Introduction

Currently, engineered microbiome integration in agricultural soils has become a hot topic due to its proposed associated benefits (Kumar and Dubey 2020). However, the soil incorporates a highly dynamic ecosystem with continuously changing environmental conditions (Jansson and Hofmockel 2020), and the predicted microbial functioning on the other hand follows highly environmentally dependent patterns (Thompson et al. 2017). In response to these changing environmental conditions, microbial communities either adapt and thrive or fail to adapt and die off (Wu et al. 2019). This varying response is the reason behind the low success rate of most lab-scale engineered microbiome technologies when applied to fields. Therefore, to improve the success rate of engineered microbiome incorporation into various soil conditions, it is necessary to build an understanding of the generalized challenges that a microbe can withstand in that particular environment. In this review, we propose the concept of a microbiome-mediated smart agriculture system (MiMSAS) and summarize some of the advanced strategies for engineering the rhizosphere microbiome to withstand the stresses exerted by the dominant microbial determinants in soil. This work will assist the scientific community in increasing the success rate of engineered microbiome technologies for enhancing crop productivity and environmental sustainability.

Complex microbial ecosystem: the rhizosphere

The rhizosphere is defined as the area that surrounds the plant and is under the direct influence of the biophysicochemical characteristics of the plant roots (Dries et al. 2021). These are microbial houses with high complexity (Chen et al. 2018; Lee et al. 2019), the so-called rhizosphere microbiome. Some of the microbiome members act as crucial partners to associated plants for their health, fitness, and metabolic functioning, such as phytohormone production, enhanced nutrient uptake, nitrogen fixation, phosphate solubilization, defence responses, and ultimately pathogen suppression (Backer et al. 2018; Compant et al. 2019; Gopal et al. 2013; Mitter et al. 2019; Ortiz and Sansinenea 2021; Timmusk et al. 2017). A healthy microbiome supports the associated plants under biotic and abiotic environmental stresses (Li et al. 2020; Priya et al. 2021). A recent study reported the role of root-associated bacterial taxa as a disease-suppressive protective layer of plant defence against potential pathogens (Carrión et al. 2019). Although a rhizosphere microbiome can promote plants by so-called induced systemic resistance (ISR) against potential pathogenic risks (Li et al. 2020; Trotel-Aziz et al. 2008), an unhealthy rhizosphere can lead to the colonization of soil-borne pathogens (Berlanas et al. 2019). Therefore, for a healthy cropping system, the primary concern is that the rhizosphere is occupied by a healthy, diverse, and active microbiome (Di Giacinto et al. 2020; Mueller and Sachs 2015; Pineda et al. 2017; Wallenstein 2017). Figure 1 depicts some of the most remarkable microbial benefits to the associated environment and host plants.

Fig. 1
figure 1

Rhizosphere microbiome-mediated advantages for associated host plants and soil nutrient cycling (image created with BioRender.com)

Full size image

Biotic and abiotic stresses modulate the recruitment pattern of the rhizosphere microbiome

The rhizosphere microbiome usually consists of a variety of symbiotic, mutual, and competitive microbial interactions (Mendes et al. 2013). Proteobacteria, Firmicutes, and Bacteroidetes are the consistently dominant bacterial groups in the rhizosphere, while Ascomycota and Basidiomycota are the most abundant fungal colonizers (Trivedi et al. 2020). All of these successful colonizers either compete for resources or form mutually cooperative coexisting communities. In the rhizosphere environment, microorganisms are exposed to a variety of abiotic and biotic stresses (Fig. 2), which may exert a number of individual or composite effects on modulating the community structure (Goss-Souza et al. 2019; Shi et al. 2018; Wu et al. 2019) and ultimately impact the associated plant health. The major determinants of the community structure include the physicochemical properties of the rhizosphere soil (Goss-Souza et al. 2019; Wu et al. 2019), the variety of plant-derived root exudations (Vieira et al. 2020), and the secondary metabolites produced from other soil microbes (Koo and Cho 2006), as shown in Table 1.

Fig. 2
figure 2

Influences of biotic and abiotic factors on soil microbiome (image created with BioRender.com)

Full size image
Table 1 Studies showing soil microbiome development triggered by biotic and abiotic factors
Full size table

Abiotic stresses tempering the rhizosphere microbiome

Although the effect largely varied in different studies, the most acceptable descending hierarchy of the abiotic factors affecting microbiome structure are salinity (Canfora et al. 2014; Hollister et al. 2010; Xie et al. 2017), pH (Bahram et al. 2018; George et al. 2019; Kaiser et al. 2016), organic carbon and nitrogen contents (George et al. 2019; Maestre et al. 2015; Siciliano et al. 2014), moisture content (George et al. 2019; Lennon et al. 2012), tillage practice (Mathew et al. 2012), and temperature (Zhou et al. 2016) (Fig. 2). However, it is worth noting that the influential intensity of these factors fluctuates across microbial domains (Deng et al. 2018). Thus, the varied microbiome in different environments would demonstrate differentiated responses to abiotic factors.

