Accelerating soil aggregate formation: a review on microbial processes as the critical step in a post-mining rehabilitation context

Keywords: bioengineering, coal, microbial processes, soil aggregate, soil formation, soil microbiome, soil rehabilitation, spoil.

Healthy soils are essential for effective rehabilitation strategies at mine sites (Zornoza et al. 2016). However, most open cut mines will present inadequate quantity and/or quality of topsoil during rehabilitation efforts. The situation is particularly acute in Australia’s open cut coal mines where topsoil was not stored prior to mining and the medium for site rehabilitation is limited to coal mine spoil (subsoil and waste rock; Emmerton et al. 2018). There is a great need and an even greater potential to improve open cut coal mine site remediation outcomes in Australia by accelerating the transformation of mine spoils to functional soils. Microbiology is central to soil formation (Kleber et al. 2015); therefore, harnessing microbial processes represents an opportunity to transform coal mine spoils to stable, fertile soils and improve rehabilitation outcomes in topsoil deficit situations. The aim of this review is to evaluate the potential for development and application of microbial biotechnologies to accelerate soil formation from coal mine spoils in Australia.

Coal is a combustible rock of organic origin composed mainly of carbon associated with several other elements in variable quantities, mostly hydrogen, sulfur, oxygen and nitrogen (Voncken 2020). It is a sedimentary rock formed from accumulated plant matter that has been altered by decay and by high temperature and pressure variations over millions of years. Interlayered with other sedimentary rocks, it forms beds ranging from less than a millimetre to many metres thick (Voncken 2020). The high carbon content of coal makes it a useful resource for heat, electricity generation and for industrial processes such as metal refining. Australia is the world’s fourth largest producer of coal and coal production represents a significant proportion of Australia’s annual income, with a total revenue of more than AUD87 billion in 2019.

The principal black coal producing basins in Australia are the Bowen (Queensland) and Sydney (New South Wales) Basins, and the majority of mines in these areas are open cut (Scott et al. 2010). Open cut mining or open pit mining is a surface mining technique of extracting rock or minerals from the earth by their removal from an open pit or burrow (Scott et al. 2010; Li et al. 2014). It is used when the minerals are found over a large area and relatively close to the surface (Li et al. 2014). The open cut mine is dug downwards in benches or steps, which slope towards the centre of the pit. To uncover the coal, the land is first cleared of vegetation and the topsoil is removed and ideally, used immediately to remediate another site versus being stockpiled for later use in the rehabilitation process (Feng et al. 2019). Open cut mining severely disrupts landforms and soils, preventing or impeding the restoration of pre-existing or functional ecosystems, because the essential properties of the original soils cannot immediately or easily be reinstated (Ngugi et al. 2018). During topsoil processing and storage, soils are subjected to intense disturbance and exposure to high temperatures and desiccation. These stresses damage the soil’s physicochemical and biological characteristics including decreases in microbial activity, organic matter content and pH (Singh et al. 2004; Ngugi et al. 2018; Williams et al. 2019). In some cases, topsoil has not been stored at all, and the underlying mine spoil becomes the substrate for rehabilitation.

The chemical properties of mine spoils can prevent seed germination and the extremely poor water-holding capacity can limit seedling development (Sarkar et al. 2017). These wastes are also very prone to erosion by wind, and water (Ulusay et al. 1995; Dawson et al. 1998). Finally, the fine-grained fraction of the waste can form soil crusts and cracks, which also adversely influence the soil structure (Hossner and Hons 1992; Sheoran et al. 2010). In Australia’s coal producing regions, mine spoils are often used as revegetation substrates even though they are usually stony, have severe chemical deficiencies, are lacking in soil structure, and are prone to erosion (Luna et al. 2016; Emmerton et al. 2018). Over decadal timescales in Australia, topsoil was not stored at all during mining operations and was simply buried, causing a severe shortage during remediation (Emmerton 2019). The lack of topsoil is the key reason why mines are forced to use spoils as revegetation substrates in the Australian open cut mining industry.

Open pit coal mining can cause severe damages to microbial communities and alter the nutritional status of the soil in the mined area through excessive leaching and stockpiling (Ram and Masto 2010). In some cases, the coal spoil abiotic conditions are vastly different from the pre-degradation baseline (unmined topsoil), and the system may not follow the expected rehabilitation projection (Barliza et al. 2018). Coal mine spoil or overburden originates from consolidated and unconsolidated materials that cover a coal seam, and its properties may vary significantly over the globe due to variation in soil texture, mineral composition and stock management (Li et al. 2014). Spoils produced during coal mining are typically nutritionally limited, weathered, sodic, and dispersive, with an acidic or alkaline condition, high salinity, and poor combinations of clay, silt, and fine sand with low coarse sand and fragmental content, which contribute to poor levels of plant growth, low infiltration rates, and high levels of erosion (Madsen and Mulligan 2006; Emmerton et al. 2018).

