Temporal Changes in Winter Diet Selection by White-Tailed Deer in a Northern Deer Yard

Abstract

We followed short-term changes in the winter foraging behavior of yarded northern white-tailed deer (Odocoileus virginianus) over 3 winters to evaluate the hypothesis that deer forage nonselectively. We used a sequential-clipping, browse-monitoring technique to measure browse use and availability at biweekly (1983-84) or weekly (1984-85, 1985-86) intervals. From initial diets composed of 1-4 browse species, deer expanded their diets to include up to 16 out of 19 species by the end of each winter. Rate and magnitude of diet expansion were influenced by the degree of browsing pressure in the study area. We found partial preferences in diet choice, but important diet items were used in nonrandom runs once adopted into the diet. Proportional representation of all browse species in the diet varied on a short-term basis. Three dominant species, red-osier dogwood (Cornus stolonifera), pussy willow (Salix discolor), and white ash (Fraxinus americana), exhibited gradual compensatory shifts in proportional use over each winter, which reflected early season preferential selection of red-osier dogwood. During the third winter, browsing was nonrandom and selective, and temporal changes in patterns of browse use occurred. Forage ratios and relativized electivity indices of the 3 dominant species also changed gradually, indicating sequential shifts in preference for redosier dogwood, pussy willow, and white ash, respectively. Overall, deer foraging in the winter yard was dynamic, with high diet selectivity at the beginning of the yarding season and gradual diet generalization as browse was depleted in the yard. J, WILDL. MANAGE. 55(3):361-376 In winter, white-tailed deer in north-temperate regions congregate in deer yards, which provide shelter during periods of deep snow and inclement weather (Telfer 1967, Rongstad and Tester 1969, Marchinton and Hirth 1984). However, deer yards only represent a small proportion of the total annual range, resulting in high population densities in areas of limited food availability. Yarded deer meet their energy and nutrient requirements from 2 main sources: catabolization of limited body fat and protein reserves (DelGiudice et al. 1988) and ingestion of woody browse (Mautz 1978). Prolonged confinement of deer to winter yards causes the gradual depletion of the browse supply, often resulting in high deer mortality by starvation late in the yarding season. The consumption of browse in winter slows the rate of tissue catabolization, a process that is critical to survival in northern regions (Mautz 1978). Efficient use of limited browse resources during periods of thermal stress is evolutionarily advantageous; presumably, the selective pressures of overwintering have contributed to the evolution of a particular foraging strategy in deer which is adapted to food-limited situations such as those in a northern deer yard. Several authors have discussed foraging behavio of large ungulates in food-limited environments from a theoretical standpoint. Westoby (1974) and Belovsky (1978) proposed that herbivore diet specialization should occur when food resources are abundant, and diet generalizati n should occur at low food levels. Using data from Wetzel et al. (1975), Nudds (1980) suggested that a shift in foraging strategy from specialist to generalist occurs as winter food resour es are depleted. Though criticized on methodological grounds by Jenkins (1982), Nudds (1980) posed a pertinent question about possible changes in foraging strategy adopted by deer in the face of ever-diminishing winter food resources. To date, most winter field studies of deer foragi g have examined total use of foodstuffs within a given habitat over the entire yarding season. Standard browse inventory techniques involve establishing transects or sample plots at a few critical times in the year (Gysel and Lyon 1980), usually fall and spring, to assess winter browse use. LaGory et al. (1985) employed seasonal totals of browse use to plot relative browse use over an entire winter as a function of estimated relative browse availability at the beginning of the winter. They found significant positiv linear regressions between use and