1254394699

105

Figurę 4.1 Seasonal variation in body composition throughout the year. Data are least square mean of lean dry mass of A) liver, B) kidneys, Q intestines, D) gizzard, E) size-independent leg muscles, F) size-independent pectoralis muscles, G) supracoracoideus muscles, H) size-independent skeletal muscles, I) heart, J) lungs, K) brain and L) size-independent skin. Least square means were obtained from a GLM including period, sex, period*sex as independent variables and body size as covariate. DifFerent letters represent significant difference between periods.

Table 4.1 Matrix of correlation showing all comparisons between lean diy masses of organ groups.

	Muscles	Heart+Lungs	Uver+Kfdney	Intestine+Gizzard Skin 1
Heart+Lungs	0.55***
Liver+Kidneys	0.31**	0.38***
Intestine+Gizzard	0.13	0.26*	0.08
Skin	0.25*	-0.10	0.28*	-0.35***
Brain	0.38***	-0.07	0.22	-0.05 0.39***

*** p < 0.0001; ** p < 0.001; * p < 0.05

Since body mass contains the mass of all organs, statistically controlling metabolic ratę for body mass in analyses testing for relationships between body components and metabolic ratę would be conceptually biased. We therefore first tested for the effects of sex and body size, rather than body mass, on BMR and Msum using a GLM. This approach considers potential