tal feeds (Doman and Rasmussen 1944), and 
differential range use (White 1960) are per- 
haps but a few of the applications. 
RECOMMENDATIONS 
Considerable knowledge about animal diets 
has been accumulated. However, the need for 
greater refinement, intensification, and diversi- 
fication is apparent. Better techniques are 
needed to provide unbiased estimates of what 
an animal has eaten. Beyond this, we need to 
know more about what foods are available to 
subject animals and where and under what 
conditions various food items are consumed. 
We need to know more about when certain 
foods are taken, particularly in relation to pe- 
riods of animal] stress and varying physiologi- 
cal demands, and about why some foods are 
preferred over others. We need to know much 
more about how dietary findings relate to both 
the environment and the animal. 
What is Consumed? 
A list of the foods consumed by an animal is 
incomplete as a contribution to our knowledge 
of animal diet. Information relating consump- 
tion to the foods available for consumption is a 
much more valuable contribution. Many early 
studies were deficient in this regard. More re- 
cent investigators have attempted to correct 
this deficiency and to relate relative consump- 
tion to relative availability (cf. Morris and 
Schwartz 1957; Loveless 1959; Ward and 
Keith 1962; McMahan 1964; Harper et al. 
1967; and Chamrad and Box 1968). Estimates 
of food availability, however, have generally 
been limited in scope and intensity. Current in- 
vestigators of animal diets should attempt to 
obtain more precise estimates of availability, 
based on adequate samples, and use an estima- 
tor comparable to that used for consumption. 
Microscopic identification of digestive-tract 
material appears to be a more refined and per- 
haps a less biased approach to estimating die- 
tary consumption for some animals. Proce- 
dures employed by Malechek (1966), Sparks 
(1968), and Sparks and Malechek (1968) seem 
particularly promising: small subsamples of 
rumen material were first ground and then 
mounted on slides, and the components were 
identified by comparing reference slides of 
diagnostic epidermal tissues under a compound 
microscope. 
Particular parts of food items consumed 
have not received much emphasis in many 
studies. Others (cf. Dixon 1934; Hill 1946; 
Van Dyne and Heady 1965) have presented 
rather detailed lists of plant parts consumed as 
well as plant species and plant groups. The 
part of the plant selected could be of more sig- 
nificance than the species eaten. 
Poisoning among wild animals may be more 
of a problem than is realized. Case and Mur- 
phy (1962) concluded that research on poison- 
ing is comparatively recent and offers many 
opportunities for valuable work. 
Statements of probability, tests of compara- 
tive data, and statistical descriptions of varia- 
bility are rarely reported in food-habits litera- 
ture. Variability estimates may be useful in 
evaluating results and as an aid in determining 
sample sizes in future work. As an objective, 
presentation of at least minimal descriptive 
statistics and probability estimates would not 
appear to require excessive labor. 
Where is it Consumed? 
Smith (1952), Hosley (1956), Boag (1968), 
and others have indicated that dietary findings 
generally have local application. Cowan (1945) 
pointed out that palatability ratings are applic- 
able only under the precise conditions existing 
when they were calculated and that their 
transfer by inference to other areas where dif- 
ferent conditions prevail can lead to erroneous 
conclusions. Errington (19382, p. 76) suggested 
that digestive-tract examinations ‘“‘at best show 
only what individual animals in a given local- 
ity have eaten at a given time. At worst they 
show only part of what individual animals 
have eaten some place, some time...” and 
that “local application . .. is one of the pri- 
mary aims of food habits research.” Difficulties 
of extrapolation may require investigations 
within specific localities, sites, or ecologic com- 
munities if reliable and applicable data are to 
be obtained. ... Also, careful, complete descrip- 
tions of prevailing environmental conditions 
should be an important segment of any food- 
habits investigation. 
When is it Consumed? 
Studies of animal diets have frequently been 
limited in conception and in time. For example, 
samples obtained during a single season from 
hunter-killed animals only provide a fragmen- 
tary picture of diet. It should be recog- 
nized, however, that some studies, by design, 
have concentrated on only a single seasonal pe- 
riod of specific interest and have resulted in 
valuable contributions. Nevertheless, more 
studies crossing the entire seasonal spectrum 
are needed. Insight gained from knowledge of 
yearlong diet adds depth and may serve to de- 
fine problems previously unrealized. 
It is uncommon to find food studies repre- 
senting more than 1 or 2 years of research. 
Gullion (1966) criticized short-term studies as 
providing an insufficient understanding of 
food requirements under different and chang- 
ing conditions. Korschgen (1958) pointed out 
139 
