Over the past decades, anthropogenic alteration of soils by the increased use of synthetic fertilizers, pesticides and land conversion in order to increase food production is causing unprecedented changes in biodiversity, and thus, rising concern on the sustainability of intensive farming systems. The agriculture intensification has a substantial impact on plant and animal diversity (Gabriel et al., 2006; Jonason et al., 2011). However, the effects of agricultural management on below-ground diversity are not well understood (Li et al., 2012). This lack of knowledge is a significant concern because soil-borne microbes, especially bacteria, represent the majority of biodiversity in soil ecosystems and are involved in multiple ecosystem functions, including nutrient cycling (Pan et al., 2014; Navarrete et al., 2015) and plant health (Mazzola, 2004; Wakelin et al., 2013).

The environmental problems associated with the intensification of agriculture have initiated research efforts for conservation strategies. Converting conventional farms to organic farming systems seems to be a potential solution to diminish the loss of biodiversity and increase sustainable food production (Gonthier et al., 2014). Organic farming system consists of low-input agro-ecosystem farms in which plant productivity and ecosystem functionality are based on the natural availability of plant nutrients, use of green manure and biological pathogen control (Lammerts van Bueren et al., 2002). In contrast, conventional farming system relies on intensive use of agrochemicals, such as synthetic fertilizers to increase crop productivity and use of fungicides and pesticides to promote plant protection against pathogens (Kremen and Miles, 2012). Effects of farming systems on microbial communities are complex and time-dependent (Jonason et al., 2011). In general, it has been reported that management practices in organic farming systems change the microbial composition toward a more fast growing community (copiotrophic community) due to nutrients (Chaudhry et al., 2012), promote habitat diversification, increase the diversity and sustainability, and benefit microbial taxa involved in plant health when compared to conventional farming systems (Esperschutz et al., 2007; Sugiyama et al., 2010; Reilly et al., 2013; Gonthier et al., 2014). However, up to date, there are no studies about microbial community heterogeneity, which we refer as microbial community variability, in different farming systems. Although positive effects of organic management have been widely reported (Liu et al., 2007; Ge et al., 2008; Jonason et al., 2011; Hartmann et al., 2015), the effects of farming systems on diversity of microbial communities are complex and commonly controversial (Kleijn et al., 2001). Ge et al. (2008) found an increase in diversity after manure amendment, and other studies reported no differences or decrease in bacterial diversity and richness when organic systems were compared to conventional management (Liu et al., 2007; Reilly et al., 2013). Bengtsson et al. (2005) argue that in most cases, organic farming can be expected to benefit the biodiversity, but the effects will differ between organism groups and landscapes.

Agro-ecosystems often face problems with plant-pathogens, such as parasitic nematodes (e.g., Pratylenchidae and Meloidogynidae), and soil-borne fungi (e.g., Rhizoctonia solani and Verticillium dahliae) that affect a large number of important crops (Back et al., 2002). A common method to control these pathogens is the use of chemical pesticides, which are under critical review due potential toxic effect on non-target organisms and environmental pollution (Oka, 2010). Therefore, the development of methods for suppression of pathogens as an alternative to chemical control is an urgent need. These methods can be applied in organic farming systems, but also enable conventional farmers to reduce the use of pesticides. Alternative approaches are organic amendments (compost) (Mehta et al., 2014), cover crops (Asteraceae plants) (Pudasaini et al., 2006), green manure crops (grass-clover) (Widmer and Abawi, 2002), composts or non-composted waste products (chitin) or those based on physical methods (soil disinfestations) (Mowlick et al., 2012). Although these management practices are environmentally friendly, they are expected to induce shifts on microbial diversity and composition (Mehta et al., 2014). At the treatment level, the microbes play an important role in above- and below- ground processes, including their potential contribution to soil suppressiveness (Cretoiu et al., 2013). In this light, the ability to understand and manage microbial community through alternative practices for pathogen control, offer a promising approach to improve sustainable crop production.

