It is proposed that the composition of waterfowl diets would be expressed better in terms of metabolizable energy (ME) than in volume or weight. The use of formulas for estimating ME from chemical data of waterfowl foods is suggested. Average ME of iree barleys fed to captive mallards (Anas platyrhynchos) was 3.173 keal/gram dry matter; of four wheats, 3.526 keal/gram dry matter; and of faX rye7 3.336 keal/gram dry matter. Values for barley and wheat are comparable to published values detennined for chickens; the value for rye fed to mallards is higher than values published for chickens. METABOLIZABLE ENERGY OF MALLARD DIETS * Sugden 781 SOWLS, L. K. 1955. Praine ducks: a study of their behavior, ecology and management. 1st ed. The Stackpole Company, Harrisburg, Pennsylvania, and the Wildlife Management Institute, Washington, D. C. 193pp. TRAUTMAN, C. G. 1960. Evaluation of pheasant nesting habitat in eastern South Dakota. Trans. N. Am. Wildl. and Nat. Resources Conf. 25 ao2-2l3. WALTERSON, J. E. 1966. Land use study. North Dakata Outdoor Recreation Agency. llSpp. WESTE:ETSKOV, K. 1950. Meiods for determining the age of game bird eggs. J. Wildl. Mgmt. 14(1):67. WOLFE, G. W., JR., AND R. D. EVANS. 1967. CharacteAstics of roadside phegant nests. Nebraska Game and Fish Dept. P-R Rept., PrOjeCt W_2FR. 14PP. Received for publication Febrry 8, 1971. LINDER, R. L., D. L. LYON, AND C. P. AGEE. 1960. An analysis of pheasant nesting in south-central Nebraska. Trans. N. Am. Wildl. and Nat. Resources Conf. 2S:21s229. MARIZ, G. F. 1967. Effects of nesting cover removal on breeding puddle ducks. J. Wildl. Mgmt. 31(2):23S247. MILONSKI, M. 1958. The significanee o$ farmland for waterfowl nesting and techniques for reducirlg losses due to agricultural practices. Trans. N. Am. Wildl. Conf. 23:219227. MOYLE, J. B. (ed.) 1964. Ducks and land use in Minnesota. Minnesota Dept. C:onserv., Div. Game and Fish Tech Bull. 8. 140pp. PAGE, R. D. 1971. Waterfowl nesting on a railroad right-of-way in North Dakota. J. Wildl. Mgmt. 35( 3 ):54S5S0. SHICK, C. 1952. A study of pheasants on the 9,000-acre prairie farm Saginaw County, Michigan. Michigan lDept. Conserv., Game Div. 134pp. Knowledge of waterfowl diets is based largely on proportions of food items found in their digestive tracts. Results are usually expressed as percentages of volume, weight, or frequency of occurrence, but these measurements do not show the relative nutntive or energy values of each food item. The primary factor governing food intake by chicks aE>pears to be the need for energy ( Hill and Dansky 1954, Peterson et al. l9S4). There is general agreement among poultry nutritionists that the energy value which it gives rise (Hill 1964). Compared with other measurements of energy for a given food and species of animal, ME is relatively unaffected by age, rate of growth, level of food intake, wide differences in diet balance, breedy and sex (Hill 1964). Thus, it seems logical that the colmposition of waterfowl diets would be expressed better in terms of ME than in volume, weight, or frequency of ocourrence. Formulas for estimaig ME have been derived from chemical data of chicken of a food is best e;xpressed as the ME to foods, eliminating the need for biological This content downloaded from 40.77.167.35 on Fri, 21 Jul 2017 14:53:46 UTC All use subject to http://about.jstor.org/terms 782 Jo1lrnal of Wildlife Managem^t, Vol. 3S, No. 4, October 1971 assay (CarpenteEr and Clegg l956, Sibbald et al. 1963). Use of such formulas for predicting ME of waterfawl foods would have another advantage, because it is not practical to collect sufficient quantities o£ some foods, particularly invertebrates, to make reliable feeding trials. Whether formulas derived for chickens are applicable to ducks depends on the relative abilities of the two species to metabolize foods. This paper describes experiments to measure ME in eight small cereal grains fed to mallards. Results are compared with ME values derived for chickens. I acknowledge the helpful advice provided by my colleagues, J. B. Gollop and E. A. Driver, and the editorial assistance of W. J. D. Stephen, Canadian Wildlife Service. MATERIALS AND METHODS Mallards, hatched from eggs of wild birds, were used in the experiment. The birds were 18-20 weeks old when first used. A feeding trial involved 12 groups of two ducks each, with 6 groups of each sex. Each group was housed in a 26x 26x 15inch pen in a finishing brooder. Each pen had a water trough at the front and a food trough at the rear. The experiment was conducted in an air-conditiolned room with a constant temperature of 20 C. An automatic control system regulated the lights to provide a 14-hour photoperiod. A feeding trial lasted 10 days. Food intake was measured and exereta were collected during the last 4 days. The ducks were weighed to the nearest 10 grams on the third day of the bial and again at the end of the trial. They were not weighed the day collection started, to avoid the disruption of feeding observed on previous occasions. Because the test diet was fed 3 days prior to the first wleighing, it was assumed that the rate of weight change during the 4-day collection period was similar to the rate measured for the 7 days. One to 3 weeks separated the trials. Between trials? to compensate for the lack of essential nutrients in the cereal grains being tested the ducks were fed a mixture of nmmercial duck pellets and the next test grain. Certified seed grains were used in the experiment. Samples of feed taken during the collection period were preserved in sealed jars and later ground and analyzed for crude protein ( nitrogen x 6.25 ), crude fat ( ether extract), crude fiber, ash, and dry matter using procedures of the AOAC (EIorwitz 1960). These analyses were contracted to the Feled Testing Laboratory, Department of Animal Science, University of Saskatchewan. Gross energy (GE) was measured by oxygen bomb calorimeter. To measure dry-matter intake, the uneaten foo{l at the end of the collection period was weighed, ground, and analyzed for moisture. The resulting estimate of dry matter was deducted from the dry matter fed. Fooid spilled on trays was reduced to dry matter aIld likewise deducted. The small amount of grain spilled on the floor was treated similarly, and the weight was divided equally among the 12 pens. Exereta collected oll stainless steel trays under screened floors were removed once daily and preserved by freezing. Because some exereta became mixed with water spilled by the ducks, it was also necessary to collect the water from the trays. The mixture was refrigerated until the end of the collection period, when it was weighed and homogenized in a blender. Duplicate samples were removed for dry-matter determination. The estimated amount that combined with the water was added to the tatal dry weight of exereta for calculating the ME. The frozen exerement was dried This content downloaded from 40.77.167.35 on Fri, 21 Jul 2017 14:53:46 UTC All use subject to http://about.jstor.org/terms Toble 1. Averoge live weights and intake of dry motter by mollards.