are taken from cattle. The time needed for 
collection depends upon the species, size of the 
fistula, rate of grazing, and type of forage. 
When materials are segregated, the whole 
sample is considered. For bird crops and some 
large animals, the various food items can be 
segregated and either weighed or measured by 
volume. With herbivores it is usually impracti- 
cal to separate the whole sample due to the 
large number of fine particles. The subsam- 
pling that is thus required has been ap- 
proached in different ways. Lesperance et al. 
(1960) describes taking 10 small grab samples 
at random from a collected sample and examin- 
ing each sample. Others (Heady and Torell 
1959; Van Dyne and Heady 1965) used at least 
half of the fistula sample, all in one 5- by 30- 
inch tray, to make their microscopic point 
analyses. The problems of subsampling are 
minimized by the point methods. Complete mix- 
ing of the sample before taking the subsample 
is the important consideration. 
A subsampling procedure is also necessary 
when using the microtechniques for Wiley- 
milled samples of fistula or stomach-content 
material and from small rodents and rabbits. 
To insure thorough mixing, the food is washed 
and stirred before small samples are taken for 
mounting on microscope slides. Each slide is 
limited to the amount of material it can hold. 
The general tendency is to use too much ma- 
terial. The specific amount will depend on the 
judgment of the investigator. When the Wiley 
mill is used, the amount can be quite uniform 
since the material is ground to uniform size. 
The number of microscope slides used has 
varied between workers. Ward (1960) and 
Myers and Vaughan (1965) used one slide for 
pocket gophers. Bear and Hansen (1966) and 
Sparks (1967) used one slide for jackrabbits. 
Malechek (1966) used two slides of steer 
rumen content that had been ground in a Wiley 
mill. Sparks and Malechek (1968) mounted 
subsamples on five slides in their study. The 
number of slides needed depends upon the var- 
iation between subsamples. If the material is 
mixed thoroughly, there should be no need to 
mount many slides. Procedures can be checked 
by comparing slides from one food sample. 
The method of estimating percentage dry 
weight in diets evaluated by Sparks and Male- 
chek (1968) recorded positive evidence for the 
presence of a plant species at a location on 
each slide. Frequency percentages (number of 
fields that the species occurred in of 100 loca- 
tions) were tabulated for each species. The re- 
quirements of random distribution and density 
of particles make it necessary to adjust the 
amount of material on a slide. The number of 
slides was determined by the difficulty of ob- 
taining 100 fields of view at 125X magnifica- 
tion without creating confusion. 
Subsampling is just as important in fecal 
analysis when using the microtechnique. Vari- 
ous procedures have been followed, and will 
vary according to the animal being studied and 
the time of year samples are collected. Consist- 
ency and shape of droppings change with the 
food being eaten. Sampling will have to be ad- 
justed to meet the requirements of obtaining a 
random subsample. 
Assessment and Presentation 
Food-habits data have been presented in 
many ways. The kind of data relates to the 
particular type of study. For foraging animals, 
the expressions of quantitative data are pre- 
sented as occurrence, volume, and weight. 
Some investigators have used the gravimetric 
method based on weight after drying the seg- 
regated materials under conditions of con- 
trolled temperature and humidity. This 
method, however, cannot eliminate all sources 
of error, and has the handicap of requiring ad- 
ditional time and expense. The use of precise 
—and costly—methods of analysis is unwar- 
ranted if equivalent precision or accuracy 1S 
not applied to other phases such as sampling 
and segregation. 
Volumes of discreet and readily segregated 
items may be measured by water displacement. 
Visual estimates of proportions are made the 
basis for percentage calculations for small, 
mixed, and fragmented materials. Whatever 
the method used, uniform procedure is impor- 
tant. Martin et al. (1946) compared the “ag- 
gregate percentage” and “aggregate volume”’ 
methods of summarizing data. Aggregate per- 
centage methods are used when volumes are es- 
timated, and percentages are the only figures 
available. Aggregate volume summaries can be 
used when volumes are actually measured. 
More recent findings by the microcopic 
point method have been converted to weight by 
regression. Stark and Malechek (1968) found 
that the percent composition based on dry 
weight of mixtures they used could be pre- 
dicted directly from the relative density. 
A common-sense approach to the interpreta- 
tion of the results of food habits studies is nec- 
essary. It often becomes impractical to satisfy 
the requirements of sample size to determine 
plant use to a precision level of 10 percent of 
the mean in all studies. Such things as individ- 
ual animal variation, the variable nature of 
grazing, the complexities that determine palat- 
ability, and the number and amount of plants 
used all affect results. Many times data with 
broader limits will meet the objectives. At the 
same time there is danger in using a small 
number of critically examined samples from 
only a few animals from a small range area to 
interpret food use over a whole range. Wildlife 
species that have freedom to select a wide vari- 
155 