Amongst the main nutrients, nitrogen was found to be the most plant growth-limiting element in coal mine spoil, followed by phosphorus, suggesting typical early successional systems (Verhoeven et al. 1996; Nussbaumer et al. 2016). Ganjegunte et al. (2009) reported that more than 65% of nitrogen can be rapidly lost from coal mine spoil compared to the previous topsoil, whereas Akala and Lal (2001) showed the same trend for soil organic carbon. In areas where the initial pH of spoil > pH 8.5, it could also reduce the availability of P, B, Cu, Zn, Fe, and Mn (Lucas and Davis 1961). This available nutrient depletion can affect plant density by limiting germination, seedling establishment and long-term survival (Nussbaumer et al. 2016; Yuan et al. 2020). The re-establishment of these biogeochemical processes, particularly of the nitrogen, carbon and phosphorus cycle, requires additional human intervention that lead to a self-sustaining ecosystem (Barliza et al. 2018).

Among all the conditions mentioned above, salinity has been registered as a common issue of coal spoils globally, being a major concern related to coal mine impact on soil rehabilitation (Szczepanska and Twardowska 1999; Madsen and Mulligan 2006; Gozzard et al. 2009; Li et al. 2014). For example, Madsen and Mulligan (2006), registered a highly sodic condition on coal spoil from Central Queensland, Australia, with electrical conductivity (EC) reaching 25 mS/cm, which is half-seawater (Park et al. 2013). Spoil in this region is outstandingly sodic compared with standard soils, limiting plant growth and being an obstacle to coal spoil bioremediation and rehabilitation strategies (Zhu 2001; Madsen and Mulligan 2006). In the Bowen Basin, sodic spoils typically contain an excess of sodium relative to calcium and other essentials nutrients (Qadir and Schubert 2002), affecting an even more important soil physical–chemical property, aggregate formation, resulting in weak aggregate stability and spontaneous dispersion of clay particles on contact with water, increasing the risk of erosion and, ultimately, slope failure (Henderson 2004; Vacher et al. 2004; Howard et al. 2011).

According to Li et al. (2014), the coal mining industry in Australia is urgently seeking information to understand the source, dynamics and management options of the spoil salinity, which has been identified as a challenge for plant growth, preventing successful responses to spoil amelioration (Li et al. 2014). In contrast, Salazar et al. (2009) found that reclamation and revegetation showed a positive effect in reducing coal spoil salinity of overburden dump soils in Spain, in opposition to other studies where decreasing spoil salinity after rehabilitation did not occur (Juwarkar and Jambhulkar 2008; Juwarkar et al. 2009).

Management of coal mine spoil salinity is still a vague area, and these studies show that much remains to be understood in order to have reliable information to determine best practices for spoil salinity amelioration (Li et al. 2014). Soil mineralogy, soil texture and climate are considered the main factors controlling salt movement in soils (Ben-Hur et al. 2009; Pankova et al. 2010; Li et al. 2014). Understanding the origin and dynamics of salts in coal spoil in Australia will also enable the definition of cohesive strategies for spoil amelioration and soil rehabilitation, where salinity is a major concern.

Current strategies for rehabilitating coal mines with sodic spoils include amelioration of excessive exchangeable sodium; the maintenance of sufficient electrolytes in the soil solution; and bonding of clay particles by organic matter (Sarkar et al. 2017).

In agriculture, the most commonly used method to ameliorate sodic soil is via the application of calcium in the form of gypsum (CaSO₄·2H₂O). The incorporation of gypsum into sodic soils addresses both primary pathways with an initial electrolyte effect (due to gypsum being a salt) and the progressive displacement of Na via the cation exchange complex over the longer-term due to the preference for the divalent cations, such as Ca²⁺ to bind with clay (Geeves et al. 2007).

The addition of organic matter as an ameliorant for sodic soils is also well described with a generally agreed effect on soil structure through the binding of soil particles into aggregates (Nelson and Oades 1998). However, while most studies report positive effects (e.g. Tisdall and Oades 1982; Clark et al. 2009; Rani and Khetarpaul 2009; Ghosh et al. 2010), some report increased dispersion through an increase in negative charge and complexing of Ca²⁺ (Aylmore and Sills 1982; Gupta et al. 1984; Sumner 1993). McL Bennett et al. (2015) found that the addition of organic matter in combination with gypsum extended the effects of the gypsum compared to gypsum alone, although this effect did not persist beyond 2 years.