The microbial threshold level for salinity is 5% (Triadó-Margarit et al. 2019); beyond this, it adversely affects cellular growth and respiration by generating shifts in core metabolic functioning (Canfora et al. 2014; Dupont et al. 2014; Iwaoka et al. 2018; Rath et al. 2019b; Wang et al. 2018). There has been a declining trend in phylogenetic diversity and species richness with increasing salinity, especially for bacteria (Canfora et al. 2014; Hollister et al. 2010; Ibekwe et al. 2010; Xie et al. 2017; Xue et al. 2018b). However, the effects on fungal communities are still contradictory (Kim et al. 2019; Rath et al. 2019b). Now that high salinity impacts active rhizosphere communities, it is no surprise that there is decreased cropping productivity in saline water irrigation systems and coastal land areas. A healthy cropping system requires a highly diverse and active rhizosphere microbiome (Di Giacinto et al. 2020; Mueller and Sachs 2015; Pineda et al. 2017; Wallenstein 2017).

Furthermore, in different soil systems, microbial distribution depends upon their tendency for pH tolerance. Extreme pH (acidic or alkaline) can significantly impact microbial networks by hindering enzymatic activities and changing the residing conditions, including nutrient supply. Scientists have reported that the highest pH soil transects are associated with the lowest microbial abundances (Xue et al. 2018b) and neutral pH conditions with higher diversities (Kaiser et al. 2016; O’Brien et al. 2019).

In addition, the nutritional status of the soil stands as another major challenge for the rhizosphere microbiome. The microbiome structure highly correlates to the content and chemistry of organic carbon, and even closely related microorganisms have different preferences for various carbon sources (Castle et al. 2016). Bacterial populations are believed to use more labile organic compounds than fungal decomposers (Rath and Rousk 2015). These fluctuations shift microbial networks by promoting the abundance of specific groups of organisms while downregulating the abundance of other members and ultimately disrupting the microbiome structure and core metabolic functioning (Nemergut et al. 2010; Zhu et al. 2017).

Drought stands as another obstacle for a healthy microbiome. Water is the fundamental ingredient for all forms of life, including microbes, and acts as a resource, solvent, and transport medium (Bailey et al. 2017; Schimel 2018). Soil microbial activities are extremely sensitive to water fluctuations, and drought conditions affect activities either directly by exerting microbial cellular stresses or indirectly by altering substrate transportation to microbes. As a direct effect, cellular stress leads to the loss of cell turgidity and eventual death by functional failure (Schimel 2018). Concerning the indirect effects, when the water potential drops to a certain threshold, insolubilized substrates become concentrated in the cell surroundings and cannot be transported passively to cells due to blockage of pores (Manzoni et al. 2012). Ultimately, to survive in such conditions, microbial cells require facilitated substrate transportation across the cell, which may cause community shifts due to varying microbial responses. The literature has already documented the strong role of water stress in restructuring the rhizosphere microbiome (Fitzpatrick et al. 2018; Xu et al. 2018).

Compared to the abiotic factors discussed above, temperature stress may have moderate indirect effects on microbiome structure, as plants are more sensitive to temperature fluctuations and follow the pattern in which root exudations increase with increasing temperature and vice versa, which may ultimately influence the abundance of sensitive microbial groups within the rhizosphere (Pathan et al. 2020). According to a metagenomics analysis of 7,560 topsoils from 189 global sites, temperate habitats (mesophilic) exhibit the highest microbial diversity (Bahram et al. 2018). As the global temperature rises slowly, it allows more time for community adaptation and might be the reason for showcasing the more temperature-acclimatized microbiome structures.

Agricultural amendments tempering the rhizosphere microbiome

In farmland soils, management practices such as tillage, synthetic fertilization, pesticides, and herbicides are being extensively used (Cai et al. 2017; Chen et al. 2017; Hartman et al. 2018; Mathew et al. 2012; Qian et al. 2018). These practices can greatly impact the soil’s physicochemical properties, resulting in significant structural and functional shifts in the microbiome (Hartman et al. 2018; Mathew et al. 2012). Cropping practices account for approximately 10% of the total structural variation in the native soil microbiome (Hartman et al. 2018). Tillage has been detected as a primary driver of the bacterial community, whereas fungal structures were found to be profoundly influenced by fertilization, with some additional effects of tillage (Hartman et al., 2018). Tillage causes disturbance in soil and distorts tolerant species by modifying their habitats, disrupting the networks of nutrient passage, and damaging the connectivity among species (Goss-Souza et al. 2019).

Synthetic fertilizers are being extensively used to improve soil fertility and crop yield. Excess and long-term inputs raise the nutrient level, which affects potential rhizosphere symbionts (such as mycorrhizal fungi and diazotrophs) (Romero et al. 2012) and the colonization of growth-promoting rhizobacteria (Wang et al. 2017). This, in turn, tends to increase the abundance of fast-growing copiotrophic taxa in soil, influencing the native microbiome structure (Ramirez et al. 2012). In parallel, inorganic fertilizers alter soil properties, for example, causing soil acidification by lowering the soil pH (Cesarano et al. 2017). This, in turn, significantly impacts the resident microbiome (Fierer et al. 2012; Leff et al. 2015; Ramirez et al. 2012; Xue et al. 2018a) and function (Ji et al. 2021; Ramirez et al. 2012; Yang et al. 2017). In short, the chemical form of the fertilizer, whether organic or inorganic, is a critical determinant of the rhizosphere microbiome.

