Master of Science
Charles R. Blem
Although the adaptive significance of daily and seasonal fattening of birds has been studied by many investigators (see Odum, 1965; King, 1972; Blem, 1976a), surprisingly little is known about the interactions of proximate and ultimate environmental factors regulating avian lipid levels. An exception is the study by Evans (1969). He found that lipid levels in Yellow Buntings (Emberiza citrinella) were more closely correlated with long-term temperature averages than with temperatures of the day of capture, or the days preceding or following capture. This is an indication that temperature may be more important as an ultimate control of fattening, through natural selection, than as a proximate cue. The relative effects of temperature, photoperiod, and morphological variables on daily lipid levels have not been assessed simultaneously in any species. Mid-winter fattening appears to be a widespread phenomenon in small birds of the temperate zone, but it is well known for only a small number of species (see King, 1972; Blem, 1976a; for reviews). It is known that lipid levels fluctuate and the amounts stored vary with weather conditions, both daily and seasonally. The amplitude of the daily cycle and the magnitude of lipids stored are maximized during mid-winter. Most of the birds have energy reserves for only a few hours of activity in the morning. It has also been found that lipid stores are not greater during favorable weather conditions. This suggests a balance between lipids stored for energy production and energy required for food gathering activities, risks from predation and lethal temperature extremes.
Statistical models of the relationship between environmental variables and the amount of lipid reserve began with research by King and Farmer (1966) and Evans (1969). Their results emphasized the importance of climatic variables as both proximate and ultimate factors (also see Vincent and Bedard, 1976). These analyses involved only sim-ple regression techniques or restricted multivariate models involving only a few or single dependent variables. More sophisticated multiple regression techniques and modern computer implementation provide a means of determining the relative importance of several independent variables in the prediction of a dependent variable. This type of analysis is particularly suited to certain aspects of fat deposition in sparrows.
Multiple regression analyses provide two potentially important pieces of information about independent variables. The coefficient of determination (R2) is an estimate of the amount of variability explained by the multiple regression coefficient (Zar, 1974) and the standard partial regression coefficients indicate that relative importance of independent variables in the prediction of a dependent variable. Such analyses do not insure that all significant variables have been included or even considered, however, such variables may be later added to the model. In this study the relative importance of a wide range of variables will be investigated.
The purpose of the present analysis is threefold: (l) Firstly, an attempt has been made to develop predictive equations which might be used to quantify lipid reserves in passerine birds from measurements of living specimens. Specific points analyzed herein include: (a) how accurately might such equations predict lipid content, and (b) how many birds must be included in the analysis to obtain maximum accuracy. (2) The second emphasis of this study is the identification and comparison of important independent variables and a comparison of the relative importance of morphometric variables with environmental measures. (3) Finally, a comparison will be made of the relative importance of various temperature measures of the day of capture with long-term averages in an attempt to assess the relative selective importance of prolonged temperature regimes.
Estimates of lipid content are useful for field studies of pre-migratory fat deposition, the energetics of overnight survival, or any life history phenomena where storage and utilization of energy is crucial. Variables will be measured that will provide relatively precise predictions of the fat content of birds. The success of this attempt will be primarily indicated by the coefficient of determination (R2). Hopefully, models will be generated that will provide a method for studying lipid deposition cycles without tedious fat extraction processess or having to sacrifice large numbers of wild birds.
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission
Scanned, with permission from the author, from the original print version, which resides in University Archives.