Diet selection is a central process in plant-herbivore interactions and is fundamental to understanding the ecological relationships between herbivores and their habitat. As well as understanding the reasons for observed patterns of diet selection by herbivores between plant species, it is important to determine why some individual plants within a species are heavily browsed while others remain undamaged. Two approaches can be used as a framework for the investigation of the selection of individual plants by herbivores. First, the relationship between the resources available to plants and the concentrations of secondary compounds in plants (and hence their palatability to herbivores) have been described by the carbon:nutrient balance (CNB) hypothesis (Bryant, Chapin & Klein 1983;Bryant et al. 1991). This hypothesis predicts that when nutrient levels are low, the products of carbon fixation accumulate relative to the levels of nutrients available for plant growth, and the concentrations of carbon-based secondary compounds such as phenolics and tannins increase. When nutrient availability increases, the allocation of fixed carbon to phenolics and tannins decreases (Bryant et al. 1987) as greater amounts of fixed carbon are used for growth. Similarly, under limited light conditions, the lower availability of fixed carbon is also predicted to lead to lower levels of carbon-based secondary metabolites.
Summary1. In addition to exhibiting preferences for particular plant species, vertebrate herbivores select particular individuals of these species whilst leaving others undamaged. This pattern of diet selection may reflect differences in the chemical composition (and hence nutritional quality) between individual plants, and/or variability in the physical constraints on intake rate, such as plant structure.2. An experiment was conducted to test the effects of environmental manipulations on the morphology and chemical composition of Sitka Spruce saplings, and to evaluate the consequences for herbivory by Red Deer. Fertilizing the trees increased tree height, branch span and leader length, twig width and needle width, and decreased the concentrations of total phenolics, condensed tannins, fibre and lignin but monoterpene content was not altered. Shading also reduced phenolic and tannin concentrations. 3. When the fertilized and shaded trees were offered to deer in feeding trials, the probability of a tree being visited by a deer and the biomass removed were influenced by tree morphological variables, as were the bite rate and intake rate of the deer. More biomass was removed from larger trees. 4. Once the effects of tree morphology had been taken into account, there was no effect of the fertilizer and shade treatments on deer browsing behaviour that could be attributed to changes in chemical composition of the trees. The relationship between intake rate and bite size at each tree varied between individual deer, but the functional response relationship between intake rate and bite size was not influenced by the treatments...