As a parallel practice in agriculture, synthetic pesticides are widely used to kill insects, rodents, and unwanted weeds. Pesticides have a direct effect on the microbiome by compromising rhizosphere immunity and fostering potential infections (Qu et al. 2020) or an indirect impact by changing plant exudations (Qian et al. 2018). However, the delivery method and mode of action of such pesticides appear to be key determinants of plant-associated microbial responses, and depending on them, some compounds have more detrimental effects, while others show moderate reversible damage (Tosi et al. 2020).

Biotic stresses tempering the rhizosphere microbiome

The rhizosphere is a habitat for complex microbial interactions that can shape the structure of the microbiome (Marques et al. 2014). Microorganisms can eliminate or suppress other comembers through a variety of mechanisms, such as antagonism, pathogen resistance, quorum sensing, and food and shelter competition (Raaijmakers and Mazzola 2016). Some microbes have a dominant role in organizing dynamic root-associated microbiomes by secreting volatile organic compounds in the form of metabolites (Bergelson et al. 2019; Toju et al. 2016). Fungi, for example, secrete antibacterial compounds, promoting the abundance of more competent bacteria while eradicating nonresilient bacteria (Boer et al. 2005). A study reported that Actinomycetes protect the host olive plant from the potential pathogen Staphylococcus aureus through various antipathogenic metabolites (Dede et al. 2020). Likewise, Bacillus subtilis specie rescues the host plant from potential fungal pathogens through antifungal activity (Walia et al. 2013).

Furthermore, saprophagous soil animals such as earthworms and protozoans have also been shown to have moderate effects on the taxonomic structuring of the soil microbiome (Braga et al. 2016; Rosenberg et al. 2009; Thakur and Geisen 2019). Although these biological interactions are generally less lethal, they can be more detrimental when coupled with abiotic stresses or pathogenic infections.

Plants exhibit some aboveground and belowground attributes that attract specific groups of microbes into the rhizosphere. Roots serve as a microbe recruitment hub for plants, and microbial species richness is related to the chemistry and amount of root exudates (de Faria et al. 2021; Vives-Peris et al. 2018). Recent studies have shown that under salinity and water stress, root exudations attract unique bacterial communities that promote the plant’s photosynthetic ability and other necessary mechanisms that help the plant withstand the proposed stresses (Cesari et al. 2019; Xiong et al. 2020). Among different plant species and in response to different stress conditions, roots can secrete unique compounds that activate or suppress certain members of the microbial community and thus determine the microbiome composition. It is well documented that the plant immune system is intimately linked to the associated microbiome, and aboveground plant pathogenic infections can therefore influence the structure of the belowground microbiome (Qu et al. 2020). When infected with Pseudomonas syringae pv. tomato, the model plant Arabidopsis thaliana changed its exudation pattern and recruited such microbes, which helped the plant resist pathogens (Yuan et al. 2018). Moreover, plant developmental stages and genotypes can also significantly shape microbial diversity within the microbiome (Chaparro et al. 2014; Marques et al. 2014; Qiao et al. 2017). These findings are clear enough to reflect the strong influence of plant phenotypic and genotypic characteristics in determining the microbiome associated with its root system.

In summary, a healthy, diverse, and active microbiome is a prerequisite for a better life of the host plant, as a healthy microbiome assists the plant in all vital functions such as defence, nutrient uptake, carbon sequestration, and phosphate solubilization (Backer et al. 2018; Compant et al. 2019; Gopal et al. 2013; Mitter et al. 2019; Ortiz and Sansinenea 2021; Timmusk et al. 2017). However, the assembly of the rhizosphere microbiome is a complex and dynamic process highly prone to the effects of a variety of soil physicochemical and biological factors. Thus, the microbiome in agricultural fields largely varies due to discrepancies in the local environment, history of tillage, and farmland management. The attempt to create a good rhizosphere microbiome in terms of agriculture has not always succeeded due to the difficulty of disentangling the impact of confounding influential factors (Hollister et al. 2010; Mark Ibekwe et al. 2017; Rath et al. 2019a; Wang et al. 2019; Xie et al. 2017). In fact, the incorporation of a more stable and active microbiome appears to be the most promising strategy for a secure agricultural future. In the following sections, we compile some of the advanced engineering strategies to target various aboveground and belowground potential attributes for developing more stable microbiomes capable of withstanding certain dominant fluctuations and stresses in various field conditions.

