tional status and familiarity with forages 
might influence choices in a given situation, 
Neff (1967b) said, “The food choices of Wall- 
mo’s pine country deer from Flagstaff when 
taken to an enclosure in the chaparral near 
Prescott in 1964, were almost identical to those 
I had observed in local deer.” 
The prevailing evidence, mostly empirical, 
indicates that captive animals exhibit prefer- 
ences for natural forage similar to those of 
wild animals. McMahan (1964) even concluded 
that the level and timing of supplementary 
feeding caused little difference in the deer’s 
grazing behavior. Watts (1964) concurred: 
“Tnitially the deer were not fed for 24 hours 
prior to field observation, but this practice was 
later discontinued because it appeared to have 
no effect upon feeding time of the deer in the 
field.” 
What animals eat is greatly determined by 
inherent characteristics of their digestive 
physiology (see Nagy 1969 in this volume). 
Longhurst et al. (1968) contend that the sens- 
ing and selection of acceptable foods by deer 
involves smell, touch, and taste. It may be as- 
sumed that any native forage that is sensed as 
palatable by a tame animal will be acceptable 
to a wild animal, and vice versa, barring 
marked differences in the prior nutritional sta- 
tus of the animals. This becomes most proble- 
matical when tame animals, maintained on an 
adequate pen ration, are grazed where wild an- 
imals are finding forage supplies inadequate. 
However, it might shed light on the true palat- 
ability of species being taken by wild animals 
under such conditions. 
SIGNIFICANCE OF INDIVIDUAL 
VARIATION 
In biological research it is accepted that at- 
tributes of an individual may differ widely 
from the mean of a population. The hazards of 
basing conclusions on the effects of or effects 
on too few individual animals are well known 
in range and livestock husbandry research. In 
comparing the forage choices of two each of 
sheep, goats, cows, and deer, McMahan (1961) 
found that “. .. no two individuals of each 
class (except the two cows) were from the 
same statistical population . . .” It was Hea- 
ly’s (1967) opinion that “Future studies . . 
should incorporate additional deer to deter- 
mine the significance of individual variation in 
feeding habits. Because only three deer were 
used, it is impossible to assign preference 
weights to deer by sex.” 
Probably each study requires either pretest- 
ing to determine the number of animals needed 
to acquire data of desired quality or, on the 
basis of existing limitations, using arbitrary 
numbers and kinds of animals and accepting 
the inherent limitations in the results. 
106 
FORAGE IDENTIFICATION 
The opportunity to identify any and all for- 
ages consumed is a most attractive aspect of di- 
rect observation of tame animals. Forages of 
unknown taxonomic identity can be collected 
or marked for future identification. Occasion- 
ally the items eaten are so small that they can- 
not be seen from a normal observing stance. In 
such cases the authors have gone to hands and 
knees and even snatched fragments from the 
deer’s mouth to see what is being taken. 
Parts of plants chosen and their phenologi- 
cal state can be noted. Equally important, spe- 
cies or plant parts that are consistently ig- 
nored or are tested and rejected by the animal 
can be noted. 
QUANTIFICATION PROBLEMS 
The reported studies by direct observation 
have used either time spent feeding * or num- 
ber of bites® to measure the relative use of 
different forages. 
Healy (1967) specified that the time meas- 
urement began with biting off or picking up 
the forage item, included chewing, and termi- 
nated with swallowing. Where the bite is the 
unit of measure, apparently most workers have 
considered it the act of breaking off or of pick- 
ing up a piece of forage. This piece might con- 
sist of one part or several parts (several stems 
or leaves in one bite). 
According to Smith and Hubbard (1954), 
. . quantity of material consumed from the 
various species cannot be accurately estimated 
by observation of feeding-minutes, for deer 
consume forage from different species with 
varying degrees of efficiency.’”’ The same might 
be said of bite counts. A better approximation 
should result if the feeding time or bites can be 
expressed in terms of weight of forage con- 
sumed. Neff (1967b) has proposed that weight 
per bite be determined by plucking “bites” by 
hand from each species of forage plant soon 
after the completion of deer-feeding observa- 
tions at each sample site. The average dry 
weight of 20 to 40 simulated bites would be 
used to estimate the weight of material con- 
sumed by the deer: the number of real bites x 
mean weight of simulated bites = amount con- 
sumed. 
Frels and Veteto (1966) attempted to deter- 
mine mean weights of portions of plants taken 
by a deer and a cow, and to record the exact 
portions of each bite taken. They have re- 
cently found this approach to be unworkable, 
and are now evaluating forage intake using 
esophageal cannulae (personal communica- 
tion). 
ee 1945; Buechner 1950; Watts 1964; Healy 
ip 
*Wallmo 1951; Brown 1961; McMahan 1964; Dciecio- 
lowski 1967; Neff 1967b. 
be 