Due to the nutrient deficiencies in coal mine spoil, topsoil addition, when available, supplying soil nutrients, organic matter, microbial populations and a seedbank of native species is necessary (So et al. 2022). So et al. (2022) also address the requirement of organic and inorganic fertiliser addition, for the establishment of nutrient cycling in the remediated ecosystem. In Queensland, Australia, the Department of Environment and Science (DES) Environmental Protection Act recently included on its Estimated Rehabilitation Costs (ERC) guideline a mention for the use of fertilisers as well as gypsum as essential practices in rehabilitation of coal mines under topsoil deficit conditions (Queensland Government Department of Environment and Science 2022). It is now approved under the cost of rehabilitation, the use of 0.2–0.25 t/ha of fertilisers in coal mine rehabilitation projects (Queensland Government Department of Environment and Science 2022). This application rate is based on a post-mining land use of grazing; however, different application rates should be studied depending on site specific soil and spoil quality and the requirements of the proposed post-mining land use (Queensland Government Department of Environment and Science 2022).

These inorganic fertilisers are usually applied to overcome inefficient vegetation cover (Bell 2001; Daws et al. 2013). However, depending on the spoil properties, ions released from inorganic fertiliser can be rapidly leached (Wilden et al. 2001) or bound, making them unavailable to plants. Thus, requiring a regular application of fertiliser to maintain a primary nutrient source (Nussbaumer et al. 2016).

A critical aspect that has been overlooked in traditional approaches to site remediation and spoil amelioration is the soil microbiota and their impact on soil chemistry. Traditionally, restoration is usually approached from the perspective of plant cover regeneration, ignoring plant–soil interactions and their consequences for plant growth (Rivera et al. 2014).

Recent studies point to the ability of microorganisms to assist in the formation or bioengineering of soils (Graf et al. 2015; Vergani and Graf 2016). Verboom and Pate (2006) define ‘bioengineering’ as any one or more of a range of biologically driven influences modifying soil profile characteristics. According to these researchers, the bioengineering principles may well predominate, particularly in relation to formative effects that major plant taxa and relevant microorganisms have on uptake, transfer and depletion or concentration of specific nutrient elements between and within soil compartments (Verboom and Pate 2006). It has been shown that soil bioengineering is an important means of restoring damaged ecosystems and follows ecological principles, investigating, designing and recreating vegetation–soil systems, enhancing soil shear strength and limiting soil particle movement on slopes by utilising the effects of rhizosphere systems on soil structure (Li et al. 2006).

The rehabilitation of mine spoils requires an understanding of both soil chemistry and microbiological activity and could be improved by bioengineering approaches (Fig. 1). In the following sections, this review will focus on the key role that soil fungi and bacteria play in soil physicochemical properties, which permits the re-establishment of soil quality after mining and consequently, the conditions conducive to successful establishment of diverse ecosystems (Table 1).

Fig. 1.  (a) Soil bioengineering explained as an interconnected system; soil microbiology influencing soil structure and chemistry to promote the soil formation, and (b) practical benefits of a microbial inoculant to bioengineer the soil, promoting soil aggregate stability through the enhancement of essential soil chemistry properties (Created with BioRender.com).

Table 1.  Opportunities for microbial processes to address biophysical limitations of typical coal mine spoils.

Recent findings show that microbes can engineer soil aggregates by two mechanisms: (1) enmeshing as a result of fungal hyphae; and/or (2) production of extracellular polymeric substances (Blankinship et al. 2016). Fungal hyphae and plant roots aid in the development and stabilisation of soil aggregates by enmeshing themselves throughout the soil (Blankinship et al. 2016). The formation of stable microaggregates via hyphae helps compartmentalise soil particles and promotes microbial niche formation (Oades 1993; de Boer et al. 2005; Deveau et al. 2018). Furthermore, Rillig and Mummey (2006), in their review on mycorrhizas and soil structure, suggested the existence of mechanisms by which arbuscular mycorrhizal fungi (AMF) (i.e. type of mycorrhiza in which the symbiont fungus penetrates the cortical cells of the roots of a vascular plant forming arbuscules) can also improve soil aggregation. The extensive extraradical hyphal growth of AMF has been described as the key property for stabilising soil aggregates (e.g. Wilson et al. 2009; Barto et al. 2010; Luna et al. 2016). The extent to which soil aggregation is promoted depends on the AMF species involved, because members of different families can differentially produce soil hyphae (Hart and Reader 2002). Nonetheless, more research is required to clarify the role of species diversity for soil structure since hyphal proliferation also depends on other environmental factors such as soil pH, water content and nutrient levels (Parniske 2008; Helgason and Fitter 2009; Pietikäinen et al. 2009).

Secondly, bacteria and fungi produce extracellular polymeric substances (EPS), exudates, secondary metabolites and organic inputs that act as glues to connect soil aggregates as a possible strategy for accessing distant resources and creating a more stable environment in which to live (Tisdall et al. 2012; Smith et al. 2014; Wang et al. 2014). EPS and other biopolymers can be very sensitive to hydration, which reflects in the soil aggregate behaviour (Billings et al. 2015; Flemming et al. 2016). Hence, the EPS matrix plays a fundamental role, both structurally and physiologically, in the soil–microbe complex. EPS-induced changes in hydrological conditions protect cells against desiccation, facilitate nutrient diffusion under dry conditions and influence soil water content at the macroscale, while the EPS matrix holds particles and cells together and promotes soil aggregation (Tecon and Or 2017).

