I have read that nitrogen fixing bacteria have the capacity to infect some plant roots and, via nucleic plasmids, transfer symbiotic enchancing genes into the recipient. Here is an example:
Abstract The symbiotic plasmid (pSym1-32) of the highly effective Rhizobium leguminosarumbv. viceae1-32 strain was identified after the conjugal transfer of replicons carrying Tn5-mobinto the plasmidless Agrobacterium tumefaciensGm1-9023 strain. Plasmid pSym1-32 was transferred intoR. leguminosarumbv. viceaestrains Y14 (showing low effectiveness of symbiosis with Vicia villosa) and Y57 (unable to fix nitrogen). Transconjugants formed Fix+nodules on roots of V. villosaand had a highly enhanced nitrogen fixing ability, increased plant weight, and increased nitrogen accumulation compared to the recipient strains. Variation of transconjugants in symbiotic properties (accompanied by alterations in plasmid composition in some of the conjugants) was detected. Moreover, the donor strain R. leguminosarumbv. viceae1-32 was shown to be more efficient in the competitiveness and acid tolerance than the recipient Y14 strain. Both these properties were transmitted upon transfer of pSym1-32 into the recipient. Thus, plasmid pSym1-32 was shown to carry genes involved in the control of the nitrogen fixing ability, symbiotic effectiveness, competitiveness, and acid tolerance in R. leguminosarumbv. viceae.
O. N. Kurchak
Email: genet@riam.spb.su
N. A. Provorov
Email: genet@riam.spb.su
B. V. Simarov
Email: genet@riam.spb.su
http://www.springerlink.com/content/w210265361863322/
The relevance to rose breeding is the potential to create low fertilizer requirements in roses, which would likely have a significant market impact. Although the success probability is low, rose breeders are known for their patience, eg. sorting through 10,000 seedlings in the hopes of finding a handful of slightly improved varieties.
My suggestion is an experiment to interplant roses with a variety of short leguminous plants (clover, alfalfa, non climbing peas and beans, …) to infect the roots with several leguminous plasmid genes. Bacteria are extremely adaptable given time. If the roses are left unfertilized, they may be receptive to the benefits of at least partial symbiosis.
Any Rosarian possessing genetics laboratory equipment could transfer the plasmids without the need of infection, possibly accelerating the process.