The broad spectrum of agricultural managements and practices used for plant pathogen control in farming systems limits comparability among different studies (Liu et al., 2007; Xue et al., 2013). Up to date, there are few long-term agro-ecosystems experiments comparing organic and conventional farming systems (Esperschutz et al., 2007), and even more seldom are studies that make this comparison on plant pathogen control. This would be ultimately required for evaluating the sustainability of agricultural practice. One exception is the experimental field with Soil Health Treatments (SHTs) in organic and conventional farming initiated in 2006. The SHT experimental site in Vredepeel is a unique experimental field reported in contemporary literature with full-factorial experimental design and replicated experimental plots, where the same soil treatments, crop varieties, crop rotations and fertilization intensities are simultaneously applied in both conventional and organic farming systems under the same sandy soil type. Korthals et al. (2014) have evaluated the potential effects of the different SHTs on plant-parasitic nematode Pratylenchus penetrans, and on soil-borne pathogenic fungus V. dahliae. However, the long lasting responses of the soil microbial community to those different managements and the potential role of microbial community in soil suppressiveness were not studied. In this context, we assessed the bacterial and archaeal communities based on 16S rRNA gene marker by next generation sequencing to examine the response of microbial communities to conventional and organic farming systems and SHTs. The objectives of this study were to address the effect of farming systems and SHTs on (i) soil microbial diversity and presence of pathogen suppressors, and (ii) microbial community heterogeneity. Based on microbial community assessment, we aimed to detect specific structural shifts and identify microbial taxa associated with specific farming system or SHT, which might be useful as a bioindicator of sustainable management of agro-ecosystems and might bring novel insights on soil beneficial agriculture practices for soil health and plant productivity.

The SHE represents a unique experiment to assess the response of microbial communities to farming systems (conventional and organic) and Soil Health Treatments (SHTs). This study was limited to temporal sampling (single time point), spatial extent (local scale), and therefore should not be generalized for the farming systems performed in all ecosystems. Although the consistent results in this study provide novel ecological insights into microbial ecology in agro-ecosystems, concrete conclusions are still difficult and need to be confirmed by long-term experiments over distinct environmental conditions, management practices and larger geographic scales. Besides this, the complexity of microbial communities and the technical constraints so far, limited our understanding of the relationship between soil microbiota and agricultural managements. However, using the approach based on high-throughput sequencing of amplified taxonomic markers, we have described the microbial community structure and found that the soil microbiome is more heterogeneous in organic than conventional farming system, and additionally identified potential microbial pathogen suppressors and individual microbial taxon associated with specific management practices.

It is difficult to draw robust and generalized conclusions on the effect of systems management on microbial diversity, but an increase in microbial diversity has been repeatedly observed in organic in comparison with conventional system (Mader et al., 2002; Hartmann et al., 2015). The increase of microbial diversity in organic systems is strongly associated with the management applied, including the organic amendments and practices related with reduction or absence of chemical inputs and biological plant protection (Sun et al., 2004; Chaudhry et al., 2012). The enhancement of microbial diversity also benefits the functional activities and a more heterogeneous distribution of species within the microbial assembly, which implies in a stable and functional redundant community, leading to an ecosystem functionality built on healthier interactions between the different trophic ecosystem levels (Brussaard et al., 2007; Postma et al., 2008; Crowder et al., 2010; Wagg et al., 2014). The decrease of microbial diversity in the conventional system may be explained by the direct or indirect long-term stresses caused by the use of pesticides, fungicides and herbicides used for plant protection. These agrochemicals reduce the total microbial diversity because of the potential to inhibit or eliminate certain groups of microbes and select members adapted or able to growth under conventional farming practices (El Fantroussi et al., 1999; Liu et al., 2007; Stagnari et al., 2014).

Our study revealed consistent farming system effects on microbial community variability, suggesting, for the first time, more heterogeneous community in organic than in conventional system. We suggest that the availability of rich substrate in soil through the introduction of cattle farm yard manure, the biological practices without the interference of synthetic compounds and the presence of weed species provide heterogeneous habitat niches, which can be occupied by a highly variable microbial community resulting in an increase of the beta-diversity. The lower heterogeneity (that is, the lower beta diversity) in microbial community in conventional system is an indication of biotic homogenization, the process of increasing similarity in the composition of communities across an array of taxonomic or functional groups (Olden et al., 2004). Biotic homogenization is a common pattern of the above-ground community in conventional systems (Gabriel et al., 2006), and recently was reported for microbial communities as a response to long-term cultivation (Montecchia et al., 2015). When poor agricultural practices are applied, such as uniformly crop monocultures, fertilization and intensive use of agrochemicals, the chain-reaction of (bio)diversity loss reduce the ecological niches leading to a homogenization of the microbial community and their functional gene pool, altering the ecosystem functioning and reducing the ecosystem resilience (Olden et al., 2004; Constancias et al., 2014; Figuerola et al., 2015). We acknowledge that the plant species planted in conventional and organic systems between 2006 and 2013 were not the same. This might have some impact on rhizosphere microbial community due to the different exudates released by different plant species. However, in this study we have focused on bulk soils and not on rhizosphere microbiome.