For several decades, the rehabilitation of mining areas has typically focused on improving the soil quality required for successful ecosystem re-establishment. In these assessments, aggregate stability is the widely used indicator for evaluating soil physical quality and susceptibility to erosion (Luna et al. 2016). Knowledge about the role of fungal hyphae and bacterial EPS in the formation of stable aggregates and, consequently, in the formation of a soil with a reduced risk of erosion and higher quality of nutrients, may prove to be of vital importance for the sustainable management of post-mining areas.

The ability of microorganisms to survive in extreme environments can be also exploited for amelioration of the post-mining salt affected soils. Despite all the benefits mentioned on the previous section, microorganisms cannot metabolically alter the sodic/saline nature that contributes to the harsh chemical nature of the spoil that reduces aggregate stability. However, salt-tolerant microorganisms can grow in these hostile environments and improve the cycling of essential nutrients and soil structure, favouring the establishment of plants that also have this capacity. Ashraf et al. (2004) also suggested that using an EPS-producing bacterial inoculum could serve as a useful tool for alleviating salinity stress in salt-sensitive plants, through the restriction of Na⁺ influx into roots. Geddie and Sutherland (1993) also showed the potential of bacterial EPS to bind cations including Na⁺, suggesting that increasing the population density of EPS-producing bacteria in the rhizosphere would decrease the content of Na⁺ available for plant uptake, relieving the stress caused by the saline environment. However, the relationship between EPS-producing bacteria and Na⁺ uptake needs further work with a range of salt-sensitive plant species (Ashraf et al. 2004).

The diversity of microbial communities found in soil clearly benefits the cycling of soil organic matter (SOM) in this environment. For example, fungi and bacteria possess different pathways to decompose organic matter (Heuer et al. 1997; Bailey et al. 2002; Ferris et al. 2004). Fungi also have a higher affinity to decompose SOM with high cellulose and lignin content and high C:N ratio through fungal-dominated ‘slow’ pathways, while moist, N-rich tissues are mainly decomposed via bacteria-dominated ‘fast’ pathways (Ferris et al. 2004). Moreover, soil C fractions produced by fungi are more stable relative to those mediated by bacteria (Bailey et al. 2002). Due to their central role in soil carbon cycling, filamentous fungi are becoming a prominent microbial group for soil structure formation studies (Oades 1993).

It is well known that microorganisms can mediate N and P cycling in soils (Aislabie and Deslippe 2013; Pajares and Bohannan 2016; Tian et al. 2021). The use of bio-inoculants containing beneficial soil microbes over chemical fertilisers is becoming more popular among agricultural practices, especially for enhancing crop productivity (Bahadur et al. 2016; Mącik et al. 2020; Fatima et al. 2022). In this group of microorganisms, cyanobacteria could also play an important role in the improvement of agriculture productivity as well as in ecological restoration of degraded lands (Singh 2014; Singh et al. 2016). By fulfilling their own nitrogen requirements through nitrogen (N₂)-fixation, cyanobacteria have the potential to not only promote crop growth, but also improve the soil nutrient status (Singh et al. 2016). Shukia et al. (2008) address the potential of cyanobacteria application in soil, which includes economic benefits, nutrient cycling improvement, N₂-fixation, bioavailability of phosphorus, mineralisation of organic molecules, improvement of the physicochemical conditions of the soil and restoration of soil fertility through reclamation.

Williams et al. (2019) also demonstrated that cyanobacteria are central to soil microprocesses, providing a strong foundation for the re-establishment of soil functions during mine rehabilitation. In that scenario, cyanobacteria photosynthesis drove the productivity and their growth as a biocrust was the starting point for C and N cycling, resulting in a considerable increase of soil nutrient concentration available for plants (Williams et al. 2019).

Phosphate solubilising microorganisms (PSMs) have also emerged, with increasing acceptance, as environmentally friendly fertilisers that improve the geochemical cycling of P (Tian et al. 2021). Bi et al. (2019a) showed that the combined inoculation of AMF and P solubilising bacteria (PSB) significantly increased available P, on aboveground and underground biomass. The same study shows that this combination had also a significant positive impact on soil acid phosphatase activities. Previous studied found that AMF alleviate root damage stress induced by simulated coal mining poor aggregate stability (Bi et al. 2019b). However, few studies focused on the effects of AMF and PSB on coal spoil. Despite the wide range of important biogeochemistry soil functions, such as mineralisation, solubilisation, desorption, dissolution, and weathering, large scale in situ field tests with these microorganisms are still urgently required (Tian et al. 2021).

