Abstract
Plants shape distinct, species-specific microbiomes in their rhizospheres. A main premise for evaluating microbial communities associated with root-soil compartments is their successful separation into the rhizosphere (soil-root interface), the rhizoplane (root surface), and the endosphere (inside roots). We evaluated different approaches (washing, sonication, and bleaching) regarding their efficiency to separate microbial cells associated with different root compartments of soil-grown rice using fluorescence microscopy and community fingerprinting of 16S rRNA genes. Vigorous washing detached 45% of the rhizoplane population compared to untreated roots. Additional sonication reduced rhizoplane-attached microorganisms by up to 78% but caused various degrees of root tissue destruction at all sonication intensities tested. Treatment with sodium hypochlorite almost completely (98%) removed rhizoplane-associated microbial cells. Community fingerprinting revealed that microbial communities obtained from untreated, washed, and sonicated roots were not statistically distinguishable. Hypochlorite-treated roots harbored communities significantly different from all other samples, likely representing true endospheric populations. Applying these procedures to other root samples (bean and clover) revealed that treatment efficiencies were strongly affected by root morphological parameters such as root hair density and rigidity of epidermis. Our findings suggest that a careful evaluation of separation strategies prior to molecular community analysis is indispensable, especially when endophytes are the subject of interest.