Besides the effects of farming systems on microbial community, we hypothesized that there is a legacy effects of the SHTs on diversity. It is expected that the differences between SHTs (e.g., organic matter composition, C/N, physical disturbances) may alter the physical, chemical and biological properties of the soil with consequent shifts in microbial diversity (Jacquiod et al., 2013). However, this study does not support evidence for the occurrence of long-term effects of SHTs on microbial diversity and richness. The first possible explanation is that different SHTs affects microbial diversity only in short-term and this effect may not be observed 3 years after the last application of the different treatments in this study. Some studies suggest a strong and fast resilience of the microbial diversity after a pronounced disturbance on soil community caused by management practices (Ding et al., 2014; Suleiman et al., 2016). Second, the continuous long-term farming system can counteract the effects of the soil health treatments, which were applied only twice. It has been suggested that long-term management practices are more likely to greatly influence the microbial community than temporal disturbances in soil (Buckley and Schmidt, 2003). Finally, we believe that the legacy effect of the SHTs occurs in specific microbial groups and cannot be resolved by determining the diversity and heterogeneity of entire microbial community, because shifts in some groups might be compensated by shifts in others.

It has been proposed that due to larger availability of organic carbon and nitrogen, organic system should favor copiotrophic bacteria, while oligotrophic should predominate in conventional systems, where the organic carbon quality is low (Ding et al., 2014; Hartmann et al., 2015). In this study, we observed that the differences in the structure of microbial communities between conventional and organic farming systems were mainly related to a large fraction of habitat specialist OTUs broadly dispersed across the phylogenetic groups belonging to almost all phyla found in soil. Only few taxonomic groups revealed to respond more uniformly to farming system. For example, most of habitat specialists assigned to Proteobacteria and Euryarchaeota were associated with conventional system and an increase of members belonging to Acidobacteria and Planctomycetes was detected in organic system. These findings are not necessarily surprising, but are an opposite trend toward the copiotrophic-oligotrophic categories expected. However, the rather dispersed OTU association between conventional and organic systems within these taxonomic groups are in agreement with the contrasting behavior of individual members within phyla reported previously (Rousk et al., 2010). Not all members in a taxonomic clade demonstrate the same ecological characteristics, implying that the general lifestyle classification might not be applied for all members in a phylum (Navarrete et al., 2013), and responses to management will occur at lower taxonomic levels rather than at major groups. Proteobacteria have been suggested to be a primarily copiotrophic phylum in soil (Li et al., 2012), while the lifestyle of microbial groups belonging to Euryarchaeota, which are predominately methanogens, are largely unknown (Angel et al., 2012). However, the increased abundance of taxa belonging to these two Phyla in conventional farming system may be promoted by the input of fertilizers, which create copiotrophic environment in nutrient-rich microhabitats and stimulate plant growth, enhancing carbon availability and favoring the growth rate of members of these phyla. Members of Acidobacteria and Planctomyces have been suggested to be adapted to nutrient-poor soils, and the input of organic amendments is expected to inhibit their activity (Buckley et al., 2006; Chaudhry et al., 2012). However, Acidobacteria and Planctomyces might be involved in the turnover of soil organic carbon and nutrient availability, pointing out that the manure addition in soil might promote the proliferation of these groups.

Microbial communities proved to be sensitive to SHTs. This is an important finding because microbial taxa strongly associated with management practices may help to elucidate the process behind soil suppressiveness. In previous study in the same SHE (Korthals et al., 2014), the SHTs were evaluated within conventional system on the potential effects on plant-parasitic nematode P. penetrans and soil-borne fungi V. dahliae. The combination, chitin, anaerobic soil disinfestation and marigold treatments were more effective in controlling P. penetrans and V. dahliae when compared with chemical control. In contrast, grass-clover, biofumigation, cultivit and compost were not effective alternatives. However, in that study, the bacterial community was not assessed. In this study, we revealed several taxa associated with SHTs distributed among major taxonomic groups, for which we have little or no information about their ecological roles. Therefore, we can only speculate the ecological importance of the detected taxa based on what has been described in previous studies and compare with findings on pathogen control (Korthals et al., 2014). A complete description of the results is beyond the scope of this study and we only focus on some consistent findings and their potential as soil microbe indicators for sustainable practices.

