work demonstrating that bacterial community composition in semi-arid agricultural ecosystems is driven by SOC (Cheng et al., 2017). Soil pH is also thought to be an important determinant of soil bacteria community structure, diversity, and richness in a range of ecosystems (Fierer, & Jackson, 2006; Ramirez et al., 2014). We observed significant positive correlations between bacterial community structure (Chao1, Observed species, and Shannon indices) and soil pH (Table S). Overall, soil pH explains 42.20% of the total variation of soil bacterial communities at the phylum level (Figure 7), suggesting that pH is of critical importance in shaping the structure of bacterial communities in reclaimed desertified soil. In addition to SOC and pH, elevated nutrient levels in cultivated land may ameliorate nutrient-dependent constraints on bacterial growth and function in otherwise nutrient-poor desert soils, thus driving increased bacterial diversity in crop ecosystems.
Herein, Ascomycota species were the dominant fungi and were positively correlated with SOC and TN (Figure 7). Soil fungal community diversity has been reported to be associated with nitrate-nitrogen, TN, SOC, and pH (Ferreira, Leite, de Araújo, & Eisenhauer, 2016; Huhe, Chen, Hou, Wu, & Cheng, 2017), in line with our data. Relative to bacteria or archaea, fungal survival is more dependent upon carbon and nitrogen sources (Schmidt, Nemergut, Darcy, & Lynch, 2014). Soil nitrogen and SOC regulate the structure of both bacterial and fungal communities in the soil. Overall, our data suggest that soil physicochemical properties are significantly altered following desert land reclamation, thereby increasing fungal and bacterial diversity, which is essential to ensure crop ecosystem stability and long-term sustainable land use.