Others measure the mass of finer items and re- 
late the proportions of coarse identified mater- 
ial to this quantity, assuming that the fine 
material is made of the larger items in the 
same relative amounts (Norris 1943; Saunders 
1955). Usually the volume or weight of this 
unsegregated fine material is measured, and 
the proportions of its constituent items are es- 
timated visually (Korschgen 1962). To help in 
this estimate, it is often advisable to separate 
small samples into small individual piles of 
known materials. 
Dirschl (1962) tested the results of screen- 
ing antelope rumen contents through 5.66, 
4.00, and 2.83-mm. sieves, separating the ma- 
terials on each screen into species, and then 
weighing them and determining percent com- 
position. He found very little difference in the 
mean composition between these three mesh 
sizes. The use of the 5.66-mesh size saved con- 
siderable time. Sieve mesh sizes of 0.078, 
0.0328, and 0.0164 inch were evaluated in the 
analysis of Newfoundland caribou rumen foods 
(Bergerud and Russell 1964). They concluded 
that the larger plant fragments were not rep- 
resentative of the entire rumen contents be- 
cause of differential digestion between plant 
groups. Scotter (1967) tried to determine 
which of three sieve-mesh sizes could best be 
used to determine composition of forage sam- 
ples from 20 barren ground caribou rumen. He 
found that data obtained from any one mesh 
size could be inaccurate and misleading. The 
proportion of lichens generally increased as 
the mesh size decreased, and grasses increased 
as the mesh size increased. He combined data 
from three screens for each sample. 
Norris (19438) analyzed stomach contents of 
sheep fed controlled diets. There were some se- 
rious variations in his results, and these are 
often cited. There were also weak points in his 
method that should be pointed out. No finely 
comminuted particles which passed through a 
window screen were identified or separated, 
and only the particles that were readily iden- 
tifiable on the screen were used in the analysis. 
A representative sample of 2 percent of the 
total on the screen was taken, and all the read- 
ily identifiable particles were picked out with 
forceps. As a result, over half of the stomach 
contents consisted of material too fragmentary 
for separation or identification, and these were 
not used in the analysis except that it was con- 
sidered to have the same composition as the 
larger, recognized and weighed material. This 
large amount of unidentified material could 
have accounted for the wide variability in the 
results. 
Another point that should be considered is 
the difficulty and added work of basing all 
analysis on a weight basis. Digestion starts as 
150 
soon as the plant is eaten. The readily digested 
materials are immediately dissolved. They are 
usually in the form of liquids or fast-dissolving 
substances that do not account for a large vol- 
ume, but are heavy. They would not introduce 
a great amount of error, but should be consid- 
ered in trying to justify the added work of 
drying materials for a weight measurement. 
Analyses using very little magnification 
have been used by many biologists. All of the 
early workers in the Fish and Wildlife Service, 
U.S. Department of the Interior, recognized 
the limitations imposed by this method, and 
they promoted the use of field methods to sup- 
plement the results of laboratory examina- 
tions. The usual procedure is to save the fine 
particles and to attempt to identify and segre- 
gate a small subsample to at least classify the 
material into the broad groups such as grass, 
forbs, or browse. Some workers examined this 
fine material under higher magnification to de- 
termine if they were missing any important 
food items that were going through the 
screens. Although the method has limitations 
and the results are approximate, the technique 
is still widely used and provides useful infor- 
mation. 
There is not much in fecal material that can 
be recognized without the use of magnification. 
For predators and reptors, many food items 
can be recognized from tooth or bone frag- 
ments. Herbivore feces are too fine for identifi- 
cation unless the food material is something 
like whole grains whose hull comes through 
the animal in large pieces. 
Identification Under Low-Power Magnification 
Low-power magnification is also used in the 
sereening methods of segregation. Magnifica- 
tion was increased in an attempt to identify 
and separate the fine materials. As better 
equipment became available, the investigator 
classified more of the finer materials. This not 
only eliminated a lot of error in the study of 
larger herbivores, but it permitted us to begin 
work on the smaller animals. 
Heady and Torell (1959) introduced the 
terms “laboratory point” or ‘microscopic 
point” in their method of determining botani- 
cal composition on clipped plots and fistula 
contents of sheep from the annual type range 
on the Hopland Field Station in Mendocino 
County, Calif. They did not state what magni- 
fication they used, but in a report of similar 
work done later by Van Dyne and Heady 
(1965), it was stated than an 18-power binocu- 
lar microscope was used. The microscope 
method was developed so that the green mate- 
rial from clipped plots and fistulas could be 
sampled by the same procedure. Eight subsam- 
ples were taken from each fistula sample. Each 
SE 
