High salt concentrations are one of the key factors limiting crop growth in coastal saline soil (Zhang et al., 2015). Although washing with fresh water is effective for reducing the salt concentration, a shortage of fresh water in coastal areas precludes the application of this technique (Li et al., 2008). Alternative methods using drip irrigation with saline water for salinity control have been developed (Sun et al., 2012; Li et al., 2017b); however, these systems are prone to suffer from emitter clogging (Yavuz et al., 2010), although some flushing regimes have been developed to cope with this issue (Feng et al., 2017; Zhangzhong et al., 2018). We have recently demonstrated that freezing saline water irrigation was an effective method for reducing salinity in surface saline soil, which has been illustrated both in a lab-scale column experiment (Li et al., 2008) and in field experiments (Guo and Liu, 2015, 2021). In combination with subsequent plastic mulching, which could help maintain the low soil salinity by preventing evaporation and salt backflow after infiltration, this technique increased the abundance of bacteria and fungi, enhanced soil respiration and facilitated seed germination in cotton (Guo and Liu, 2015; Li et al., 2017a).

The freezing saline water irrigation technique involves the use of saline water during winter to form an ice layer on the surface of the soil. As saline water freezes, the salt is not incorporated into the ice crystal but concentrated in the interstitial liquid brine (Notz and Worster, 2009). The resulting saline ice is therefore a mixture of freshwater ice crystals, brine pockets, air and solid impurities (Xie et al., 2009; Gu et al., 2012). Previous research has revealed that as the temperature gradually increases in the spring, the ice salinity declines with decreasing ice volume (Guo and Liu, 2015), possibly due to the brine pockets draining from the ice under the influence of gravity (Gu et al., 2012) and/or flushed out by relatively fresh surface meltwater (Notz and Worster, 2009). During this process, salinity gradually declined in the percolated meltwater, which infiltrated into the underlying saline soil and eventually resulted in a relatively low salinity in the surface soil to facilitate seed germination and crop growth. Although this method could dramatically reduce the salinity in surface soil, due to the shallow, high salinity groundwater table in coastal areas, saline soil treated solely with freezing saline water irrigation did not support plant (cotton) growth, as indicated in prior studies (Guo and Liu, 2015; Li et al., 2017a). To maintain the low salinity level achieved by infiltration of meltwater from saline ice, mulching (M) has been introduced and proven to be an efficient way to control the salinity in the root zone (Guo and Liu, 2015). Mulching could reduce soil-water evaporation and thereby prevent soluble salt from moving up with water flow and accumulating in the topsoil (Sadegh-Zadeh et al., 2009; Fall, 2017).

While salinity is altered by amelioration practices, other physical and chemical properties are also inevitably affected, as indicated in our previous studies (Guo and Liu, 2015; Li et al., 2017a). In addition to soil physicochemical characteristics, soil microbial properties have gained attention recently and have been increasingly used as bioindicators of soil quality (Epelde et al., 2014). Previous studies have demonstrated that changes in microbial community composition during saline soil amelioration are concomitant with changes in soil salinity or other soil properties (Elmajdoub and Marschner, 2015; Lu et al., 2015; Wang et al., 2020). However, how soil microbial community responds to freezing saline water irrigation and mulching in coastal saline areas remains unclear. In particular, due to alterations in salt distribution and soil aggregation along the soil profile during these amelioration processes (Ju et al., 2018), it is of great interest to understand the accompanying vertical changes in soil microbial properties (Zahran, 1997).

Soil microbes play important roles in driving soil biogeochemical cycling (Nielsen et al., 2011; Graham et al., 2014) and maintaining soil quality (Kennedy and Smith, 1995; Clergue et al., 2005). In agricultural systems, it has been well documented that soil microbial communities are sensitive to a number of environmental factors, including soil pH (Valentin-Vargas et al., 2014; Sun et al., 2015b), moisture (Evans and Wallenstein, 2012; Ma et al., 2015), and temperature (Lauber et al., 2013; Auffret et al., 2016), and could be altered greatly by agricultural practices, such as fertilization (Poulsen et al., 2013; Hartmann et al., 2014; Sun et al., 2015b,2016) and tillage (Zuber and Villamil, 2016; Sun et al., 2018). Salinity is the major stress for microorganisms in saline soils and has been found to be negatively associated with microbial abundances and activities (Siddikee et al., 2011; Wu et al., 2015; Singh, 2016). As reviewed by Singh (2016), prior studies have demonstrated that bacterial growth clearly responded to salt concentration, but soil salinity was not proposed as a decisive factor for alterations in bacterial growth or community structure in the investigated soils because microbial communities have evolved the ability to survive saline conditions through physiological and structural modifications (Zahran, 1997). These survival strategies increased the uncertainty as to how soil microbial populations respond to soil salinity changes. In particular, how soil microbial communities shift and reassemble with decreasing salinity achieved by long-term saline soil amelioration remains poorly understood.

