8 • Alternatives to Animal Use in Research, Testing, and Education 
as genetic background, prior experience, and 
environmental conditions, all of which affect 
behavior and can obscure the influence of the 
factor under study. 
• The short lifespans of certain animals allow 
scientists to study behavior as it develops with 
age and across generations. 
• Some animal behavior is less complex than hu- 
man behavior, facilitating an understanding 
of basic elements and principles of behavior. 
• The behavior of certain animals holds particu- 
lar interest for humans . These animals include 
companion species, farm animals, and agri- 
cultural pests. 
Although behavior is a biological phenomenon, 
behavioral research differs substantially from bio- 
medical research in that researchers have fewer 
opportunities to study mechanisms isolated from 
living organisms. There is little prospect, for ex- 
ample, of using in vitro cultures to look at aggres- 
sion, habitat and food selection, exploration pat- 
terns, or body maintenance activities— all topics 
studied by behavioral scientists. Yet in each of these 
disciplines, reduction or refinements of animal use 
may be possible. It is the continued, but modi- 
fied, use of animals that holds the most prom- 
ise as an alternative in the field of behavioral 
research. 
ALTERNATIVES IN TESTING 
Several million animals are used each year 
in testing substances for toxicity and establish- 
ing conditions for safe use. The resulting data — 
together with information about use and ex- 
posure, human epidemiologic data, and other 
information — are used in assessing and man- 
aging health risks. 
As a reduction in the number of animals is a prin- 
cipal alternative, proper statistical design and anal- 
ysis in testing protocols play an important role (see 
ch. 7). The total number of animals needed for sta- 
tistically significant conclusions depends on the 
incidence of toxic effects without administration 
of the test substance, the degree of variation from 
animal to animal for the biological effect that is 
of interest, and the need to determine a quantita- 
tive relationship between the size of the dose and 
the magnitude of the response. Statistical analy- 
sis plays a similarly important role in research. 
One of the oldest and, perhaps for that reason, 
least sophisticated tests is the LD 50 ("lethal dose" 
for "50” percent of the test animals). In this short- 
term, or acute, test, a group of animals, usually 
rats or mice, are exposed to a single substance, 
and the measured end point is death (although 
other observations may be made). The LD 50 is the 
dose at which half the test animals can be expected 
to die. A range of doses is administered to some 
30 to 100 animals and the LD S0 is calculated from 
the results. Tests providing the same informa- 
tion have recently been developed using as few 
as 10 animals, i.e., a 3- to 10-fold reduction. 
The LD S0 is used to screen substances for their 
relative toxicity and mode of toxic action. Scien- 
tists and animal welfare advocates have criticized 
it in recent years, in part because it cannot be ex- 
trapolated reliably to humans, and in part because 
the imposition of a highly toxic or lethal dose seems 
particularly inhumane. 
Another often-criticized acute toxicity assay is 
the Draize eye irritancy test. This involves plac- 
ing a test substance into one eye of four to six rab- 
bits and evaluating its irritating effects . Results are 
used to develop precautionary information for sit- 
uations in which exposure of the human eye to 
the substance is possible. Substances with certain 
properties— e.g., a caustic pH — could be assumed 
to be eye irritants and not tested. Draize proce- 
dures may also be modified to reduce pain, and 
in vitro methods to test for irritancy are under 
development. A promising new bioassay for tis- 
sue irritancy makes use of the chorioallantoic mem- 
brane of the chick embryo (see fig. 1-1). 
Other common tests include those for long-term 
chronic effects, carcinogenicity, reproductive and 
developmental toxicity, skin irritancy, and neuro- 
toxicity. In addition to such descriptive toxicology 
(i.e., tests that focus on the response of the organ- 