The soil microbial ecosystem is complex, and all members need each other, functioning in multiple metabolic feedback processes. For example, the soil microbial biomass rivals the aboveground biomass of plants or animals, with soil often containing >1000 kg of microbial biomass carbon per hectare (Fierer et al. 2009; Serna-Chavez et al. 2013). These soil microorganisms have crucial roles in nutrient cycling and the maintenance of soil fertility (Fierer 2017). Given the significant role fungi and bacteria play in aggregate formation and carbon sequestration, which are critical in spoil stabilisation and spoil quality improvement, a multiproxy approach is necessary to identify and understand soil microorganisms as potential drivers for accelerated spoil to soil transformation.

To be able to harness the microbial processes already mentioned, it is important to be able to monitor and evaluate microbial processes in spoil. Traditional approaches have focused on cultivation-based studies, but these are limited by our inability to culture many microbes (Fierer 2017; Lewis et al. 2021). Many of these microbial processes are catalysed by exoenzymes produced by metabolically active fungi and bacteria present in soil (Tabatabai 1994; Veres et al. 2015; Cenini et al. 2016). Understanding the enzymatic activity patterns related to soil fractions with different physicochemical characteristics is of vital importance for the management of soil quality restoration practices.

As a result, several studies have proposed or demonstrated the usefulness of soil enzymatic activity assessments in determining their ‘health’ and quality (Dick 1994; Gregorich et al. 1994; Dick et al. 1996; Garcia et al. 2002). Despite these benefits, enzyme activity assays do not allow identification of the microbial species involved in these processes (Nannipieri et al. 2003). However, recent advances in molecular approaches open up a lot of opportunities for quantitation (Fierer et al. 2005; Noyce et al. 2015; Beule et al. 2019; Xiong et al. 2019) and comparative analyses at both a diversity level (Theron and Cloete 2000; Nannipieri et al. 2003; Malik et al. 2008; Sánchez 2017; Khan et al. 2018; Kumar et al. 2019) and a functional level (Bastida et al. 2014; Keiblinger et al. 2016; Sánchez 2017; Mandalakis et al. 2018; van der Heyde et al. 2020). Due to functional redundancy, changes in microbial composition might not alter soil ecosystem function (Souza et al. 2015). Soil processes have been related to absolute numbers of taxa, genes or gene products rather than relative percentages in a community, suggesting the use of alternative approaches (including quantitative PCR-based methods and proteomics) to follow up the absence or presence of these processes in soil (Fierer 2017). Therefore, the assessment of the actual expression and activity of functional proteins is of paramount importance for understanding the soil ecosystem functioning (Prosser 2015; Delgado-Baquerizo et al. 2016). However, the identification of functional proteins involved in the metabolic processes of soil ecosystem still presents many challenges for microbial ecologists, especially because of the difficulty imposed by the strong binding of these proteins with chemical compounds in the soil, which interfere on the identification of peptides, such as humic substances (Bastida et al. 2014; Keiblinger et al. 2016).

The soil microbiome is extremely variable across the whole soil environment. Fierer (2017) attributed this spatial variability to soil’s heterogeneous environment and the specific characteristics of the sampling site, pointing out that the choice of the experimental methods used is crucial to predict correctly this variability, since there are many factors directly or indirectly influencing the spatial structure of soil microbial communities (Fierer 2017). In addition, soil sampling is usually destructive, making it impossible to sample in the exact same location again. Thus, sampling of adjacent areas may generate ambiguous results simply due to the spatial heterogeneity in soil microbial communities (O’Brien et al. 2016).

It is also extremely important to evaluate the degree of temporal variability in soil. Understanding the temporal variability may help to describe spatial patterns and address, on a timescale, how soil microbiomes respond to environmental changes (Fierer 2017). Lauber et al. (2013) showed that temporal analyses of soil microbial communities can provide key information about factors influencing soil microbial diversity and the environmental niches inhabited by the large proportion of soil microbes that were not yet described. According to Frostegård et al. (2011), the best chance to address the temporal variability in soil microbial communities will depend on the experimental methods performed and the sampling design. Temporal pattern determination in soil microbial communities will rely on analyses of a relatively large number of samples, which will allow the discrimination between temporal changes and changes related to the inherently high level of spatial heterogeneity observed in soil (Lauber et al. 2013). In this scenario, high throughput sequencing-based techniques will be crucial to detail microbial communities’ patterns across a relatively large number of samples (Knight et al. 2012; Gellie et al. 2017), making it possible to describe the temporal variability in soil microbial communities and compare its effect with spatial variability.

Although there is evidence of successful attempts to manage the soil microbiome for several purposes (Kinkel et al. 2011; Chiquoine et al. 2016; Wood et al. 2016), it is still difficult to establish a best practice and an ‘ideal inoculum’ without fully understanding all the variabilities involved in soil biogeochemical process. Understanding soil/spoil microbiology will be the key to identify relevant microbial interventions that will improve spoil characteristics.

Evaluation of rehabilitation success, soil aggregate formation, relies on measuring the occurrence and distribution of the soil microbial community, which is regulated by interactions between carbon and nutrient availabilities. According to Sheoran et al. (2010), reclamation success also measures the structure and functioning of mycorrhizal symbiosis and various enzymatic activities in soil. However, most reclamation projects currently do not assess the microbiological conditions of the stored topsoil/spoil.