Since coastal areas are located close to the sea, a considerable portion of microorganisms in coastal soils are marine-sourced, and we hypothesized that the effects of marine environments on the microbial community in coastal saline soil will be weakened and microorganisms living in lower-salinity environments will be enriched. To test this hypothesis and bridge the knowledge gap regarding the responses of soil microorganisms to saline soil amelioration, microbial community composition, diversity and assembly mechanisms under different amelioration practices were evaluated through high-throughput sequencing and bioinformatic analyses.

Saline soil remediation has recently gained increasing attention due to its potential for agricultural production (Li et al., 2017a), and a number of remediation techniques, including organic amendment (Wang et al., 2014; Zhang et al., 2016; Wu et al., 2018; Yang et al., 2018; Sun et al., 2020), phytoremediation (Hasanuzzaman et al., 2014; Shao et al., 2019; Wang et al., 2020), and microorganism inoculation (Zhang et al., 2014; Arora et al., 2016), have been developed recently. Freezing saline water irrigation, which utilizes local saline water, is an economical technique for saline soil remediation and has been successfully applied in saline soils of coastal areas (Guo and Liu, 2014) and inland irrigation areas (Guo and Liu, 2021). While the effects of this technique on the chemical and physical properties have been intensively investigated (Guo and Liu, 2015, 2021; Li et al., 2017a), sparse information is available about how the microbes in soil respond to amelioration practices. The current study filled this knowledge gap and, for the first time, illustrated different response patterns among bacteria, fungi and archaea.

In this study, the different responses of soil archaeal, bacterial and fungal communities to three amelioration practices were observed. The differences in the response patterns among the taxonomic groups have also been revealed in a previous study where saline soil was amended with biochar poultry-manure compost in conjunction with a pyroligneous solution (Lu et al., 2015). These differences might be associated with their community composition and different tolerances to salinity. The archaeal communities include many neutrophilic/halophilic taxa, such as Halogranum, Halohasta, Halorubrum, and Salinigranum (Figure 2), that can only survive in environments with a high salt concentration (Lizama et al., 2002; Kim et al., 2011; Mou et al., 2012; Corral et al., 2016; Wang et al., 2016). These taxa were absent in the ameliorated soil, where a much lower salt concentration than that of the original soil was achieved (Table 1), and resulted in a lower diversity in the archaeal community. The significantly positive correlation between archaeal richness and soil EC (Spearman’s correlation coefficient: 0.734, P < 0.05) also suggested that amelioration practices decreased archaeal diversity primarily by reducing soil salinity. Consistently, our recent study on phytoremediation of coastal saline soil has also demonstrated a significant decline in archaeal richness through planting salt-tolerant plants (Wang et al., 2020). However, the amelioration practices only slightly affected fungal richness (Figure 3). One possible reason for this result is that soil fungi are more resistant to salinity stress than soil bacteria (Rath et al., 2016), as fungi have a chitinous cell wall and different energy generation systems from bacteria, both of which offer protection against osmotic stress (Thornber, 1987; Strickland and Rousk, 2010; Rath et al., 2016). Another possible reason for this phenomenon is that halophilic taxa normally fluctuate dramatically along with salinity changes; however, only one known halophilic taxon, Hortaea werneckii (Plemenitaš et al., 2014), was detected in the soil fungal community. Therefore, changes in halophilic fungi, specifically, fungal species replacement in general, contributed minimally to the changes in community richness, and the shift in fungal communities between treatments primarily resulted from changes in species abundance. For bacterial communities, although the relative abundances of some halophiles, such as Halomonas (Gunde-Cimerman et al., 2018), also decreased in the ameliorated soil, the increased bacterial richness (Figure 3) associated with the amelioration treatments indicated that these practices promoted the colonization of new bacterial taxa. An increase in bacterial diversity was also discovered in saline soil remediation with farm manure (Shi et al., 2019) and with planting Jerusalem artichokes (Shao et al., 2019).

Understanding the mechanisms that shape soil microbial community structure is a central topic in microbial ecology (Zhou and Ning, 2017). The β-null deviation results showed that microbial communities in the studied saline soil deviated from the null model in which assembly occurred at random (Figure 4), indicating the profound contribution of the deterministic process governing microbial community assembly. This result was expected since the high concentration of salt is a strong filter for most microbial taxa. This notion was supported by the significant correlation between soil EC and the soil microbial community structure (Table 2). Accordingly, the decline in the importance of deterministic processes for community assembly in the ameliorated soils (Figure 4) could probably be due to the decrease in the salt concentration, which weakened niche selection and led to an environment suitable for more diverse microorganisms. Additionally, in the current study, plants (cotton) managed to grow in the treatments with mulching (M and WIM), and the nutrient input (especially carbon) from plants might have also contributed to the decrease in the importance of deterministic processes in microbial community assembly, as nutrient input has been perceived to enhance random changes in the relative abundances of taxa within a microbial community, weaken niche selection by reducing resource competition, and strengthen priority effects (Chase, 2007), thereby increasing the unpredictability of microbial composition and the stochasticity of microbial assembly (Zhou et al., 2014).

