Techniques for evaluating fecal analysis were investigated using examples from studies on European waterfowl. Visual classification of droppings in the field on the basis of color or the presence of macroscopically identifiable material gave a good indication of the use of particular foods. Quantitative microscopic analysis techniques were investigated in detail. Identification of epidermis in waterfowl fecal material posed no great problems with relatively small numbers of plant species. Point sampling gives an accurate area measurement, and corrections can be made for differential fragmentation of plants and for differences in weight/surface area ratio of food items. Much work is needed before diet analysis can be accurately related to the energy budget of wild birds. J. WILDL. MANAGE. 39(2):271-279 The technique of identifying plant epidermal fragments in herbivore feces to identify the foods eaten is a very useful one for the study of protected species and has been extensively employed since Baumgartner and Martin (1939) used it in squirrel studies. Most interest in the method has been shown in studies of mammals, particularly ruminants (e.g., Martin 1964, Hercus 1960), and the method was described and tested in detail by Stewart (1967). He concluded that, although useful in recognizing the plant species ingested and indicating their relative abundance, the detailed quantitative work was laborious and difficult, owing to the large number of plant species in most habitats and the necessity for correcting for differential digestion of different plants. The technique has been used with success on animals ranging from locusts (Nomadacris septemfasciata) (Chapman 1957) to red grouse (Lagopus 1. scoticus) (Eastman and Jenkins 1970). In waterfowl, Lynch (1939) identified fragments of marine algae in the feces of baldpate (Anas americana), and other workers have used the technique for studying waterfowl foods (Gillham 1956, Ranwell and Downing 1959, Luther 1963, Pollard and Walters-Davies 1968, Owen and Kerbes 1971, Owen 1973). Grazing waterfowl digest vegetation relatively inefficiently, have rapid throughput rates, and often, especially in winter, feed in areas with a limited number of plant species. This paper assesses the usefulness of fecal analysis techniques in these situations and proposes methods for field and detailed laboratory analysis. All the droppings mentioned below refer to discrete solid feces. Cecal material is voided regularly by geese; this was not considered however, because cecal contents usually do not include food fragments. The evidence presented was collected during studies on grazing waterfowl caried out between 1967 and 1973 when I held a post at the Wildfowl Trust financed by the Natural Environment Research Council. I am indebted to the following people who in some way assisted with field or laboratory work: J. S. Adams, D. A. Cadwalladr, C. R. G. Campbell, L. Colley, R. S. Cook, G. B. Cross, K. Derrett, J. Frith, S. Gagnon, S. Macauley, J. V. Morley, and A. J. Stewart. I also thank many others, especially J. Kear and G. V. T. Matthews, for valuable discussion and advice. Useful criticisms of the manuscript were provided by G. V. T. Matthews and D. S. Ranwell. J. Wildl. Manage. 39(2):1975 271 This content downloaded from 157.55.39.224 on Sat, 27 Aug 2016 04:25:32 UTC All use subject to http://about.jstor.org/terms 272 FECAL ANALYSIS TECHNIQUE Owen