Since soil microbial communities are involved in a wide range of functions and are key agents in the rhizosphere system, they represent essential targets for restoration monitoring (van der Heyde et al. 2020). Their high sensitivity to variations of soil chemistry (Šmejkalová et al. 2003; Leff et al. 2015), soil structure disturbance (Kabiri et al. 2016; Dong et al. 2017), and plant establishment (Burns et al. 2015), also make them good bioindicators of the success of the rehabilitation process (van der Heyde et al. 2020).

Recent studies have been advocating for the inclusion of functional and composition analyses of microbial communities to not only improve our understanding of the microbial responses to restoration, but also improve the efficacy of restoration interventions (Gellie et al. 2017; van der Heyde et al. 2020). In addition, van der Heyde et al. (2020) stressed the importance of using non-sequencing analyses, such as enzyme activity assays and plant bioassays to validate high-throughput sequencing results.

While previous research has examined the physical and chemical benefits of various organic matter amendments, little is known about the relative benefits of microbial inoculants. The potential of microorganisms to be early indicators of changes in ecosystems, allows them to be used as an early warning of the likelihood of success in rehabilitation. In this way, they prove to be important tools for documenting evidence of microbial changes in soils as part of the environmental assessments necessary for land rehabilitation after mining (Wildman 2009). However, within an industrial context, there is insufficient understanding of the complex microbial interactions as well as rhizosphere environments (Burns et al. 2013). These limitations together with limited research on different aspects of soil extracellular enzymes makes it difficult to convert the knowledge acquired thus far into applied viable biotechnologies (Burns et al. 2013; Sekhohola 2015).

Given the critical importance of microorganisms in promoting soil quality and our ability to monitor soil microbiology with great precision, are there opportunities for targeted microbial intervention to improve the characteristics of coal mine spoil and improve rehabilitation outcomes? The answer to this question is centred on the capacity that this understanding will have in enabling the use of microbial inocula as tools for soil bioengineering.

Conventional rehabilitation techniques commonly focus only on aboveground ecosystems and the stabilisation of waste piles. However, long-term sustainability of ecosystems on coal mine spoil dumps requires an integrated scientific approach, which includes arrangement of spoil with organic waste and inoculation with microbial biofertilisers to achieve the restoration of fertility of these dumps (Juwarkar and Jambhulkar 2008). Selective inoculation of microbes at such sites has been shown to improve soil fertility, increasing the survivability, growth and biomass of plants (Burns et al. 2015). Microbial inoculation has also shown value for improving plant tolerance of saline conditions in agricultural situations and much could be gleaned from these studies, for application in saline, sodic coal mine spoils (Ashraf et al. 2004). The most direct way to alter the soil microbiome is through inoculation. Products containing one or several species of bacteria or fungi have been commercially available for decades (Calvo et al. 2014). However, most of these species have been isolated under traditional culturing conditions that do not mimic the chemical environment of the soil.

Reductionist approaches (one-component-at-the-time) have dominated bioremediation research. However, more fundamental research is needed to better understand the interactions among soil, plants and the microbiome as a whole, before scientifically sound interventions can be carried out in the field (de Lorenzo 2008). This is because the application of techniques developed under controlled conditions in the laboratory and in greenhouses versus the highly complex and heterogeneous field conditions have proven to be quite complex potentially producing adverse results, lacking consistency, which restricts microbiota applicability (Khan et al. 2013; Sessitsch et al. 2013; Sessitsch et al. 2019). In part, this is due to the wide range of conditions experienced by microbes in the field. Not only do key attributes such as pH, nutrient stoichiometry and texture differ among soils, but the climate regime experienced by microbes in the field covers a wide range of environmental conditions (Wallenstein 2017). These conditions must overlap with the multidimensional niche of any inoculated microbes so that they have a chance to survive, reproduce and function. According to Shi et al. (2016), an effective inoculant must form associations with the rest of the microbiome, emulating heavily structured networks in the native rhizosphere soils (Shi et al. 2016). That is why the careful selection of appropriate genera is essentially useful for application in the field; even the basic conditions experienced during fieldwork can be extremely valuable for directing future research about the use of these remediation methods applied at larger scales (Wildman 2009).

The most recent research proposes that the application of microbial inoculants in the field requires consideration of several aspects, from the appropriate design of the formulation to new concepts based on understanding the complexity and ecological behaviour of the natural microbiota (Sessitsch et al. 2019). Generally, microorganisms are produced in rich media, without considering the conditions that prevail in the soil environment after mining, which generally has low levels of nutrients. Physiological adaptation can reduce the competitive capacity of an introduced strain and limit its establishment. Therefore, field application of microorganisms requires the consideration of several aspects, ranging from the design of the appropriate formulation to allow the survival and shelf life of microorganisms, to new concepts based on understanding the complexity and ecological behaviour of natural microbial communities.

