ENVIRONMENTAL DEFENSE FUND 
257 Park Avenue South 
New York. NY loom 
(212)505-2100 
September 4, 1987 
1616 P Street. NW 
Washington. DC 20036 
(202)387-3500 
1405 Arapahoe Avenue 
Boulder. CO 80302 
(303(440-4901 
2606 Dwight Wav 
Berkeley. CA 94704 
(415)548-8906 
1 108 East Main Street 
Richmond. VA 23219 
(804) 780-1297 
Director 
Office of Recombinant DNA Activities 
12441 Parklawn Drive 
Suite 58 
Rockville, MD 20852 
Dear Director, 
Please consider the following comments on the 
containment guidelines for plants and animals proposed 
for NIH's "Guidelines for Research Involving 
Recombinant DNA Molecules." (Federal Register, August 
11, 1987). 
We are impressed by the thoroughness of the 
guidelines for research with animals (Appendix Q and 
related amendments). We hope the RAC will recommend 
that NIH adopt them. On the other hand, we urge the 
RAC make a number of changes to the plant guidelines 
(Appendix P and related amendments) , as detailed below 
(the numbers refer to the associated sections of the 
Federal Register notice). 
4. BL1-P, the lowest level of containment for plants, 
is recommended for "experiments in which there is no 
recognizable and predictable risk to the environment in 
the event of accidental release or for which there is 
convincing biological evidence that precludes the 
possibility of survival, transfer, or dissemination of 
the recombinant DNA molecules into the environment." 
BL2-P, the next higher level, is "recommended for 
experiments in which there is a recognized possibility 
of survival, transmission, or dissemination of the 
recombinant DNA-containing organisms but the 
consequence of such a release has a predictably minimal 
biological impact in the event of inadvertant release." 
The "or" in the description of BL1-P should be changed 
to "and." Otherwise BL1-P and BL2-P are roughly 
equivalent; both are recommended when there is 
no/minimal effect predictable from inadvertant release. 
Since one can always come up with a scenario in 
which an engineered organism can have minimal impact, 
we believe distinguishing between "no" effect and a 
"minimal" effect will be impossible. Furthermore, 
Recombinant DNA Research, Volume 13 
[333] 