Microbiome-mediated smart agriculture system

Remodelling the existing microbiome by cultivation practices: a traditional approach

Soil management practices can be designed in such a way to recruit microbial networks in a desirable fashion (Fig. 3). The important agronomic measures in this regard include cultivation practices (Edwards et al. 2015; Lourenço et al. 2018; Schmidt et al. 2019; Ullah et al. 2020) and cropping practices (Banerjee et al. 2017; Bell et al. 2019a; Venter et al. 2016; Zhang et al. 2019).

Fig. 3
figure 3

Strategies for microbiome manipulation, 1) SynCom inoculation, 2) routing the existing soil microbiome, and 3) plant-based soil microbiome steering (image created with BioRender.com)

Full size image

Crop rotation

Typically, after harvesting conventional crops, most fields are left bare until the new season arrives. It has been observed that after a crop is harvested, the loss of moisture content and other organic and inorganic essential nutrients causes soil cracks and infertility (Bhuyan et al. 2020). Consistent crop rotations, such as growing maize after rice, can promote the recruitment of a beneficial rhizosphere microbiome and, ultimately, promote microbial-assisted nutrient cycling in the soil (Gan et al. 2015; Venter et al. 2016). According to a report, crop rotation can induce approximately 10% of the variation in the native microbiome (Hartman et al. 2018). Nitrogen, for example, is an essential nutrient for plant growth and productivity, but it is also a limiting nutrient in most terrestrial ecosystems (often lost via ammonia volatilization) and is supplied as an external source. In the process of symbiotic nitrogen fixation, biological nitrogen fixer diazotrophs inhabit the root nodules of legumes and aid in the biotransformation of inert dinitrogen (N2) into ammonia (an acceptable form of nitrogen for plants) (Ryan et al. 2009). These heterotrophic diazotrophs use carbonaceous root exudates and photosynthetically sequestered carbon from the host plant and enzymatically catalyse nitrogen fixation (Biswas and Gresshoff 2014). Such crop rotation with suitable legumes and cereal crops promotes the level of available nitrogen not only for the crop itself but also based on the “niche complementarity effect”, where the remaining fixed nitrogen can be used by the subsequent crops (Adhikari and Missaoui 2017; Ryan et al. 2009). A detailed metagenomics investigation revealed that, when compared to other sites, Taoyuan fields have more nitrification functional genes and a high abundance of nitrogen fixation taxa, which contribute to extensive nitrogen fixation, launching Taoyuan as the site with the ultrahigh rice yield in China (Guiñazú et al. 2010). Therefore, appropriate cropping rotations improved soil fertility by strengthening microbiome and microbial-assisted nutrient cycling.

Organic fertilization

The drawbacks associated with chemical fertilization during extensive use in agricultural systems have become a major concern. Excess fertilizer inputs can have a detrimental impact on native microbiome structures (Cai et al. 2017; Cesarano et al. 2017; Hartman et al. 2018; Ji et al. 2021; Mathew et al. 2012; Schmidt et al. 2019; Zhou et al. 2017). In contrast, the plant-associated microbiomes in nutrient-deficient soils are extremely susceptible to colonization by latent opportunistic competitors (Finkel et al. 2019; Pagé et al. 2019). These nutrient-stimulated microbial shifts are likely to have significant negative impacts on associated plant health. Organic amendments, on the other hand, can be used as efficient replacements (Sankar Ganesh et al. 2017). Traditional nutrient-enriched organic inputs such as manure and cropping residues can promote microbial diversity by improving soil nutritional status and physicochemical properties (Saeid and Chojnacka 2019). The addition of sheep manure has been shown to improve crop yields in contaminated soils by increasing microbial diversity and decreasing heavy metal toxicity (Elouear et al. 2016). Compost, such as biocomposts or vermicomposts made from farming waste, is a suitable substitute for synthetic fertilizers (Hellequin et al. 2019). They are eco-friendly, nutrient-rich biostimulants that can promote a healthy microbiome while suppressing soil-borne plant pathogens (De Corato 2020). Further a moderate derivative, biochar is a carbon-rich product of the partial or complete pyrolysis of residual organic waste that improves soil nutritional status and water-holding capacity (Arif et al. 2020). It has the ability to suppress a wide range of fungal and parasitic plant pathogens. The rhizosphere microbiome of lettuce plants was altered by soil amendment with wood biochar, which suppressed the potential plant pathogen Fusarium oxysporum (Bonanomi et al. 2018).

Yet to achieve the best fertilization rates for improving yields while minimizing short- and long-term environmental impacts, a thorough understanding of each unique scenario and a comprehensive assessment of the individual effects of each fertilizer practice on soil and plant-associated microbiome health are required.

