Survival and dispersion of genetically modified rhizobia in the field and genetic interactions with native strains

A genetically modified strain of the symbiotic nitrogen‐fixing bacterium Rhizobium leguminosarum biovar viciae was used to inoculate a typical host, pea, and a control non‐host cereal crop in the field. The inoculant was monitored for survival and spread from the site of application, and for genetic interactions with the native population. It could be identified by chromosomally located antibiotic resistance markers and additional markers conferred by the transposon Tn 5 inserted on its conjugative symbiotic plasmid. These markers facilitated enumeration of the strain on selective agar, enabling survival and spread to be monitored over a six year period. Although culturable cell numbers dropped two to three orders of magnitude after the first year, subsequently they remained around 10 2 viable cells per g soil, even in subplots where only the non‐host cereals had been grown. However, peas did give the inoculant a small survival advantage compared with non‐hosts. Soil cultivation appeared to play a major role in inoculant dissemination from the site of application. Transfer of the Tn 5 marker to other rhizobia could be monitored by screening for isolates with Tn 5 ‐encoded antibiotic resistance in the absence of the inoculant chromosomal markers. Over three years, more than 4000 pea root nodules were screened for indigenous rhizobia that had acquired the Tn 5 ‐marked symbiotic plasmid from the inoculant. None were detected, although overall about 2% of nodules contained the inoculant strain, and transfer of the Tn 5 ‐marked symbiotic plasmid to three out of four R. leguminosarum biovar viciae isolates from the field site could be demonstrated under laboratory conditions.