42 
the U.S. every year; and (2) billions of the bacterium Bacillus thurengiensis 
are released to control insects such as the Japanese beetle. Dr. Vidaver 
said: 
"...all of these preparations, furthermore, most certainly contain a 
minute number of variants or mutants. The laws of probability and the 
principles of microbiology apply universally and these variants are to 
be expected. So, in these two examples mutants are released all the 
time. " 
She pointed out that since the mutants possess no selective advantage the 
wild- type predominates . 
Dr. Vidaver said AGS proposes to field test a specific INA - mutant strain 
of Pseudomonas syr ingae . The INA~ mutant strain was obtained from a parent 
Pseudcmonas syr ingae INA + strain by the process of double-reciprocal 
recombination . 
Dr. Vidaver said AGS preposes to spray less than two-tenths of an acre with 
2 x 1C>11 colony forming units. Because 99 percent of the original inoculum 
is expected to die, the approximate effective concentration will be 2 x 10^ 
viable bacteria. An elementary calculation based on simple assumptions 
suggests the natural population of Pseudcmonas syringae in the test plot 
is approximately equivalent to the applied effective concentration of 
mutant INA~ bacteria. 
Dr. Vidaver said greenhouse data show the test strains are not harmful to 
the test plant nor are they harmful to a variety of economic plants grewn 
in the area. Dr. Vidaver said AGS had adequately designed the test plots 
and proposes to monitor bacteria on test plants, on nearby plants, and in 
soil by a combination of tests. AGS has proposed prudent disposal methods 
should this be necessary. 
Dr. Vidaver recommended approval of the AGS proposal with five stipulations: 
(1) Plants should be monitored for bacteria at shorter intervals than the 
proposal suggests. Sampling should be performed at the time of application 
and then two or three days later to obtain data on the fate of the initial 
inoculum. (2) Approval should be given to test a specific crop. (3) Approv- 
al should not be given for continuing tests until information has been 
reported to RAC showing no problems arose during the first limited field 
test. (4) RAC should request information on dissemination, persistence, 
and efficacy of the released strain. This may be done on a confidential 
basis. (5) Initial approval should be for one growing season. If the data 
reported back to RAC show no problems, RAC cculd recommend approval for 
two additional growing seasons. Dr. Vidaver said generally three growing 
seasons are necessary to obtain adequate predictive data in agriculture. 
Dr. Vidaver said she wished to say for the record that she was more 
concerned with certain experiments over which RAC has no control than over 
the AGS proposal. Dr. Vidaver explained that one potential commercial use 
[495] 