Biofertilization

The term biofertilization refers to biological amendments in soil intended to improve soil health by boosting soil microbiome efficiency (Mahmud et al. 2021). It consists of individual or grouped inoculations of known microbial strains (e.g. Rhizobium spp.). Biological fertilizers have been shown to be the best alternative in terms of soil biochemical properties and microbial diversity. As we discussed in the previous section, nitrogen is an essential but limited nutrient for plants, so soil amendments containing nitrogen-fixing bacteria are an example of the most suitable biofertilizer for the crops. In addition to plant nodule-residing symbionts, free-living nitrogen-fixing bacteria such as Azospirillum, Azotobacter, Rhizobium, Pseudomonas, and Burkholderia play an important role as biofortifying agents (Ryan et al. 2009; Timmusk et al. 2017). Another limiting nutrient in the rhizosphere is plant-available phosphate. The solubilization of mineral phosphate in soil is performed by a variety of soil bacteria (Arif et al. 2020; Mahmud et al. 2021), such as the genera Bacillus, Rhizobium, Pseudomonas, Burkholderia, and Enterobacter (Richardson et al. 2009). A study showed that biological amendment carrying Burkholderia spp. induced plant phosphate accumulation under phosphate starvation (Finkel et al. 2019). Bacterial community dynamics in sugarcane crop soil revealed that amendment with organic vinasse (a mixture of organic nutrients and exogenous microbes) significantly improved soil properties, such as nutrient availability, pH, cation exchange capacity, water retention, and soil texture, which ultimately improved the microbial community structure and allowed beneficial microbes to thrive (Lourenço et al. 2018). Improving soil health by developing resistance against pathogens is another advantage associated with biofertilization. A diseased soil amended with Bacillus- and Trichoderma spp.-based biofertilizer increased the abundance of native microorganisms carrying antifungal activity, thus suppressing the Fusarium oxysporum pathogen (Xiong et al. 2017).

To date, it is not possible to completely eliminate reliance on chemical fertilization, but a balanced use of chemical and bio-organic fertilizers has the least negative impact on the rhizosphere microbiome while improving soil nutrient status (Cui et al. 2018). It has been suggested that replacing at least 25% of synthetic fertilizer consumption with bio-organic amendments would be beneficial in maintaining a relatively stable rhizosphere microbiome (Cai et al. 2017). Thanks to advances in microbial ecology, bioinformatics, and molecular technologies, we can now design multispecies biofertilizers by combining multiple strains with complementary properties that can withstand environmental stresses.

Tillage intensity

Tillage has a significant impact on the physicochemical properties of soil (Hartman et al. 2018; Mathew et al. 2012). Frequent tillage deteriorates microbiome structure by disrupting nutrient passage networks, resulting in a loss of connectivity among individual species (Goss-Souza et al. 2019; Young and Ritz 2000). Long-term no-tillage, on the other hand, can promote the distribution of stable communities (Goss-Souza et al. 2019), which can further promote the enrichment of unique microbial taxa strongly linked with key nutrient cycling processes in the rhizosphere (Smith et al. 2016). A global-scale meta-analysis identified no-till practices with crop residue retention as an important strategy for improving soil quality by increasing the microbial biomass and nutritional status of soil regardless of climatic and edaphic variables (Li et al. 2018). However, completely avoiding tillage is not recommended in fields with an elevated level of organic amendments (no pesticides/herbicides), where weeds, herbs, and pests can thrive and reduce crop yield. However, in practice, a reduced-tillage practice combined with biofertilization can result in a relatively stable functional microbiome.

Reconstructing the existing microbiome by microbial manipulations

Simplified microbial inoculations and associated challenges

In agricultural systems, inoculating beneficial microorganisms has been identified as a key strategy for manipulating the plant microbiome (Sessitsch et al. 2019). Next-generation sequencing and advanced quantitative and qualitative molecular methods such as qPCR, TRFLP, FISH, and DNA arrays have allowed scientists to develop a better understanding of the structural and functional composition of soil microbial communities (Compant et al. 2019). Such advanced technologies can be used to identify and isolate the suitable microbial targets that are the primary drivers for the associated host plant fitness (Qiu et al. 2019). Various microbial inoculation approaches exist, ranging from single-strain inoculations to multi-strain consortium inoculations, as well as transferring complex unidentified microbiomes (Tosi et al. 2020). The introduction of a single strain or a combination of two or more previously isolated and defined strains as a simple consortium is a classic approach for targeted inoculation. Successful inoculation relies on survival, establishment, and performance within a given environment, and single-strain inoculations are usually susceptible and extremely prone to inconsistency (Tosi et al. 2020). One reason is their oversimplification, which allows them to be quickly outcompeted by the indigenous microbiome following inoculation. Another reason is that transitioning from an in vitro to a field environment can temper their capabilities and produce opposing traits (Kaminsky et al. 2019).

In contrast, by integrating multiple distinct microorganisms, a large functional genetic pool can be created, providing stability against environmental challenges while reducing the burden of risks and uncertainty associated with single-strain or lower complexity inoculations. Under low-phosphorus greenhouse conditions, soil amendments with biocompatible multispecies consortia of phosphate rock-solubilizing bacteria promoted maize seedling growth by providing sufficient solubilized phosphorus (Magallon-Servin et al. 2020). However, contradictory results have also been reported when inoculating microbial consortia. For example, in herbivore control applications, multi-strain consortia attribute inferior performance to insufficient triggering of chemical changes in the host plant when compared to single-strain inoculants (Gadhave et al. 2016). This could be due to unexpected effects from interactions between the introduced microbes, the resident microbiome, and the host plant (Gadhave et al. 2018).