Successful long-term establishment of microbial inoculants, as well as expressing the relevant soil health-promoting effects, are also key issues to be considered. Taking into account the large number and diversity of microorganisms in the soil, where the inoculant must be established, it is necessary to apply an appropriate number of active cells. This requires the presence of microorganisms within suitable formulations, which should protect microbial cells from desiccation and other adverse conditions (Sessitsch et al. 2019).

Recent studies highlight certain doubts about the widespread use of soil microbial inoculants mainly because the ecological understanding is not sufficiently integrated into microbial selection and production and also because of the difficulties assessing in-field success (Owen et al. 2015; O’Callaghan 2016; Hawkes and Connor 2017; Hart et al. 2018; Ryan and Graham 2018). At the same time, there is an increase in demand for effective inoculants with agriculturally-relevant traits (e.g. nutrient mineralisation), as food needs will increase for the coming decades (Godfray et al. 2010), as well as the concern about the reduction in the use of synthetic fertilisers and pesticides (Keeler et al. 2016; Milner and Boyd 2017). The importance of ecology in inoculant survival has been reported (Dejonghe et al. 2001; Thompson et al. 2005; Verbruggen et al. 2013); however, researchers continue to focus on targeted functional traits rather than establishment/survival traits, while the academic literature suggests that such an integrated view is still not sufficiently considered (Kaminsky et al. 2019).

Commercially useful microbes can be rapidly characterised and identified with new omics tools, demonstrating the feasibility of using microbial consortia as biofertilisers and bioremediators, energising this industrial sector (Thavamani et al. 2017; Kaminsky et al. 2019). However, according to Kaminsky et al. (2019), the biggest challenge in developing effective inoculants is the inherent difficulties created by the long and varied processes necessary to develop an efficient and safe product formulation. To achieve this goal, inoculants must go through at least five essential stages: (1) capture and refinement; (2) production; (3) establishment; (4) function; and (5) environmental impact assessment, which demand different microbial traits (Kaminsky et al. 2019).

Prior to the initiative to solve the key problems in the field (e.g. nutrient solubilisation), the inoculant must be cultivated, tested and mass produced (e.g. in a liquid bioreactor). In addition, microbial inoculants must survive formulation and storage, and adapt to diverse environmental conditions, which probably requires greater replication and growth rates in the environment (Kaminsky et al. 2019). Those with rapid growth rates, particularly those suitable for mass production, can replicate several times after field application, with products forced to adapt to the varied biotic and abiotic conditions at each application site (Kaminsky et al. 2019). Therefore, the feasibility of producing an ideal microbial inoculant is perhaps the biggest challenge in this sector because we still have little information about the ideal microorganisms for a general product that will be applied in the field. The most successful tests until now involve the capture of microorganisms from the environment where the product will be applied and their possible modification through adaptations to the growth conditions in the laboratory or in the field. This approach has led to microbial products from new companies that can quickly reach the market (Waltz 2017).

With respect to plant growth promotion, several studies in different systems have demonstrated the positive effect of using consortia (either multiple bacteria or combinations of bacteria and fungi) as biofertilisers instead of a single strain (Artursson et al. 2006; Sánchez et al. 2014; Egamberdieva et al. 2016). These results can be explained by the fact that the consortia may present a metabolic redundancy, which promotes the ‘metabolic division of labour’, decreasing the demand for energy to complete certain metabolic pathways in the soil (Tsoi et al. 2018). Elevated species richness and diversity also produces high functional redundancy within the soil microbiome, allowing it to quickly recover during stress (Yin et al. 2000; Nannipieri et al. 2003; Chaparro et al. 2012). However, the reasons for consortium benefits are not always well known, but complementary functional mechanisms and/or greater chance of environmental colonisation would likely result in more reliable functional outcomes, although the development of synergistic consortia under a wide range of environmental conditions is far more complicated in terms of screening and production (Kaminsky et al. 2019).

The use of microbial consortia that act synergistically, occupy different niches, co-exist and do not show potential negative interactions, have been shown to be more efficient in some systems (Mikesková et al. 2012; Thijs et al. 2014; Teng et al. 2015). In addition, the use of mycorrhizal helper bacteria, such as a biodegradative bacteria that promote rapid colonisation by ecto- and endo-mycorrhizal fungi might have a synergistic effect (enhancing nutrient mobilisation from soil minerals, regulating water holding capacity, contaminant degradation, and protection of plants against root pathogens (Tarkka and Frey-Klett 2008; Thijs et al. 2017).

