be counted by the beaters or by observers sta- 
tioned along a terminal line. Obviously, this 
type of census is expensive and unsuitable for 
large-scale operations, although the Indians 
used it effectively for herd control. A similar 
approach that of simple visual observations 
made along predetermined and_ well-defined 
transects by an observer walking these lines, 
provided valuable census data in the recent 
study of the white-tailed deer in Texas (Teer, 
Thomas, and Walker 1965). 
Spotlighting or night-shining, in which the 
presence of an animal is often indicated merely 
by the reflection of the projected light from 
the eyes, is another form of remote sensing. In 
open areas in the forest, this form of census is 
useful, although numbers are usually underes- 
timated. 
The woodcock peenting census is a form of 
remote sensing in which the ear is the sensor 
as often as the eye. A similar technique is the 
widely used ruffed grouse census method, the 
drumming count. Problems involved in these 
techniques are well known and have been dis- 
cussed in detail elsewhere. 
Any form of visual or audio observation 
should be classed as remote sensing, whether it 
be searching for eagles from the ground or 
using a dog and his sensitive nose to flush and 
thus census grouse or pheasant. Today then, 
most of our techniques for determining popula- 
tion size and distribution are not based on con- 
tact, but on indirect or remote approaches. 
AERIAL CENSUS METHODS 
The aerial surveys of big-game animals more 
closely approach the modern concept of remote 
sensing. For some years Minnesota and On- 
tario as well as other Canadian provinces and 
Alaska have regularly censused moose by in- 
tensive aerial surveillance over designated 
sample areas. They have found it more effec- 
tive to use trained observers than to take and 
then interpret aerial photographs. The aerial 
observer can zero in on a questionable spot, 
and the pilot will circle lower until he 
frightens the animal out of the brush. 
Aerial observation for census purposes has 
been used extensively in Africa; the work of 
Darling (1960) is a good example. 
The aerial photography (both motion and 
still pictures) of wolves on Isle Royale, taken 
under the direction of Allen and Mech (Mech 
1966) provided census data on the wolfpack. 
The pictures are supported by visual counts in 
which individuals are identified. Aerial obser- 
vation is presently being used for many other 
census needs. For example, Wisconsin conducts 
a visual aerial survey of breeding waterfowl on 
60 transects that are each 50 miles long. There 
is added benefit in using a human as the sen- 
sor; i.e., the observer can often identify the 
species of bird seen. 
The usefulness of aerial photography in enu- 
merating waterfowl in large concentrations 
has been demonstrated, particularly for the 
Pacific Coast flyway (Chattin 1952). With 
low-level photography, the interpreter can 
readily count birds and, when necessary, sub- 
sample the pictures. Recently Kadlec and 
Drury (1968) reported on a method to deter- 
mine trends in gull populations by combining 
visual estimates of numbers with photography 
of both the island nest areas and the flocks of 
birds utilizing the islands. They state, ‘Test 
predictions of the numbers of nests from 
counts of gulls on photographs gave acceptable 
results even for most individual colonies,” and 
“Neither visual estimates nor photography 
will reliably detect annual change of less than 
about 25 percent.’”’ (in the breeding gull popu- 
lation) 
Wisconsin and other States use aerial sen- 
sing for checking on a variety of water activ- 
ity. Boating and fishing surveys made from the 
air provide knowledge of people-pressure on 
lakes of different sizes and at different times of 
day. Water skiing is also censused by aerial ob- 
servation. 
Thus, aerial sensing techniques are not new 
or unusual in wildlife population census, and 
they can be expected to increase. The question 
is how effective and economical will we find the 
highly sophisticated sensing methods now 
being developed for other aspects of the evalu- 
ation of natural resources. 
As Poulton indicates in his conference paper, 
few range and wildlife researchers have fully 
appreciated or fully used the potential of con- 
ventional aerial photography. There is much 
useful information on these photos for habitat 
evaluation, and perhaps also opportunities for 
population evaluation. Aerial photography has 
little application to pellet group counting; on 
the other hand, Reppert and Driscoll (1968) 
suggest in their conference paper that it might 
be very useful for species that produce casts or 
burrows. Photography is effective only when 
the object to be photographed can be detected in 
the picture. So much of the ground surface is 
screened by vegetation in the Lake States and 
Northeast that not only is aerial photography 
of limited value for counting animals, but other 
forms of remote sensing, such as radar and 
infrared scanning, may also be severely limited. 
Seasonal changes in cover with spring, fall, or 
winter aspects sometimes provide good oppor- 
tunities. However, animal habits in relation to 
cover, wide distribution of conifers, and other 
factors tend to amplify the influence of the tree 
canopy. 
Nonetheless, aerial photography is useful in 
195 
Se ee 