A less traditional to modernized approach: constructing a synthetic microbiome

Synthetic community development, which refers to the controlled design and establishment of microbial communities with specific functions, is one option for mitigating the uncertainties associated with consortia inoculation (Fig. 3). Synthetic communities, so-called SynComs, are usually constructed following a bottom-up approach (Lawson et al. 2019), which involves the identification of keystone microbial taxa, interactive network development, and the use of combinations of microbial isolates as a composite inoculum (Mueller and Sachs 2015; Vorholt et al. 2017). The reported synthetic microbiome applications in agricultural systems are summarized in Table 2.

Table 2 Applications of synthetic microbiome in agriculture
Full size table

Identification of the keystone microbial taxa

The core microbiota is comprised of members of the microbial community that are abundant and persistent in virtually all of the communities associated with a specific host and carry functional genes that are essential for host fitness (Trivedi et al. 2020). Fortunately, many common core microbiota members, such as Agrobacterium, Pseudomonas, and Methylobacterium, are observed in multiple plant species, such as rice, barley, soybean, sugarcane, and grapevine (Trivedi et al. 2020). Microbe-microbe and host-microbiome interactions are mediated by a few members of the core microbiota. Even at low relative abundance, these hub microbial members may represent keystone species that are critical in shaping the microbiome structure (Carlström et al. 2019). The identification of these keystone taxa may represent potential targets to build SynComs. However, due to the relatively lower proportion of cultivable microbes in nature, it always carries an associated risk of missing keystone functional members (Vorholt et al. 2017). Since these hub members have a strong regulatory influence on microbial interaction networks, their removal leads to a loss of interactions. For example, a study clearly demonstrated that Enterobacter was a keystone member for maize root microbiome assembly and that its loss resulted in the dissociation of all other consortia members (Niu et al. 2017).

Interactive network development and creating the SynComs

Rapid advancements in culture-independent analysis, gene editing technologies, next-generation sequencing, meta-omics, and bioinformatics tools have offered a top-down strategy for creating efficient SynComs. Briefly, next-generation sequencing is used to characterize the core microbiomes, which are then statistically analysed and fitted into network analysis models. Subsequently, by employing metagenomics, interlinking core microbial members with crucial functions  are identified, followed by isolation and application of the key players as a microbial cocktail (Mahmud et al. 2021).

For the robust microbial screening and cultivation required for the development of a SynCom involve high-throughput and axenic culturing technologies (Qu et al. 2020). Since the core microbiome carries some species that are crucial in shaping the microbiome structure, even at a very low relative abundance (Carlström et al. 2019). To reduce the risk of missing such keystone members due to a lower proportion, it is always necessary to incorporate the selection of proper cultivation media, as well as pay particular attention to less cultivable microbial members. A study used high-throughput cultivation and identification technology and successfully characterized 7943 bacteria isolated from roots and leaves of the A. thaliana plant under natural soil conditions (Bai et al. 2015). The microbial isolates were identified using two-sided barcode labelling and independent library preparation followed by Illumina high-throughput sequencing. Another group of scientists employed upgraded Illumina HiSeq technology and successfully characterized 13,512 bacterial isolates, accounting for almost 70% of rice root microbiome members (Zhang et al. 2019). Furthermore, phenotypic microbial characteristics (e.g. bacteria with aerobic or anaerobic metabolism) can be retrieved by using databases such as BugBase to find growth media types for growing the rarest taxa (Ward et al. 2017).

SynCom complexity: a strong pillar of its effectiveness

High complexity is particularly important for SynComs in terms of stability. Complex SynComs build more stable metabolic networks and can substitute functional species as the environment changes (Carlström et al. 2019; Mueller and Sachs 2015; Tsolakidou et al. 2019; Vorholt et al. 2017). Simpler SynComs, on the other hand, may be less persistent in such environments, which may be the reason why successful laboratory-scale low-complexity microbial inoculants fail when exposed to field conditions (Raaijmakers and Mazzola 2016). Complex consortia-mediated enhanced effects of SynComs have been documented, as increased diversity promotes the quantity and quality of pro-plant functions (Saleem et al. 2019; Tsolakidou et al. 2019). A study built a complex bacterial SynCom consisting of 185 members and revealed a key genus involved in inducing host plant phosphorus stress under phosphorus starvation conditions (Finkel et al. 2019).

To date, most SynComs have only been developed and evaluated in simplified ways (Compant et al. 2019; Vorholt et al. 2017), and there is insufficient experimental evidence to warrant the use of synthetic microbiomes on a broader scale in agriculture. One key point of contention is whether these SynComs should be built particularly for each unique case or if they should instead aim for broader applicability. To develop robust SynComs, some factors must be considered, such as development, accessibility, applicability, and production costs (Bell et al. 2019b). Future research should focus on the integration of novel analytical technologies, omics, and bioinformatics to develop higher-complexity SynComs that incorporate more microbe-mediated plant fitness parameters with the goal of translating them into agricultural applications.