A number of species found in the Control soil have been found in sea water or marine sediment, such as Marivirga (Lin et al., 2015), Gramella (Lau et al., 2005), Pelagibius (Choi et al., 2009), Idiomarina (Ivanova et al., 2000), Marinobacter (Raddadi et al., 2017), and Marinobacterium (Chang et al., 2007), indicating the pronounced influence of the ocean on the microbial composition of coastal saline soil. However, the relative abundances of these taxa decreased in the ameliorated soils, especially in WIM-treated soil (Figure 2), suggesting that the amelioration practices weakened the effect of seawater on the microbial communities. A similar propensity was also observed in the functional prediction analyses, where the relative abundance of aerobic chemoheterotrophy, a function that has been found to be dominant in marine systems (Louca et al., 2016), was significantly decreased by amelioration (Figure 5).

Among the three amelioration treatments, WIM resulted in the greatest changes in soil microbial communities, with a decreasing trend in autotrophs and an increasing trend in heterotrophs in the community. This could be attributed to the plant (cotton) growth in WIM introducing exogenous carbon inputs that support the growth of heterotrophic microorganisms. Profound effects of plant growth on the composition of the rhizosphere microbiome have been revealed in previous studies (Mukhtar et al., 2018; Wang et al., 2020). Higher relative abundances of arbuscular mycorrhizal fungi (AMF) were observed in treatments that support plant growth (M and WIM) than in the Control soil (Figure 5C). This could be due to the increase in the abundance of AMF that were in symbiosis with cotton, since most of the AMF were within the family Glomeraceae (Supplementary Table S3), members of which have been reported to be commonly associated with cotton (Nehl and McGee, 2010). Furthermore, previous studies reported that AMF can alleviate salt stress and promote growth and yield in cotton (Khaitov and Teshaev, 2015).

Nitrogen is one of the most important elements influencing crop yield (Greenwood, 1982), and nitrogen transformations in soil are primarily driven by soil microbes (Simon and Klotz, 2013; Sun et al., 2015a). The current study revealed that some functional groups involved in nitrogen cycling were significantly impacted by the amelioration practices. Notably, some archaeal and bacterial functional groups catalyzing the same nitrogen cycling processes responded differently to the amelioration practices (Figure 5). The archaeal aerobic ammonia oxidation functional groups increased in WI- and WIM-treated soils, whereas their bacterial counterparts were not significantly affected. This observation was supported by qPCR results from the archaeal and bacterial amoA genes, which showed that the abundance of the archaeal amoA gene was significantly increased by WI and WIM ameliorations, but bacterial amoA gene abundance showed minimal variation across treatments (Supplementary Figure S2). The different responses of oxidizing archaea (AOA) and bacteria (AOB) to agricultural management and their niche separation have been widely reported (Schleper, 2010; Prosser and Nicol, 2012; Stahl and de la Torre, 2012; Segal et al., 2017; Xiang et al., 2017; Sun et al., 2019). In the present study, the higher abundance and greater sensitivity of AOA compared with AOB might indicate their important role in ammonia oxidation in this coastal saline soil.

In this study, the microbial community composition, diversity, assembly mechanisms and potential functions in costal saline soil under the effects of freezing saline water irrigation and plastic mulching were investigated. Microbial communities in coastal saline soil are composed of a high relative abundance of marine-associated taxa, which were reduced by amelioration practices. Amelioration practices exerted contrasting effects on the richness and vertical distribution among bacterial, archaea and fungal communities. β-null deviation analysis showed that the deterministic process played a dominant role in microbial community assembly in this coastal saline soil and that its effect was decreased in ameliorated soil. These results, combined with the significant correlation between soil microbial communities and soil salinity, indicated that the amelioration practices led to a transition of microbial communities primarily through weakening the strong selectivity of high salinity and forming a microbial community more suitable for agricultural production. These findings advanced our understanding of the mechanisms underlying coastal saline soil amelioration and provided valuable information for the development of potential biofertilizers for use in salt-affected soils.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ebi.ac.uk/ena, PRJEB29290.

BL, XL, and RS conceived the project. RS, XW, and YT performed the laboratory experiments and analyzed the data. KG, XF, HS, and XL designed and setup the field experiments. RS and BL wrote the manuscript. All authors discussed the results and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.