| A range of soil-borne diseases are increasingly threatening agricultural production around the world.Plant root-associated microbiomes are increasingly seen as a possible driver of natural pathogen resistance and have become a target for innovative strategies aiming at improving crop protection.However,the ability of rhizosphere microbial communities to keep diseases under control is influenced by many factors,including the microbial interactions within these communities.Unfortunately,we still have relatively little insight into how microbial interactions affect community assembly and how such interactions eventually impact plant health.This thesis seeks to examine how microbial interactions within the rhizosphere microbiome impact the ability of plant pathogens to proliferate and thereby cause plant disease.To this end,we use bacterial wilt disease in tomato plant,which is caused by pathogenic bacteria Ralstonia solanacearum(Rs)as a relevant model system.We started this thesis by presenting a field survey of tomato bacterial wilt disease across six regions in China where Rs infestation has created a mosaic of disease and healthy tomato plants,and compared the differences of interactions between resident rhizobacteria and Rs in healthy and diseased rhizosphere soil;we then tested pairwise interactions between resident species,predicted the interactions in multi-species bacterial communities and the ability of these bacterial communities resist to Rs invasion;at last,we examined the importance of direct(effects of rhizobacteria on pathogen growth)and indirect(effects of rhizobacteria on helper growth)pathways affecting pathogen growth.The main results are listed as following:1.We recovered bacterial community data from six tomato fields(Changsha,Ningbo,Nanchang,Nanning,Nanjing,Wuhan)across a wide geographic range where tomato wilt disease was observed.Rhizosphere samples were examined from both diseased and healthy tomato plants in each field.We first compared Rs density,and bacterial community composition between healthy and diseased rhizosphere samples(DRS).For all sites,we detected significantly higher densities of Rs in diseased rhizosphere soil samples as compared to healthy rhizosphere soil samples(HRS),the relative abundance of Proteobacteria was higher in DRS than HRS,while the relative abundance of Actinobacteria was higher in HRS than DRS.Then 515 bacterial strains from Nanjing rhizosphere soil were isolated,identified and tested for their effects on the growth of Rs.Our 515 isolates were classified into four main phyla,with the following distribution:Proteobacteria 34.4%,Firmicutes 31.7%,Bacteroidetes 17.5%and Actinobacteria 16.5%.This collection contained a total of 26families and 52 genera.A total of 37.9%of these isolated rhizobacteria were shown to inhibit pathogen growth,while 53.6%of them significantly improved pathogen growth,and these categories were referred to as pathogen antagonists and helpers,respectively.Although HRS or DRS both yielded strains with a wide range of effects on Rs growth,we isolated a higher proportion of Rs antagonists from healthy rhizosphere soils(40.4%)as compared to from diseased rhizosphere soils(34.7%).To relate the result of our cultivation-independent and-dependent approaches,we identified bacterial isolates within microbiomes of the rhizosphere soils,and compared co-occurrence patterns between bacterial community members and Rs in the field with the effects of matching bacterial isolates on the growth of Rs,and we observed discrepancies.2.We subsequently used controlled systems with synthetic microbial communities which allowed us to disentangle the role of specific organisms and their interactions in the ecological processes associated with bacterial wilt disease suppression.By comparing the productivity of six rhizosphere resident bacteria cultured separately and co-cultured respectively,the interaction between the two strains was classified as antagonism and facilitation,and it was found that the interaction between the two strains in this study was mainly driven by direct antagonistic competition between bacteria,rather than nutritional competition.We then predicted interactions among multi-species resident bacteria based on the interaction between two resident bacteria,and the prediction was verified.It showed that the interaction between two species can predict the interactions among multi-species(R~2=0.44,P<0.0001).3.In order to explore the effects and mechanisms of the interactions among rhizosphere resident bacteria on the invasion of R.solanacearum,the microplate experiment and the greenhouse pot experiment were combined to detect the Rs invasion in the complete combination of 6 resident rhizosphere bacteria.The results showed that the interactions between resident bacteria can predict the invasion of Rs:compared with facilitated resident communities,the antagonistic resident communities were more resistant to the invasion of Rs(P<0.01).Moreover,antagonistic resident communities were more inhibitory not only towards themselves but also against the invader.4.The above results showed that there are both bacterial inhibitors and helpers of R.solanacearum in the rhizosphere.To explore the role of the bacterial helper of Rs in the control of soil-borne bacterial wilt,we first selected two representative helper strains of Rs and 46 tomato rhizosphere bacteria that have different effects on the growth of pathogen Rs from 515 resident bacterial isolates.Then,the tripartite interactions between Rs,two model helper strains and the collection of 46 bacterial isolates recovered from the tomato rhizosphere were examined.This system allowed us to examine the importance of direct(effects of rhizobacteria on pathogen growth)and indirect(effects of rhizobacteria on helper growth)pathways of pathogen growth.We found that the interaction between rhizosphere isolates and the helper strain was the major determinant of pathogen inhibition in vitro and in vivo(P<0.01).We therefore propose that controlling microbiome composition to prevent the growth of pathogen helpers may become part of sustainable strategies for pathogen control.In conclusion,the composition of rhizosphere bacteria and their interaction characteristics affect the invasion of soil-borne pathogen.The pairwise interaction can not only predict the interactions among multiple strains,but also predict the communities resistant to the invasion of pathogens:compared with facilitated resident communities,the antagonistic resident communities were more resistant to the invasion,which was mainly mediated by direct antagonistic competition between bacteria rather than by nutritional competition.There are a large number of helper strains of soil-borne pathogen,which can stimulate the growth and colonization of pathogen,and promote the severity of bacterial wilt disease.The inhibition of the helper strains was a more effective path toward pathogen reduction than direct inhibition effects on the pathogen itself.These results will help us to understand the survival mechanism of soil-borne pathogen and provide theoretical support and new ideas for seeking strategies to suppress soil-borne pathogen. |