Soil is one of the most complex and diverse habitats as one gram of soil can contain from 10e3 to 10e7 bacterial species.
Apple replant disease is a complex syndrome that causes reduced growth and diminished production in apple trees that are replanted in the soil of previous orchard. Moreover, plants show significantly shortened internodes, discolored roots, root tip necrosis and a reduction in root biomass, which can ultimately lead to plant death within the first growing season and production caused by ARD may decrease profitability up to 50% throughout the life cycle of the orchard, incurring thus significant losses to agriculture.
In addition, distinct bioinformatic approaches (e.g. MGRAST, QIIME, mother, custom made pipelines) with their inherent assumptions, myriad of settings, algorithmic approximations and biases, distinct levels of congruency and benchmarking lead to contrasting results and interpretations. These limitations make the comparison and generalization of the results published in unrelated studies difficult.
The aim of this study was to identify the presumed microbiological drivers in soils affected by apple replant disease (ARD) based on reanalysis of a larger cohort of molecular data. A literature search was conducted to identify the relevant deep-sequencing datasets from ARD-affected soil microbiomes next to the data on environmental variables and molecular techniques. This resulted in 140 and 72 datasets for Bacteria and Fungi,respectively (n=212 datasets).The sequencing datasets of bacteria and fungi were analyzed using a taxonomic approach in mothur, using SILVA and UNITE databases, respectively. Variation partitioning and network analysis were used to identify the extent of variability related to environmental, spatial, spatially structured environmental or methodological sources.
The re-analysis and inclusion of metadata in this study confirmed that healthy soils contained distinct and more diverse communities than soils with a reduction in growth due to ARD. A change in the microbial association networks linked to nitrogen metabolism was observed suggesting that the differences in microbial community structure apparently reflected changes in the soil environmental chemistry and metabolite composition. As large portions of variability in bacterial and fungal communities remained unexplained our results also point to the importance of precise and more unified descriptions of soil environments to identify the key soil parameters or metabolites associated with changes in soil microbial communities towards infectious phenotypes In agreement with previous studies, our results indicate that for the multifactorial origin of ARD a model of component-necessary-sufficient causes is more appropriate as ARD can be classified as an opportunistic microbial infectious disease mostly shaped by a various complex constellations of environmental parameters with microbial constituents, affecting microbial physiology and their mutual interplay, culminating ultimately in plant disease. Future concerted studies aiming at systems level linking of information from microbial metagenomes, soil metabolites and soil physico-chemical parameters in relation to plant genotype will thus have the potential to disentangle the causative and associative network of parameters leading to the development of systemic effects that are reflected at the plant level and are known as ARD.
The MACH-2 architecture enabled the extension of our studies to a large number of datasets and large scale databases. Additional projects from the field of amplicon sequencing and environmental and medical metagenomics are in pipeline.
References