by animals and may indicate the relative pro- 
portions in which food items are consumed. 
However, the method is not without its limita- 
tions. 
The fact that certain foods are digested 
more quickly and more thoroughly than others 
creates problems in evaluation. Although this 
difficulty was recognized by early food-habits 
workers (cf. McAtee 1912), it remained 
largely unassessed. Davison (1940) compared 
food items found in the crops and gizzards of 
bobwhite quail and concluded that the crop 
was the only part of the alimentary canal in 
which food items remained in the same propor- 
tion as in the food eaten. Jensen and Korsch- 
gen (1947), however, found that even in crops, 
proportions of foods differed appreciably from 
those fed. 
A controlled-feeding study with sheep by 
Norris (1943) has been frequently cited to 
indicate limitations of stomach content analy- 
ses as a quantitative estimate of forages con- 
sumed. Differential digestion was apparent, 
and percentages of food items found in the 
stomach were poor estimates of items con- 
sumed. Only the larger, readily identifiable 
plant fragments were segregated and used, 
however; this procedure has more recently 
been criticized by Courtright (1959), Bergerud 
and Russell (1964), and Scotter (1966). 
Bergerud and Russell (1964) compared ru- 
men contents of four sacrificed caribou fed 
known rations from 30 minutes to 72 hours be- 
fore death. Their findings revealed that identifi- 
able fragments of some food items disappeared 
quickly from the rumen, and digestion rates 
varied between plant groups and to some ex- 
tent between species within groups. Plant frag- 
ments retained by sieves with a 0.078-inch 
mesh size were separated and used in calculat- 
ing digestion ratios. 
Limitations imposed by differential digest- 
ibility, although not directly investigated, have 
been recognized by other workers (Errington 
1932; Hartley 1948; Smith 1952; Cole 1956; 
Jensen 1958; Davison and Hamor 1960; Ed- 
wards and Ritcey 1960; Brown 1961; Martin 
and Korschgen 1963; Anderson et al. 1965; 
and Scotter 1967). 
Most investigators have accepted digestive- 
tract analysis methods as providing a reasona- 
ble estimate of at least the kinds of food con- 
sumed. The validity of even qualitative results 
has not been accepted without question. Rumen 
content analysis of caribou (Murie 1933) and 
mule deer (Anderson et al. 1965) did not re- 
veal food items known to have been taken. 
Partially digested foods are not only likely 
to have undergone substantial alteration of 
their original proportions, but they are also 
more difficult to identify. Hill (1946), Martin 
138 
(1949), Brown (1961), and Anderson et al. 
(1965) have mentioned limitations of identifi- 
cation. Dietary analyses are tedious and re- 
quire patience and considerable study. The 
quality and accuracy of results depend on the 
person(s) making the examination. Investiga- 
tor error in food-habits studies is apparently 
largely unassessed. 
Recognition of limitations imposed by diges- 
tive-tract analyses has promoted use of correl- 
ative methods. Dixon (1934), Deen (1938), 
DeNio (1938), Hahn (1945), Buechner (1950), 
Smith (1954), Wilkins (1957), Lovaas (1958), 
Brown (1961), Bishop and Hungerford (1965), 
and Chamrad and Box (1968), are but a few 
who have used field methods to complement 
laboratory analyses or, conversely, used diges- 
tive-tract material to supplement field observa- 
tions. 
Utility 
Imperfections of method and lack of preci- 
sion in results have not prevented digestive- 
tract examination from being a useful and 
widely applied means of investigation. Kalm- 
bach (1954, p. 276) maintained that “When- 
ever we are seeking the identity of food items 
. . Or whenever we aim to determine merely 
the presence or absence of particular items of 
diet, analysis of stomach contents is the only 
direct and reliable method of approach.” Rec- 
ognition that findings are only approximations 
(Martin 1949) does not nullify the value of the 
method. Under some conditions (Talbot 1962; 
Dirschl 1963), stomach analysis is the only 
method feasible. 
The method has been found applicable in 
several related, yet distinct kinds of studies. 
Evaluation or appraisal of the “economic” sta- 
tus of animals (McAtee 1933; Gross 1944; 
Lumsden and Haddow 1946), environmental 
damage (Cole 1956; Cole and Wilkins 1958; 
Ward and Keith 1962; Browning and Lauppe 
1964), predator-prey relationships (Errington 
1932; Murie 1935; Robinette et al. 1959), ani- 
mal competition (Halloran and Kennedy 1949; 
Davis 1952; Morris and Schwartz 1957; Ju- 
lander 1958; Sparks 1968), life histories (Dix- 
on 1934; Buechner 1950; Murie 1951; Loveless 
1959), ecological relationships (Rasmussen 
1941), environmental manipulation (Martin 
and Uhler 1939; Hungerford 1957; Biswell 
1961), food palatability (Bellrose and Ander- 
son 1948; Cowan 1945; Hill 1946), animal 
preference (Deen 1938; Hill and Harris 1948; 
Talbot 1962; Chamrad and Box 1968), pois- 
onous foods (Forbes and Bechdel 1931; Buech- 
ner 1950; Bak and Lewandowski 1959), ani- 
mal condition (Harris 1945; Anderson et al. 
1965), food quality (Lehmann 1958; Bissell 
1959; Klein 1962), dependence on supplemen- 
