Body Weights of Birds: A Review

Although body weights are relevant to many branches of avian biology, they have not been recently reviewed.I here survey sources an d uses for weight data, with reference to such variables as time of day, season, year, sex, migration, geography, climate, buoyancy, food size, and habitat.In addition, I consider uses of weights as standards for body size, in systematics, in assessing the physiological condition of individuals, and in analyses of ecological communities.Excellent previous reviews of this subject by Nice ( 1938), Baldwin and Kendeigh ( 1938), Amadon ( 1943), Fisher ( 1955), Blake ( 1956), and Perrins (in Thomson 1964) do not cover many recent discoveries and interpretations.OBTAINING WEIGHTS Pettingill (1970:446), Bub (1967:61-63), and Giles ( 1963 : 4-5) outlined weighing techniques in which beam balances or small spring scales are used (Berger 196S), with live birds usually being confined in tubes or chambers for weighing.Records of age, sex, fat levels, time, locality, and other variables are useful in analyzing weights.Baldwin and Kendeigh ( 1938), Owen ( 1962)) and others have properly emphasized the need for large samples of weights in comparing similar populations.For some studies weights need to be taken as uniformly as possible, e.g., sampling diurnally feeding species early in the morning to minimize variation from daily weight cycles (Kontogiannis 1967).Breeding birds are best compared at known stages, because some species (e.g., certain penguins) vary greatly in weight through the breeding cycle, and females of many species vary considerably during the period of egg laying.After being banded, many, but not all, migratory birds lose weight for a day or two (Leberman and Stern 1977).This phenomenon can, of course, be detected only in recaptured individuals who are ordinarily a minority (commonly less than 20%) of the migrants initially captured.The cause of weight declines in many recaptured migrants is not well understood, and the reasons for the individual variations are unknown.Possible causes of the weight decreases include ( 1) illness or other abnormalities, (2) recovery time for migrants that have completed particularly stressful flights, (3) failure to locate food in an unsuitable or unfamiliar area, (4) failure to obtain enough food due to social subordination (Rappole and Warner 1976), (5) reduction in weight that is adaptive for prolonged stopovers, (6) loss of opportunity to feed through time spent in nets or being processed by banders, and (7) handling shock (Mueller and Berger 1966, Leberman and Stern 1977).Determining the relative contribution of these factors in any particular case can be difficult.Further attention should also be given to the possible occurrence of weight decreases due to handling of nonmigratory birds (Rogers and Odum 1966).To avoid possible effects of handling, wild birds might in some cases be weighed without capture by use of weighing perches (Moore 1953, Ellis and Varney 1974) and uniquely marked with dye applied by remote control.Ornithologists have assumed that a freshly killed bird weighs the same as a live one, but this may not be valid.Von Brijckel (1973) weighed 237 Garden Warblers (Sylvia horin) alive and 80 others within 15 min after death and found that the dead birds averaged more than 1 g (5.5%) lighter.Possible differences in weight between life and death need more study.Under ordinary laboratory conditions dead birds gradually lose weight over a period of hours by desiccation.However, dead birds that are weighed, then frozen in sealed plastic bags, kept in humid conditions, and later thawed and reweighed, lose relatively little weight (2% or less) over a period of weeks (Grant 1965, Holmes 1976).Researchers sometimes cannot directly obtain weight data and therefore turn to the literature.Although weights have been published for thousands of species, they are often difficult to find.Examples of publications listing many North American weights are Irving ( 1960), Hartman ( 1961)) and Clench and Leberman (1978), but others are too numerous to mention.Brief summaries of weights should ideally include the sources of data, the mean, standard deviation, range, and sample size for ten or preferably more adults of each sex in specified conditions, e.g., early in the day at a particular stage in the breeding season.Samples from representative geographic localities are also desirable.