Soon after the isolation of nodule bacteria in 1888, differences were recognized in the ability of bacterial strains to form nodules on particular host plants and in the nitrogen-fixing ability of the nodules so formed. These and other symbiotic heterogeneities were attributed, sometimes correctly, to bacterial strain differences, not then thought to be open to formal genetic analysis. The realization that the host plant was an essential component of this variability came only gradually, stimulated by observations of host varietal differences and by the demand for reliable and homogeneous material for experimental work. Only within the last two decades has host variability been studied by plant breeding, and bacterial strain differences by some of the methods of microbial genetics. This review, except for a brief reference to earlier work of some historic interest, will consider only genetic problems open to investigation by these methods. The developmental sequence in all legume nodules is broadly similar. The initial infection phases are followed by the induction of the nodule, the invasion of part of the nodular tissue and culminate in bacteroid formation and nitrogen fixation; the genetics of symbiosis will be considered in this context.
The genetics ofUstilago maydis
