Table 1.

Possible ways to improve phytoremediation research using a metaorganism approach.

	Metaorganism approach	Example research strategy
Selecting the right plant host	In function of the microbiome	A pyrosequencing study identified that native willow cultivars were associated with a different microbial community than non-native cultivars, across a hydrocarbon contaminated soil (Bell et al., 2014). Fungi were more sensitive to hydrocarbon contamination then bacteria, and reacted different to willow introduction, suggesting that plant species selection (and evolutionary history) are not to underestimate with regard to their effect on microbiome establishment and influence on phytoremediation activity.
	Subsequent microbial inoculation practices	Fungi grown in their soil of origin with native plant species have been shown to be more mutualistic (more arbuscules), which can enhance the introduction of cooperative strains at a later stage (Johnson et al., 2010)
	Breeding, and transgenesis of plants for high biomass/rapid growth, high tolerance, uptake and detoxification potential	Transgenic tobacco cultivars over-expressing a bacterial nitroreductase resulted in improved TNT detoxification, and additionally increased the functional diversity of the rhizosphere microbial community (Travis et al., 2007), a double positive effect which warrants further investigation.
Interfering with root-exudates (diet)	Rhizoengineering	There are promising outlooks to change the quality and quantity of root exudates in the rhizosphere to optimize plant growth (Zhang et al., 2015) and biodegradation (Narasimhan et al., 2003). Transgenic Arabidopsis plants that exuded the xenotopic compound octopine, significantly increased the ratios of octopine degraders (Mondy et al., 2014).
	Selecting plant traits for their global interaction with the rhizosphere microbiome	Maize seedlings were shown to exude a high concentration of the compound DIMBOA which exerts antimicrobial activities in the rhizosphere (Neal et al., 2012). In addition, DIMBOA also attracts a catabolic (Neal et al., 2012), but also recruiting a catabolic, plant-beneficial rhizobacterium Pseudomonas putida KT2440. Can we identify more plant traits that globally interact with the rhizosphere microbiome?
Modify the driving forces	Elucidating the main factors that influence the plant–microbiome interactions	Antibiotic administration altered the community structure of gastrointestinal microbiota (Robinson and Young, 2010), similarly antibiotic addition to soil changed the soil microbiome leading to increased hydrocarbon degradation rates (Bell et al., 2013a). It is assumed that reduced interspecies competition enhances the catabolic activity of degradative strains.
	Studying interactions at the single microbial cell level in situ, and extrapolate/confirm findings at the population and community level (bottom-up).	Remus-Emsermann et al. (2012), used gfp-tagged individual Erwinia herbicola cells as bioreporter, to better understand bacterial colonization of the leaf surface of Phaseolus vulgaris plants. They suggested that the ‘carrying capacity,’ can be understood as the sum of local carrying capacities (Remus-Emsermann et al., 2012) whereby leaves contain a few sites where individual cells can produce high numbers of offspring, and the remainder of the leaf offers sites with low and medium reproductive success. Using such a bottom-up approach in phytoremediation, can help to better understand bacterial colonization of the rhizosphere at the population and community level.
Feeding the supply lines	Isolation of previously difficult to culture degradative strains	Continued efforts in culture-based techniques including the use of improved culture media and intelligent devices such as the i-Chip, have enabled the cultivation of a broader collection of previously difficult to cultivate microorganisms (Nichols et al., 2010; Stewart, 2012).
	Selection of strains/consortia to inoculate	The identification of the core rhizosphere microbiome and core root microbiome (Lundberg et al., 2012; Yeoh et al., 2015), i.e., strains that sufficiently depend on host genotype, but remain consistent across different soil types and developmental stages, allows selecting strains for their intimate interaction with the host, which should facilitate their introduction/enrichment in the native microbial community.
	Time point of inoculation matters, prioritization	Preemptive colonization of plant leaves with beneficial bacteria was found to reduce, but not completely exclude, the ability of secondary colonizers to reproduce and proliferate (Remus-Emsermann et al., 2013). These findings have direct relevance to phytoremediation, based on preemptive exclusion of opportunists and pathogens when there is a high level of catabolic strains at the start.
	Exploiting horizontal gene transfer	Introduction of the endophyte Burkholderia cepacia VM1468, equipped with the pTOM-Bu61 plasmid coding for toluene degradation, in the rhizosphere of yellow lupine has shown to dramatically reduce toluene evapotranspiration. Interestingly, the catabolic plasmid was found to transfer from the inoculated strain to different members of the endogenous plant endophytic community (Taghavi et al., 2005). As such harnessing horizontal gene transfer is a simple and inexpensive way to enrich catabolic traits.