Qualitative Identification of Forage Remnants in Deer Feces

Abstract

Seventeen tree, shrub, and herbaceous plant species were individually fed to captive mule and white-tailed deer (Odocoileus hemionus and 0. virginianus). Remnants in feces were identified foi all species by magnification varying from 7x to 100X. At best, only low percentages of recognizable material were found despite a single species diet. Unstained epidermal mounts prepared from fresh plant material and photomicrographs required less time and proved superior to staining techniques for identification. Ruminant food habit studies have emphasized direct observation of feeding and analyses of stomach contents and samples taken from esophageal fistulas. Fecal analysis, with the advantage of unlimited sampling, could supplement these methods. This study adapts and extends to deer, procedures developed for fecal studies of birds and small mammals. Specifically, this study reports on forage plant structures that can be qualitatively identified in deer feces, and the development of a useful technique for such analyses. Foremost among the several cooperators who have contributed to this study are D. J. Neff, Arizona Game and Fish Department, Flagstaff, and J. L. Tinker of the Phoenix Zoo. REVIEW OF THE LITERATURE Fecal studies of mammalian carnivores and birds have been used for many years to determine diet composition. These studies largely restricted identification to gross undigested structures, such as chitinous insect exoskeletons, fur, bones, seeds, and pine needles. Herbivorous mammals often masticate 1 Forest Service, U. S. Department of Agriculture. Central headquarters maintained in cooperation with Colorado State University, Fort Collins. Research reported here was conducted at Tempe, in cooperation with Arizona State University. 506 and degrade food so finely that histological microtechniques are necessary for species identification. The first widely-used technique was developed for analysis of squirrel stomach contents (Baumgartner and Martin 1939). Material in stomachs was compared with permanent reference slides of stained leaf and stem epidermis from plants in the study area. Later, Dusi (1949, 1952) adapted this technique for fecal analysis of cottontail rabbits (Sylvilagus floridanus). Adams (1957) modified a technique used by Scott (1941) on red fox (Vulpes sp.) for analysis of snowshoe hare (Lepus americanus) diets. Hares were fed a stock diet plus a forage plant. A species constant for the forage plant was ascertained as a factor correlating the number of recognition items in the excreta to the amount of forage plant eaten. More sophisticated statistical methods of analysis were later developed (Adams et al. 1962). Gross have been recognized in deer feces. Adams (1957:156) found bitterbrush (Purshia tridentata) leaf and stem fragments in mule deer pellets. Remains of various acorns and seeds were identified from whitetailed deer fecal material in Texas (Lay 1965). A European study of deer damage to forest trees utilized fecal examination (Burckhardt 1959). Storr (1961), in Australia, fed perennial This content downloaded from 157.55.39.54 on Thu, 30 Jun 2016 03:59:21 UTC All use subject to http://about.jstor.org/terms QUANTIFYING FORAGE REMNANTS IN DEER FECES Zyznar and Urness 507 and annual dicotyledonous plants to penned quokkas (Setonix brachyurus), marsupials with ruminant-like digestive systems. He concluded (p. 161) that epidermis survived digestion by virtue of the encasement of entire cell walls in cutin, and that perennial dicots showed no differential breakdown. Succulent annuals, however, survived digestion very poorly. Hegg (1961) made microscopic counts of cuticle and epidermal fragments found in red deer (Cervus elaphus), roe deer (Capreolus capreolus), and chamois (Rupicapra rupicapra) fecal samples collected in the Swiss National Park. He found that such counts gave satisfactory data on diet composition. However, a more thorough study in Uganda and Kenya by Stewart (1967) contests this conclusion. He fed perennial grasses to a number of wild ungulates. Stewart (1967) questions the validity of remnant counts because some species fragment more readily than others and consequently would appear more important, although intake was the same or less. Statistical support is presented for using point quadrat or area cover to estimate frequencyof-species occurrence. He further states (p. 100) that perennial grasses consistently forming more than 5 percent of the diet can be quantitatively identified in any fecal sample with standard errors within 10 percent of the expected mean. Kiley (1966) used epidermal characteristics to qualitatively identify individual grass species and some dicotyledonous plants from waterbuck (Kobus sp.) feces in Uganda and Kenya. Difficulties in quantitative analysis were implied but not discussed.