In anaerobic soil disinfestation treatment most of habitat specific OTUs were represented by taxa belonging to Bacillales and Clostridialles (Firmicutes), whose dominance is linked to their spore-forming capability, a competitive advantage under anaerobic conditions. Members belonging to family Bacillales have been described to be responsible for suppression of soil-borne disease-causing fungi (Verticillium, Rhizoctonia and Fusarium) and plant-parasitic nematodes (Meloidogyne and Pratylenchus) through production of antimicrobial compound and pore-forming toxins (crystal proteins) (Wei et al., 2003). Thus, this treatment selected Firmicutes taxa that might be involved in suppression of fungi and nematodes. In addition, habitat specific OTUs belonging to phylum Nitrospira, nitrite-oxidizing bacteria, were also associated with this treatment. This may be an indication of previous accumulation of ammonia (NH₃) and production of nitrite (NO₂), both nitrogenous compounds released due to decomposing of organic material known to play an important role in the suppression of fungi and nematodes (Tenuta and Lazarovits, 2002; Oka, 2010).

The genus Lysobacter, chitinolytic bacteria, was found to be associated with chitin treatment and have been described to have an important role in soil suppressiveness, with a potential antagonistic property against Rhizoctonia and nematodes plant pathogens (Tian et al., 2007; Postma et al., 2008). The genus Virgibacillus, another chitinolytic bacteria (Cretoiu et al., 2014), was also found to be associated with chitin treatment, but its role in soil suppressiveness is not described yet. Chitin is a major component of nematode egg shells and cell wall of most plant-pathogenic fungi, and it is assumed that chitin amendments increase the number of chitinolytic microorganisms and chitinase activity, which in turn suppress nematodes and fungi. Members of Flavobacteriales and Chitinophagaceae associated with marigold may also suppress soil nematodes by their chitinase activity (Glavina et al., 2010; Kharade and McBride, 2014), suggesting that besides its ability to produce nematicidal compounds, marigold can also recruit nematode-antagonistic microorganisms (Hooks et al., 2010).

The potential plant pathogens antagonists Pasteuria, Pseudomonas and Burkholderiales were associated with cultivit and grass-clover treatments. Bacterial taxa belonging to these groups have been described to be potential against plant-parasitic nematodes and fungi (Tian et al., 2007). However, our results suggest that multiple mechanisms may accounted for an effective soil suppressiveness and the simple presence of taxa with antagonistic behavior against plant pathogens is not a sufficient proof for successful suppression of a pathogen in soil (Weller et al., 2002). Thereafter, the alternative methods to control plant pathogens require more detailed studies in combination with molecular and traditional approaches used in plant pathology and microbiology.

Altogether our results indicate that conventional and organic farming systems had a major influence on soil diversity and composition of microbial communities while the effects of the SHTs were of smaller magnitude. Organic farming system promoted beneficial effects on biotic aspects regarding to microbial diversities, richness and community heterogeneity. However, the response of microbial community to farming systems is diverse and complex, and simple conclusions like “organic systems increased the soil biodiversity” may not be directly synonymous with concomitant increase in soil health and plant productivity. Furthermore, impact of the diversity losses in conventional system is not yet known; it is not clear how microbial diversity is related to ecosystem function and whether the changes in diversity we observed are reversible and the long-term consequences remain to be unexplored. Moreover, we detected that there is a legacy of the SHT which selects for treatment-specific microbial members that are consistent with the existing knowledge, but the limited phylogenetic and functional information precludes more definite conclusions about the beneficial impact of individual taxonomic groups with soil suppressiveness. However, the observed shifts in microbial diversity, community structure and individual taxon bring novel insights into the potential of managing the microbial community for sustainable agricultural productivity.

Design the experiment: GK and EK. Obtain and process the data: ML, TJ, and EK data. Analyze the data: ML and MH. Wrote the paper: ML and EK with contribution of all co-authors.

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