The use of consortia can also help to compensate for characteristics that are not ideal for early product survival; e.g. improving desiccation tolerance in the formulation (Berninger et al. 2018) or facilitating the establishment of a target strain. However, unsuccessful microbial applications may modify soil receptiveness to further inocula (Mallon et al. 2018), suggesting that pioneer strains could hinder the establishment of target microorganisms. The in situ integration of several microorganisms with different metabolic functions (e.g. rhizobia and mycorrhizal fungi) can meet the demand for soil nutrients (van der Heijden et al. 2016) or at least, promote the identification of microorganisms that additively enhance a single function (Wallenstein and Bell 2018). It is also important to use indigenous microbial strains that are best adapted to actual soil and climatic conditions to produce site-specific inoculum (Mummey et al. 2002; Khan 2004).

Successful inoculants will need to thrive over the long term and under varying soil conditions, which presents a challenge. Changes in approaches to screening (e.g. co-selection for survival and function) and new approaches to commercial production and application of many soil microorganisms that are adapted to site-specific conditions rather than nutrient-rich media should help in the development of such products (Kaminsky et al. 2019).

According to Wildman (2009), monitoring microbial and in particular, filamentous microbial status in soil stockpiles can assist in determining new approaches to soil storage and management. Microbial survival and their reintroduction during soil rehabilitation programmes can be improved with a reliable monitoring, increasing the likelihood of success through the more rapid re-establishment of soil processes that assist revegetation and plant succession. The effects of these approaches on microbial communities can be monitored using the methods outlined above. Wildman (2009) also recommended the development of an integrated database from these analyses and additional microbial research. This database would link all information on microbial numbers, bacterial and fungal microbial community data, site details, soil physicochemical information and environmental data, developing a predictive tool that would be useful to improve the efficiency of soil handling, storage and re-application on mine sites undergoing rehabilitation.

Post-mining, there are certain ecological factors such as soil–microbe and plant–microbe interaction systems that play a major role in improving the landscape (Thavamani et al. 2017). Thus, engineering those ecological inputs (i.e. introduction of specific microbial consortia) will be the key for sustainable rehabilitated mine lands in future. However, limited reports are available on these interactions on mined lands (Thavamani et al. 2017). Thereafter, complex investigations on identification of microbial communities thriving in mine lands would be of great importance in this direction. According to Thavamani et al. (2017), future research in mine site microbiology must focus on engineering the rhizosphere and non-rhizosphere microbiomes at mine sites to harness the full potential of microbiota in the rehabilitation of unused/abandoned mine sites. A better understanding on microbial diversity in mine sites can pave the way for alternate reclamation approaches like the use of biofertilisers (Thavamani et al. 2017).

Wallenstein (2017) pointed out that in the coming decades, we will be purposefully engineering the rhizosphere with increasing sophistication. This indicates that successful rhizosphere engineering will require a systematic approach, encompassing as many variables as possible. Engineering the interactive rhizosphere can enhance the efficiency and sustainability of crop production only by emulating the symbiotic interactions between plants, soil, and microbes that may have evolved over millions of years in nature (Wallenstein 2017).

A better understanding of these interactions, the establishment of microorganisms in the soil environment, their ecology and the ability to perform the necessary functions in specific agricultural environments need to be addressed to allow application in the field (Mitter et al. 2019). The selection of microbial inoculants or combinations based on their integration/position in microbial networks, rather than primarily functional characteristics, will identify microbial inoculants that can reliably establish in the field due to more stable interactions with the soil-resident microbiota (Compant et al. 2019). Finally, a constant transfer of knowledge from soil microbiome research, considering field conditions, could boost innovation in microbiome applications for soil recovery and help bridge the gap between laboratory results and field performance (Sergaki et al. 2018).

Post-mining rehabilitation is underpinned by effective and healthy soils. Given the critical role that microorganisms play in carbon, nitrogen and phosphorus cycling in soils, there is potential for their application to improve the quality of nutrient-deficient mine spoils. Furthermore, microorganisms play a key role in promoting soil stability and aggregation and in stimulating plant survival through rhizosphere interactions. These functions are critical with regard to rehabilitation of saline, sodic dispersive mine spoils that are prone to erosion. As such, there is both an urgent need for, and an opportunity to harness, microbial processes that promote soil aggregate formation, in rehabilitation of open cut coal mines.

Understanding and monitoring soil microbiology during rehabilitation will shed light on how to promote the microbial processes that accelerate the spoil to soil transformation and reveal ways to use microorganisms in soil reclamation projects. Monitoring using a multiproxy approach provides reliable evidence of microbial community changes in spoils and soils and could also form part of the environmental assessments that are required for the rehabilitation of mined lands. The development of an integrated database of information from these geomicrobiological analyses would be useful in the development of suitable microbial inoculants that will improve spoil quality and improve post-mining rehabilitation processes. This review provides a reliable and valuable compilation of data that must be taken into account prior to, during, and after the application of microbial inoculants on mine spoil as part of mine rehabilitation efforts.

Data sharing is not applicable as no new data were generated or analysed during this study.

The authors declare no conflicts of interest.

We acknowledge support from The Australian Coal Industry’s Research Program (ACARP Project C28035). Guilherme O. A. da Silva acknowledges the support from the Australian Government Research Training Program.