Résumé
So far, the analysis of microbial populations associated with wheat monocropping-induced decline of take-all disease (Gaeumannomyces graminis var. tritici) has focused mainly on culturable biocontrol pseudomonads. The objective of this study was to develop a taxonomic rrs (16S rRNA gene) microarray to assess the changes in Pseudomonas populations taking place during take-all decline. The microarray contains 12 probes for five Pseudomonas phylogenetic clusters chosen because they include well-known plant-beneficial pseudomonads. Four of the clusters are within the 'Pseudomonas fluorescens' species complex. PCR primers were selected to target these five clusters, and they were validated using 53 pseudomonads belonging or not to these clusters. Microarray analysis of the pseudomonads enabled discrimination between strains from several Pseudomonas clusters. Rhizosphere samples were collected from field plots grown with wheat for 1 (low level of take-all disease), 5 (high level of disease) or 10 years (low level of disease, suppressiveness reached). Microarray data could distinguish Pseudomonas populations from some of the wheat plants grown in the same plot. When comparing treatments, there was a difference between years 1 and 10. Cloning-sequencing of rrs enabled to define more precisely this difference by identifying two major Pseudomonas populations, one associated with year I and the other with year 10 (disease suppressiveness), which represent new clades within the 'P. fluorescens' complex. These populations may be useful as soil quality indicators. In conclusion, the combination of microarray and cloning-sequencing approaches highlighted changes in the prevalence of two major Pseudomonas populations, giving new insights on the dynamics of root-associated pseudomonads during take-all decline. (C) 2007 Elsevier Ltd. All rights reserved.