Host plant-mediated microbiome routing

Phytochemical-exudations as potential target

Plants have been shown to have a strong association with rhizosphere microbiome assembly (Edwards et al. 2015; Lebeis et al. 2015; Marques et al. 2014). Based on physiological and genotypic traits, plants regulate their metabolism and produce a variety of chemical attractants in the form of root exudation. Root exudates include primary and secondary metabolites, such as sugars, amino acids, enzymes, organic acids, and mucilage and account for nearly 10% of the plant total sequestered carbon and approximately 15% of total plant nitrogen (Haichar et al. 2016; Jones et al. 2009; Preece and Peñuelas 2016). These compounds selectively increase microbial biomass and their metabolic activities in the rhizosphere. The amount and composition of exudates vary at different plant developmental stages (Chaparro et al. 2014) and with different plant species (Naylor et al. 2017), resulting in the selective recruitment of associated microbiome members. For example, the two rice varieties, indica and japonica, recruits significantly distinct root microbiota (Zhang et al. 2019). Another study observed that different developmental stages of cotton plants promoted specific bacterial phyla and thereby induced significant variation in the rhizosphere microbiome at different stages. They also witnessed strong intactness of microbial community composition with different cotton plant cultivars irrespective of soil type, which reflected the influence of plant genetic variation on the microbiome (Qiao et al. 2017).

One category of exudate is the defence phytohormones salicylic acid, jasmonic acid, and gaseous ethylene (Lebeis et al. 2015). These phytohormones promote the plant's ISR and restrict colonization to only potentially beneficial, nonpathogenic microbes (Naylor and Coleman-Derr 2018). A study demonstrated the contributions of phytohormones to root microbiome recruitment in the model plant A. thaliana. They grew isogenic plant mutants with altered immune systems (each mutant lacking the expression of at least one of the defence phytohormones discussed above) in wild soil and found that the recruitment of the plant’s root microbiome members was dependent on the foliar defence phytohormone salicylic acid (Lebeis et al. 2015). Modifying the chemical exudations of plants can direct the plants to steer the existing microbiome in a favourable way.

Mimicking the bacterial autoimmune signal can be another efficient phytoexudation target for manipulation. Bacterial members in the microbial community regulate their genomic expression and cooperate with each other via a signalling process called quorum sensing (QS). It involves the synthesis and exchange of small signalling molecules responsible for the activation of transcriptional regulators. N-acyl homoserine lactones (AHLs) are the most common signalling molecules released by most gram-negative plant pathogens (Gao et al. 2003). During pathogenic invasions, some plants have the unique ability to mimic the bacterial-specific AHL-QS system to manipulate bacterial autoinducer signals, which can either stimulate or disrupt bacterial sensing as a stress response (Gao et al. 2003; Liu et al. 2019; Pérez-Montaño et al. 2013). Synthetic construction of plant-based microbial signal mimicry may be a viable option because artificially stimulating bacterial sensing can improve plant-microbe interactions and may ultimately maximize the recruitment of beneficial microbiome members. This knowledge constitutes a solid foundation for developing successful plant-based microbiome engineering strategies, as given below.

Plant-based strategy for microbiome engineering:

Plants can be directed to secrete specific exudates that attract specific microbes. Various manipulation strategies have been used to alter these exudation patterns, with genetic engineering, plant breeding, and meta-organism-based manipulations being the most commonly used (Arif et al. 2020; Dries et al. 2021; Kumar and Dubey 2020). Plant cultivars, or “designer plants”, are created by incorporating breeding/engineering strategies to attract and maintain a stable and efficient microbiome and, ultimately, tolerance against abiotic stress such as drought and salinity, as well as biotic stress such as pathogens. A study compared plant cultivars with varying levels of resistance to wilt-causing diseases and revealed that some disease suppressant bacteria were more abundant in the rhizosphere of resistant cultivars (Kwak et al. 2018). Another study reported the capacity of watermelon cultivars to establish a type of rhizosphere microbiome and considerably suppressed Fusarium wilt disease in that soil (Liu et al., 2018).

The quantitative trait locus (QTL) mapping technique can be used to search the host genome for specific gene pools associated with distinct phenotypic traits. Once a suitable genes pool determined in the plant genome, advanced genome editing tools such as site-specific transcription activator-like effector nucleases (TALENs), zinc finger nucleases (ZFNs), and clustered repeatedly interspaced short palindrome repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) can be incorporated for site-directed manipulations to modify the traits for recruiting the desired rhizosphere microbiome (Kaul et al. 2021; Kumar and Dubey 2020). These tools are used to target key genes involved in biosynthetic and metabolic pathways at the transcriptional and translational levels (Fig. 3). Numerous success stories have demonstrated that the CRISPR/Cas9 system is the most advanced technology for knocking out or inserting key genes across multiple plants. A study used this technology to develop a tomato plant resistant to bacterial speck disease caused by the pathogen Pseudomonas syringae pv. tomato (Pto) DC3000 (Ortigosa et al. 2019). Furthermore, Shan et al. (2014) successfully developed a wheat cultivar resistant to powdery mildew disease by knocking out the TaMLO gene (associated with fungal pathogens colonization) found in its protoplasts. Many more achievements have been made with the CRISPR/Cas9 system as a genome editing tool, and there is still much more to come.

In short, plant breeding approaches and advanced plant genome editing-based methods are promising ways to accumulate favourable alleles that can act as a selective force in directing dynamic microbiome assembly. However, technical gaps such as varying root exudate composition, ISR, host genotype, competition by indigenous soil microflora, environmental stress conditions, and a lack of in situ manipulation tools continue to impede these manipulation strategies. Despite all, incorporating advanced omics and synthetic biology approaches for in situ microbiome engineering may aid in the establishment and success of healthier microbiomes.

Engineered microbiome adapted to abiotic and biotic stresses

Understanding the complex nature of microbiome-mediated protection will result in long-term strategies to ensure healthy management practices that take into account interactions of the microbiome with the soil, plant, and environment. Synthetic microbiome-based disease suppression is already well documented in human gut microbiota studies, where faecal microbiota transplantation (FMT) application is frequently used to alter immune responses and to suppress diseases (Nazmul Huda et al. 2020). With respect to plant health, the soil microbiome has also emerged as a key component to induce systemic resistance in plants against pathogens and insects (Pineda et al. 2017). Some studies successfully developed SynComs at laboratory and pot scales to improve plant fitness (Niu et al. 2017; Tsolakidou et al. 2019). However, for this technology to be successfully used at the field scale in a variety of soil conditions, it still has a long way to go.

In agricultural soil systems, salinity is a growing environmental concern, as over 840 M ha of arable land is already affected by salinization and predicted to increase (Rath et al. 2019b). Implementation of halo-tolerant synthetic microbiomes into saline soils could alleviate the salinity impact. Despite the fact that the majority of microbes are halo-sensitive, some halophytic plant-associated members are halo-tolerant and can be considered potential targets for synthetic microbiome development. It has been demonstrated that inoculated halotolerant rhizobacteria improve the native microbial community’s resilience to salinity stress and, as a result, improve plant growth and stability in unfavourable saline conditions (Bharti et al. 2015; Yuan et al. 2016). Engineered microbiome approaches are recommended for use in areas with saline water irrigation systems.

Drought is among the worst obstacles to agricultural productivity. Plant stress tolerance must be improved to allow for satisfying crop growth in the face of limited water resource availability under drought conditions (Liu et al. 2019). Certain root-associated bacteria have been reported to have the capability to improve plant drought stress tolerance (Ngumbi and Kloepper 2016; Rolli et al. 2015), but except for a few model plants (Xu et al. 2018; Zolla et al. 2013), their mechanism for stress alleviation is still not well documented with respect to a specific host. For example, the model plant Sorghum bicolor secretes specific metabolites, which facilitates bacterial ATP-binding cassette transporter gene expression and, in turn, modifies the root-associated microbiome composition by promoting the abundance and activity of monoderm bacteria, which has a positive impact on the growth and development of Sorghum bicolor plants facing drought stress (Xu et al. 2018). This is a potential blueprint for developing SynComs from such plant-associated microbiomes to increase crop productivity in arid areas with low precipitation and poor irrigation systems.

Concluding remarks

A number of abiotic and biotic factors influence microbial interactions with the associated environment. Single or composite effects of these factors impact the fitness and performance of rhizosphere microbiomes; in response, they behave differently and cause functional unevenness in diverse soil conditions. However, a healthy cropping system entails the presence of a healthy, diverse, and active microbiome in the rhizosphere. In fact, we advocate for the reshaping of the microbiome to improve stress tolerance and the development of smart and sustainable agricultural systems. Soil microbiome manipulation involves either isolation and application of external beneficial microbes or in situ steering of the existing microbiome. Here, we explored different microbiome manipulation strategies and proposed the term MiMSAS to explain the development of smart and sustainable agriculture systems based on reshaping the rhizosphere microbiome in cropping soil. In brief, using a nongenetic strategy, such as integrating phytobiome-based soil management methods with cultivation practices, is an efficient way to direct the existing microbiome for cropping benefits, which can eventually promote a soil ecosystem with a stable microbiome for critical functioning. We further reviewed SynCom inoculation as a tool for regulating existing microbiomes and indicated the positive correlation of increasing SynCom complexity with its success rate in field applications. However, to be successful with this approach, smart microbes with smarter delivery systems, as well as advanced approaches to monitoring the ultimate fate of the applied consortia in the soil, must be developed.

As a realistic solution to the problems associated with microbiome manipulation, host-plant-based microbiome steering is suggested as a more attractive approach. The QTL mapping technique can be used to search the host genome for specific gene pools associated with distinct phenotypic traits. Next is the advanced genome editing tool CRISPR/Cas9 system to knock out or insert key genes across multiple plants. In short, plant breeding approaches and advanced plant genome editing-based methods are promising ways to accumulate desired cultivars in variety of crops to occupy healthier microbiomes.